{"Bibliographic":{"Title":"Proceedings of Workshop on Priority Great Lakes Environmental Research Laboratory, Ann Arbor, Michigan","Authors":"","Publication date":"1975","Publisher":""},"Administrative":{"Date created":"08-17-2023","Language":"English","Rights":"CC 0","Size":"0000290251"},"Pages":["GB\n1627\nPROCEEDINGS OF WORKSHOP ON\nG8\nPRIORITY\nW6\n1974\nGREAT LAKES ENVIRONMENTAL RESEARCH\nINITIATIVES\nHELD OCTOBER 10-11, 1974,\nAT THE\nGREAT LAKES\nENVIRONMENTAL RESEARCH LABORATORY\nAnn Arbor, Michigan\nDr. Eugene J. Aubert, Program Chairman\nDr. Arthur P. Pinsak, Proceedings Editor\nauthorization\nOF\nCOMPENSATION\nU.S. DEPARTMENT OF COMMERCE\nNational Oceanic and Atmospheric Administration\nWHO\nAvenue\nEnvironmental Research Laboratories\nSTATES OF\nAPRIL 1975","GB\n1627\nG8W6\n1974\nAND ATMOSPHERIC\nU.S. DEPARTMENT OF COMMERCE\nRogers C. B. Morton, Secretary\nNOAA\nNATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION\nRobert M. White, Administrator\nS DOCUMENT OF\nENVIRONMENTAL RESEARCH LABORATORIES\nWilmot N. Hess, Director\nProceedings of Workshop on\nPriority Great Lakes Environmental Research\nInitiatives Ann Arbar, mich., 1974.\nHeld October 10-11, 1974\nat the Great Lakes\nEnvironmental Research Laboratory\nAnn Arbor, Michigan\nDR. EUGENE J. AUBERT,\nProgram Chairman\nDR. ARTHUR P. PINSAK,\nProceedings Editor\nGLERL Contribution No. 26\nApril 1975\nMARINE AND EARTH\nSCIENCES LIBRARY\nMAY 4 1976\nN.O.A.A.\nU. S. Dept. of Commerce\n16 2037","DISCLAIMER\nThe Environmental Research Laboratories do not approve,\nrecommend, or endorse any proprietary product or proprietary\nmaterial mentioned in this publication. No reference shall\nbe made to the Environmental Research Laboratories or to this\npublication furnished by the Environmental Research Labora-\ntories in any advertising or sales promotion which would in-\ndicate or imply that the Environmental Research Laboratories\napprove, recommend, or endorse any proprietary product or\nproprietary material mentioned herein, or which has as its\npurpose an intent to cause directly or indirectly the adver-\ntised product to be used or purchased because of this Envi-\nronmental Research Laboratories publication.\nii","FOREWORD\nA workshop was held on October 10-11, 1974, at the Great Lakes Environ-\nmental Research Laboratory (GLERL) of the National Oceanic and Atmospheric\nAdministration (NOAA) in Ann Arbor, Michigan, to identify future priority Great\nLakes environmental research initiatives for GLERL. The International Field\nYear for the Great Lakes (IFYGL), an ongoing major multidisciplinary research\ninitiative for which NOAA has lead agency responsibility, has completed the\ndata collection phase; research, analysis, and simulation will continue until\napproximately June 1977. The following question thus becomes timely: What\nGreat Lakes environmental research should be pursued by GLERL as a follow-on\ninitiative to IFYGL on an in-house and contract basis, through multiagency\ncoordination, and with a possibility for joint United States-Canadian partici-\npation?\nCentral to the GLERL mission is the development of improved methods of\nenvironmental simulation and prediction in its broadest sense. Several sugges-\ntions have been made for future GLERL research initiatives which invoke data\ncollection, analysis, and modeling of nearshore environmental dynamics, near-\nshore and lake-scale water movements, aquatic ecology, environmental dynamics\nof Lake Michigan, Great Lakes water levels, flows in connecting rivers, and\nflooding. Deliberation needs to be given to these and other subjects to arrive\nat the most pertinent research priorities both from the scientific viewpoint\nand from the aspect of environmental information to support Great Lakes research\nmanagement.\nThe workshop was convened for the following purposes:\n(1) To identify future Great Lakes environmental research initiatives\n(i.e. major research programs of the multimillion dollar, 3- to 5-\nyear duration variety--beyond the normal GLERL resource) required\nto provide a satisfactory state-of-the-art in environmental simula-\ntion and prediction to support the decision process for Great Lakes\nactivities.\n(2) To provide the university research community an opportunity to\ndiscuss and recommend future Great Lakes environmental research\ninitiatives.\n(3) To consider possible United States-Canadian joint research initiatives.\n(4) To identify logical research follow-ons to IFYGL.\n(5) To provide background for subsequent development of a preliminary\nresearch plan by the GLERL staff. This plan will be coordinated\nwith other agencies as appropriate to consider joint research\ninitiatives. The ensuing GLERL program document will be submitted\nthrough NOAA channels for support in the FY 77 budget.\nThe workshop format included a plenary session with perspectives and\nstructured response in eight fields followed by five work group sessions and\niii","a final plenary session. The principal speakers in the plenary opening session\ntook stock of the accomplishments and deficiencies of Great Lakes environmental\nresearch in IFYGL as well as in other research programs or projects in order\nto set the stage for the workshop discussion sessions. Of concern is what has\nbeen learned and what are the proper scientific questions that should be asked,\nwhat research objectives are now logical, and what achievable products can be\ndefined to meet these objectives. Of concern also are the user needs for\nenvironmental information. The responders either presented their views on the\ntopic to expand the perspective given by the principal speaker and/or reviewed\nmajor points raised by the principal speaker. With the background provided in\nthe plenary session, the individual work groups discussed and identified\nresearch initiatives in terms of the following guidelines:\n(1) Discuss the state-of-the-art of simulation and prediction and identify\nthe research required to further it.\n(2) Discuss IFYGL and other research programs/projects in terms of research\naccomplishments, deficiencies, and the logical next research step.\n(3)\nIdentify scientific questions, objectives, and products.\n(4) Identify user needs for improved environmental information.\nGenerally consider methods of approach related to the research\n(5)\nsequence (data collection, analysis, modeling, evaluation).\nEach work group developed a priority listing of recommendations which was\npresented at the final plenary session for discussion and reached consensus\non a coordinated listing of priority research initiatives.\niv","CONTENTS\nPage\nINTRODUCTION\n1\n1. PERSPECTIVES OF GREAT LAKES RESEARCH\n3\n1.1\nView\nof IFYGL Research\n3\n1.1.1\nResponse\n9\n1.1.2\nResponse\n10\n1.2 Water Movements: Simulation of Lake-Scale\nCirculation\n15\n1.2.1\nResponse\n21\n1.2.2\nResponse\n22\n1.3\nDiscussion\n23\n1.4\nWater Movements: Nearshore Circulation\n26\n1.4.1\nResponse\n32\n1.4.2\nResponse\n34\n1.4.3\nResponse\n36\n1.5 Simulation of Aquatic Ecology\n38\n1.5.1\n43\nResponse\n1.6\nSimulation of Water Quality\n45\n1.6.1\n58\nResponse\n1.6.2\n63\nResponse\n1.7 Simulation of Environmental Dynamics of the\n64\nGreat Lakes\n71\n1.7.1\nResponse\n76\n1.7.2\nResponse\n1.7.3\n78\nResponse\n1.8 Discussion\n79\n1.9\nLake-Atmosphere Boundary Layer Processes of\n80\nLarge Lakes\n1.9.1 Response\n89\n1.9.2 Response\n91\n1.10 Simulation of Great Lakes Water Levels and Flows\n93\nin Connecting Channels\n1.10.1 Response\n99\n101\n2.\nWORKSHOP DISCUSSION SESSION\n101\n2.1\nGuidelines for Work Groups\n106\n2.2\nRecommended Research Initiatives\n106\n2.2.1\nWater Movements\n2.2.2\nAquatic Ecology and Water Quality\n108\n113\n2.2.3\nLake-Atmosphere Interactions\n116\n2.2.4\nEnvironmental Dynamics\n119\n2.2.5\nWater Levels and Flows\n120\n3. PRIORITY RESEARCH INITIATIVES\n127\nAPPENDIX A. WORK GROUP MEMBERSHIP\n128\nAPPENDIX B. LIST OF WORKSHOP ATTENDEES\nV","FIGURES\nPage\n11\n1. Lake Ontario drainage basin.\n17\nLake Ontario water temperature (°C) at -15 m, July 1972.\n2.\n3. Lake Ontario current (full barb is 1 cm sec -1) at -15 m\n18\nand dynamic height (mm), , July 1972.\n20\n4. Lake Ontario simulated temperature field, July 1972.\n20\n5. Lake Ontario vertically averaged flow, July 1972.\nEastward component of Lake Ontario current, July 1972.\n20\n6.\n7. Isotherm contours (°C) at Oshawa, Ontario, flag\n28\nstations, October 10, 1972.\n8. Constant speed contours (cm sec -1, at Oshawa, Ontario,\n29\nflag stations, October 10, 1972.\n46\n9. Definition of an aquatic ecosystem.\n46\n10. Important ecological processes for modeling.\n11. An idealized hydraulic element versus a laboratory stirred\n48\ntank reactor.\n51\n12. Physical representation of a stream.\n52\n13. Geometrical representation of a reservoir.\n53\n14. Segmentation for river run reservoir.\n15. Hypothetical horizontal sectioning of Lake Michigan\n54\ninto uniform zones.\n55\n16. Vertical sectioning of a lake into uniform cells.\n56\n17.\nSegmentation for Lake Erie.\n59\n18. Segmentation for Lake Ontario.\n19. Phytoplankton chlorophyll as a function of day of year\n60\nfor selected segments of Lake Ontario (0-4 m) .\n20. Preliminary comparison of Lake Ontario output to observed\n61\ndata at two segments, June 1972.\n67\n21. Model of wind-driven motions in a small stratified lake.\n68\n22.\nMomentum transfer of heat in Lake Ontario.\n77\n23.\nResearch feedback loop.\n81\n24. Temperature distribution in the surface boundary layer.\n81\n25. Surface pressure distribution.\n82\n26. Surface wind distribution.\n27. Vertical velocity (cm sec -1, distribution at a height\n82\nof 1 km.\n28. Schematic flow diagram of the boundary layer prediction\n84\nproblem.\n86\n29. Relationships between evaporation and large-scale parameters.\n30. Relationships between sensible heat flux and large-scale\n87\nparameters.\n31. Relationship between the Richardson number and the bulk\n88\nRichardson number.\nvi","TABLES\nPage\n1.\nProgress in accomplishment of IFYGL objectives\n4\n2.\nSelected future lines of research including modeling\n7\n10\n3.\nIFYGL schedule\n10\n4.\nIFYGL projects and objectives\n5.\nMajor planned scientific-technical - products\n12\n6.\nIFYGL accomplishments and deficiencies to date\n13\n14\n7.\nQuestions and objectives for future research\nFive platitudes of environmental (lake system) modeling\n65\n8.\n114\nSummary of priorities assigned to various problems\n9.\nvii","INTRODUCTION - E. J. Aubert\nWelcome to this workshop at the Great Lakes Environmental Research Labora-\ntory (GLERL). We are the newest research component of the National Oceanic and\nAtmospheric Administration (NOAA) Environmental Research Laboratories. We plan\nto have a workshop proceedings; that is why a tape recorder is being used. We\nfelt it might not be reasonable to ask all the speakers to come with manuscripts,\nso our plan is to record and transcribe the workshop. Proceedings will then\nbe published for limited distribution. The proceedings will be primarily for\nthe attendees and are therefore viewed as a working document. I would like to\nexpress appreciation to several people in GLERL who have participated in\norganizing this workshop up to this point. I would like to acknowledge\nArt Pinsak, who has been my deputy on this; Bob Bramlet, my Administrative\nOfficer, who has solved many logistics problems; Marlene Hein and Jean Grasso,\nwho helped register you; and Dave Norton and Steve Bermick, who are assisting\nwith projection and recording. Many others are also involved.\nI have identified five objectives for this workshop. The first objective\nis to determine future Great Lakes environmental research initiatives. Many of\nus have been involved in the International Field Year for the Great Lakes\n(IFYGL), and although IFYGL is not yet completed, it is desirable at this point\nto take stock of what we have accomplished both in IFYGL and in other Great\nLakes research investigations that we expect to complete in the near term.\nThen we can look at where to go from here. I frequently use the word simulation\nin the program outline. The word is used in a broad sense to represent the end\nproduct of a sequence of research endeavors involving both field and laboratory\nobservations, analysis to organize the information and to better understand the\nprocess and phenomena, and simulation to organize this information in a predic-\ntive model. A predictive or simulation model may be theoretical or numerical.\nA feedback mechanism exists in this research sequence and requires many feed-\nback loops. Modeling is part of the learning process, and as I view it, the\nsimulation or prediction precision represents the state-of-the-art for the\nenvironmental science involved. If we simulate or predict poorly, we do not\nunderstand very well.\nThe second objective is to provide the university research community an\nopportunity to discuss and recommend future Great Lakes environmental research\ninitiatives. We do plan, as a result of this workshop, to prepare an initiative\nfor the budget process through this Laboratory.\nThe third objective is to consider possible United States-Canadian joint\nresearch initiatives. Recognizing that the international boundary runs down\nthe middle of four of the Great Lakes, one must, if one plans to undertake a\nlake-scale investigation, recognize that advantages may exist for a joint\nUnited States-Canadian research program. All initiatives, however, need not\nbe lake-scale. Likewise, one major lake is in the United States.","The fourth objective is to identify logical follow-on research to IFYGL,\nand the fifth objective is to explore priority research needs of some of the\nmajor NOAA users, such as Sea Grant, Office of Coastal Zone Management, and the\nNational Weather Service. In other words, as I see the scope of this workshop\nand the scope of this Laboratory, some of our environmental research is to\nsupport NOAA users. It is postulated that a suitable joint research effort\nbetween GLERL and these users would improve the total NOAA environmental pro-\nduct. We visualize the program of GLERL as a joint in-house and grant or\ncontract venture, rather than solely an in-house effort.\nThe attendees at this meeting fall into two groups: people from NOAA and\nGreat Lakes researchers from universities and private institutions. GLERL has\na large representation at this workshop. The following NOAA units are also\nrepresented: Environmental Monitoring and Prediction, Marine Resources, Sea\nGrant, Coastal Zone Management, the National Weather Service, and the Environ-\nmental Research Laboratories.\nThe workshop will start off in plenary session, which will continue\nthrough this afternoon. Work group sessions will convene this evening and\nwill continue in the morning. We will then reconvene in plenary session\ntomorrow afternoon to hear the major points and recommendations of each work-\nshop group. Workshop sessions will not be recorded on tape. The summary of\nthe recommendations presented in plenary session and the discussions which\nfollow will be recorded.\nThe first plenary session has been structured to give perspective to the\nworkshop. If you look closely at the program, you will see that all the topics\nin the plenary session today line up with the work group sessions. Item 1.1,\nthe View of IFYGL Research, and item 1.7, Simulation of Environmental Dynamics\nof the Great Lakes, both are background for the Environmental Dynamics work\ngroup session. Two topics also back up the Water Movements work group session:\nitems 1.2 and 1.4, Simulation of Lake Scale and Nearshore Circulation,\nrespectively. Likewise, two topics provide background for the Aquatic Ecology\nand Water Quality work group; they are designated under those titles.\nSimulation is being used in the broad sense I described earlier. One plenary\nsession topic lines up with each of the work group sessions on Lake-Atmosphere\nInteractions and on Water Levels and Flows. I view simulation modeling as\nthe research process which organizes all the knowledge gained from observation\nand analysis into a predictive framework. Simulation is the end product,\nalthough the ability to simulate requires the complete research process.\n2","1.\nPERSPECTIVES OF GREAT LAKES RESEARCH\n1.1 View of IFYGL Research - C. H. Mortimer\nWhat have we learned and where do we go from here? It is really too\nearly to say what we have learned from IFYGL. We are just beginning to look\nat some of the results, but planning must go on and budgetary planning must\nbe done several years ahead of time; this meeting is therefore timely. But,\nin order to cut the speeches and get to the debate, I have prepared a handout\n(table 1) which is by no means comprehensive. There are others here who can,\nof course, go back into the history of IFYGL when it was a gleam in the eyes\nof the founding fathers of the steering committee. Dr. Chandler is in a much\nbetter position than I to tell you about that. I joined at a much later stage,\nwith the Water Movements working group.\nThe IFYGL program was in many ways unique in that it was the first large-\nscale attempt to study the physical limnology of the Great Lakes. The major\ninstitutions on both sides of the border took part; six research vessels, a\nnumber of smaller craft, and over 600 scientists and technicians from both\nsides of the border were involved. So, on the Great Lakes scale at least, it\nwas \"big science. It started as a component of the International Hydrological\nDecade program and therefore hydrology had an important role. Meteorologists\ncame into the program early and played a great part in it. I am not competent\nto speak of the results in that field, but I am sure others will do so.\nThen, at a fairly late stage, resulting from proposals some of us\nmade at a meeting of the International Association for Great Lakes\nResearch, biology and chemistry were added. Of course, IFYGL, like many such\nlarge scientific projects, had the task of selling the programs to governments.\nThe presence of a water quality component helped, but it was also significant\nthat this was an important attempt to weld the program in physical limnology\nto biology and chemistry and to bring the principal users into the picture at\nthe beginning of planning.\nTable 1 lists a number of themes. It is not complete, and I am sure\nothers will be added.\nHow are we going to measure the progress of IFYGL? I believe progress\nwill be measurable mainly in three main categories listed as columns I, II, and\nIII. As this workshop continues, I hope you will be able to fill in some of\nthese columns--they are left blank at the moment. It is a game you can play as\nthe workshop proceeds. Column I lists progress in estimating known effects or\nbetter estimates of things we have known already but need to know with a greater\nprecision. Column II relates to new discoveries or improved understanding of\noperating mechanisms, and column III, already referred to by Gene Aubert,\nrelates to predictive modeling capability which environmental management needs\nand is willing to pay for.\n3","Table 1. Progress in Accomplishment of IFYGL Objectives\nPROGRESS IN:\nIFYGL THEMES\nI. Estimat- II. Discovery III. Pre-\ndictive\ning known\nor improved\nmodeling\neffects\nunderstanding\nof mechanisms\ncapability\nWater Quality:\n(a)\nin basin; inflow, out-\nflow; evaporation-\nmethods compared\n(b)\nin air mass\nEnergy fluxes in and between air\nand water. Lake heat budget\nAir motions: air/water\ninteractions, **M\nWater motions:\n(a)\nsurface waves - short\n- long,\n**M\n(b)\ninternal waves - short\n- long,\n**M\n(c)\ncurrents--whole-basin\ncirculation patterns,\n**M; nearshore patterns\ndiffusion and dispersal\n(d)\nSediment distribution and inputs\nWater quality inputs and exchanges\nof dissolve materials - nutrients,\n**M\n-- toxins\nBiological studies:\n(a) surveys\n(b) dynamics, **M\n* on a variety of space and time scales.\n**M, substantial modeling effort attempted.\n4","The first theme is Water Quantity, i.e., basin inflow, outflow, and evapo-\nration estimated by various methods. What was attempted was a large-scale\nLake Hefner experiment. In the classic experiment on Lake Hefner, the water\nbudget and energy budget methods were used to provide estimates of evaporation,\nthat important but usually ill-defined term in the water balance equation.\nIt will be interesting to see how much further progress has been made as a\nresult of the IFYGL work. My guess is that we can put a small plus in column\nI there. Others will be talking about the attempt to estimate water quantity\nin the air mass.\nA lot of effort was put into another theme, energy fluxes in and between\nair and water. Most of the ship time was taken up in measuring thermal\nstructure of the lake and its changes with time. It will be interesting to\nsee how much closer those estimates are and how much further we have proceeded\n1\nbeyond\nSweers'\n(1969)\nsummary of knowledge of Lake Ontario heat budgets,\npublished before IFYGL started. My own guess here is that we shall be able\nto put a small plus under column I, a query in column II, and a small plus\nunder column III because an improved estimate, of course, gives improved\npredictive modeling capability. My strategy in making such sweeping and\ncertainly debatable statements is to generate discussion. If I may insert a\nconclusion from later remarks on \"Where Do We Go From Here,\" I believe that\nfuture investigation should concentrate on the dynamics of key mechanisms,\nrather than repeating the use of research vessels for large-scale surveying\nof quantities that we know already to a fair degree of accuracy.\nI will touch briefly on the remaining headings in table 1. Where substan-\ntial modeling efforts were attempted, I have inserted **M.\nWater motions fall into various classes depending on the space and time\nscales that were considered; there were programs on short surface waves, long\nsurface waves, and seiches and associated modeling efforts which were quite\nsuccessful. There have been a number of notable advances, for example, of\nD. B. Rao's normal mode analyses and Paul Hamblin's treatment. A good set of\nwater level measurements is available for verification, and new results con-\ncerning both the gravitational and rotational modes of Lake Ontario have been\nobtained. A large set of observations of internal waves, both short and long,\nwas made using several instruments, particularly Farrell Boyce's thermistor\nchain and the undulating transducers that we towed from the research vessels.\nAnalysis has only begun, but some of the patterns are beginning to emerge. I\nbelieve these will be focused more clearly when we recognize the episodic\nnature of the forcing functions. We can already put a small plus in column II\nas a result of the discovery of internal surges on the upwelling fronts.\n1\nSweers, H. E. (1969), , Structure, dynamics and chemistry of Lake Ontario,\nMarine Sciences Branch, Department of Energy, Mines and Resources,\nOttawa, Canada, Manuscript Report Series No. 10, 227 pp.\n5","Whole-basin circulation patterns have been successfully modeled (Bennett,\nSimons), warranting large pluses in columns II and III. Another strong feature\nof the IFYGL program was the attention paid to coastal circulation and coastal\ncurrents through the setting up of coastal chains both on the Canadian side\nand on the United States side under the respective leadership of Profs.\nCsanady and Scott. They will present some of the initial results, so little\nneeds to be said here. I predict that we shall be able to log some large\npluses in columns II and III.\nThe work done on diffusion and dispersal by Murthy, Kollenberg, Csanady,\nand colleagues has provided much improved estimates of the horizontal and\nvertical dispersion and diffusion coefficients. Therefore I believe that all\nthree columns will register pluses.\nDeep lake sediments were not studied in detail under the IFYGL program,\nalthough there was some work, particularly at Canada Centre for Inland Waters\n(CCIW), in sediment distribution and inputs. Water quality and biological\nstudies were added to the program at a later stage, and these will be reviewed\nlater in this workshop. For these I believe we can insert pluses in column I\nnow, perhaps a plus in column II later, and in due course a plus in column III.\n\"Where to From Here\" is the main theme of this workshop and I have made\na few suggestions of my own on table 2. You may wish to scan first the\nmaterial at the top and then go to the \"Preparatory Work in Advance of\nNew Field Programs.\nIf I may expand a little on some of the points summarized above, I make\na strong plea for thorough analysis of the IFYGL findings to exploit fully the\ninvestment. We all know of examples where this was not done because funds\ndried up after the field work was completed. For example, consider a\n$13 million program on the Great Lakes just over 10 years ago. Few results\nhave been published; others are still in limbo; great efforts have largely\nbeen dissipated. That, I sincerely hope, will not happen to IFYGL.\nTherefore, we should strongly press for thorough digestion and exploitation\nof the present findings, if only because we can take steps forward on their\nfoundation.\nMy second plea is for \"prior modeling,\" i.e., modeling before the experi-\nment is designed. We had this in mind during IFYGL planning, but we did not\nhave time or funds to do it.\nMy third plea is for prior instrument development and reliability testing.\nIFYGL has had its successes and its failures. Perhaps we do not want to dwell\non the failures, except to learn from them. On the U.S. side, the planning of\nwater movements instrumentation left a great deal to be desired. Instrumenta-\ntion requirements were considered by the water movements panel, but the design\nplans were not. The scientists should have made a recommendation as to whether\nthe adopted real-time telemetry was worth the additional cost. In the end,\n6","Table 2. Selected Future Lines of Research Including Modeling\n1.\nVertical motions and structure--air/water interactions; vertical fluxes\nof momentum and heat (buoyancy); two- and three-dimensional modeling of\nthermocline history.\n2.\nHorizontal motions and dispersal--scales and mechanisms; horizontal\nshear effects (see 3, which follows). .\n3.\nInshore/offshore exchanges and partition of energy--mechanics of upwell-\ning and subsequent whole-basin responses; shore-trapped long waves;\ngeneration and decay of nearshore currents.\n4.\nAssembly and critical review of all available chemical and biological data\nfor the purpose of model testing, model development, and design of\neffective long-term monitoring strategies.\nPreparatory Work in Advance of New Field Programs\n1.\nThorough analysis of present IFYGL findings to exploit fully previous\ninvestment.\n2.\nPrior modeling to focus on key questions and to improve design of\nexperimental programs along the selected research lines.\n3.\nPrior instrument development and extensive reliability testing, under\nrigorous field conditions, designed to provide answers to key questions\nidentified under 2, above.\n4.\nEncourage interagency and interinstitutional planning to optimize use of\nresearch platforms and funds.\nlittle use was made of this feature. Also, as is so often the case, what\nappear to be small details of design and seamanship can largely determine\nsuccess or failure. For example, during IFYGL, it was not possible to change\nthe gas cylinders on the Texas Instrument buoys in rough weather because\nthe buoy casing was awash. Breakdowns, coupled with a 30-hour limit in the\nbackup tape, led to considerable loss of data during particularly\ninteresting episodes. Standard, well-tried, self-contained instruments of\nconventional design used by CCIW were more successful.\n7","My final plea is for interagency and interinstitutional planning of the\nkind we are starting today. We must recognize that a number of agencies,\non the U.S. side at least, are developing plans for Great Lakes research on\nrather a large scale. Although they have different aims and missions, each\nagency is looking to a similar type of lake research to answer particular\nquestions. Therefore, in order to obtain maximum benefit from expenditure of\nthe federal dollar, coordination is called for.\n8","1.1.1 Response - D. C. Chandler\nIn an overview of the IFYGL program, it seems to me there are two\ncategories of benefits derived from that experience: (1) direct and concrete\ngains in the form of scientific data and technological advancements and,\n(2) indirect and intangible benefits in the form of attitudes, viewpoint, and\ngeneral philosophy about the Great Lakes.\nThe first category has been summarized by Dr. Mortimer and the remainder\nof the day will be given over to a discussion of specific and detailed\nscientific gains. Therefore, I will confine my brief remarks to the second\ncategory--indirect and intangible benefits.\nI feel that the IFYGL program influenced the attitudes and viewpoints\nof Great Lakes investigators in many ways, but I will comment on only four\naspects for the sake of brevity:\n(1) It was the first successful attempt at a multidisciplinary study of a\nGreat Lake with special emphasis on the total system (biological, chemical,\nand physical processes and phenomena of the lake water and the inter-\nactions of the water with its atmospheric and geologic boundaries). The\ncomponents or elements of the program were not necessarily original or\nimaginative, but rather they consisted of current procedures, methods,\nand technology. However, it demonstrated without question the advantages\nof this approach over the results of individual or small group effort.\n(2) It demonstrated that the Great Lakes are mesoscale aquatic systems,\nrequiring the application of oceanographic methods, equipment, design of\nfield study, management procedures, and level of funding in the conduct\nof research. It further demonstrated the feasibility and desirability of\nmultiple ship synoptic coverage of the lakes.\n(3) It also emphasized the importance of an international cooperative effort\ninvolving government agencies, industry, and academic institutions. No\nsingle organization possess the total capabilities to study adequately\nthe Great Lakes as a total system.\n(4) It created an opportunity for many interested scientists to become\ninvolved in a way totally closed to individuals or small groups.\nAs one who has been involved in Great Lakes research for more than three\ndecades, I am greatly impressed by the present-day wide acceptance, among Great\nLakes researchers, of the multidisciplinary, cooperative approach to Great Lakes\ninvestigations. Prior to the field year, the predominant attitude among the\nacademic scientific community was one of emphasis on individual effort with com-\nplete freedom to pursue a specific problem which required low levels of funding.\nThe field year program afforded an opportunity for the first time for academic\nscientists to participate in a cooperative effort involving their specific in-\nterests and with higher levels of funding. I believe these scientists are in a\nfavorable mood to continue this kind of involvement, and I sincerely hope that\nin the near future another multidisciplinary cooperative Great Lakes program will\nbe launched. It is to be hoped that such a program would build on the experience\nof the field year by avoiding the mistakes of that effort and strengthening the\nareas of success.\n9","1.1.2 Response - E. Aubert\nI will give a brief overview of IFYGL. The point has been made that a\nlot of analysis must still be accomplished. The IFYGL schedule includes plans\nto continue the IFYGL analysis phase until 1977. It is desirable to give\nperspective to the research accomplishments and deficiencies, although I\nrecognize that my perspective is incomplete. IFYGL is so broad that I doubt\nanyone, Professor Mortimer excepted, can adequately define all the IFYGL major\naccomplishments and deficiencies. I suggest three questions or objectives\nfor future research. These objectives are limited. Participants will have\nfuture research ideas to suggest for consideration and discussion in this\nworkshop.\nIFYGL addressed Lake Ontario and the Ontario Basin (fig. 1). At the time\nof this workshop, near the end of 1974, we have completed the first four\nscheduled activities (table 3) and the data management-archive generation is\nnearing completion. A large data base is being generated in both the United\nStates and Canada. Several years remain in the analysis phase. We anticipate\nmany more results from the research analysis phase than what we have\naccomplished to this point.\nTable 3. IFYGL Schedule\n1976\n1977\n1974\n1975\n1971\n1972\n1973\nACTIVITY\nDevelop technical plan\nPrepare for field program\nField year operations\nEngineering, tests, and data\nsystem comparisons\nData management-archiver\nAnalysis\nTable 4 is an overview of the IFYGL scientific objectives and projects.\nEight different major projects address these three objectives.\nTable 4. IFYGL Projects and Objectives\nPROJECTS\nSCIENTIFIC OBJECTIVES\nAtmospheric water balance\nTo determine large-\nLake heat balance\nscale processes\nTerrestrial water balance\nEvaporation synthesis\nMaterials balance\nAtmospheric boundary layer\nTo determine small-scale dis-\nWater chemistry and biology\ntribution, variability, processes\nAtmospheric boundary layer\nTo model limnological, hydro-\nTerrestrial water balance\nlogical, and meteorological\nWater chemistry and biology\nproperties\nWater movement\n10","YORK\nFigure 1. Lake Ontario drainage basin.\nONTARIO\n77\"\nNEW\nPennsylvania\nNew York\n78°\n19\"\n100\n75\nKILOMETERS\n50\n80\n25\no\n42","Table 5 shows the major planned scientific-technical products.\nTable 5. Major Planned Scientific-Technical Products\nAnalysis of budgets--lake, basin, atmosphere\nWater, heat, materials\nAnalysis of natural distribution and variability--what\nand why\nPhysical quantities, chemical concentrations,\nbiological properties\nDevelop and test models for analysis, diagnosis,\nprediction and simulation of interdependent physical,\nchemical, and biological properties\nReports\nIFYGL scientific reports, articles and agency scientific\nreports, technical reports on data acquisition systems\nOnly a few of these products have been achieved at this time; all are antici-\npated by 1977. Eight major IFYGL international summary scientific reports are\nplanned for completion during 1975, 1976, and 1977.\nTable 6 lists the major IFYGL accomplishments and deficiencies as I see\nthem at this time. The data collection phase is completed; we had some successes\nand some failures. A large data archive will result. The natural distribution\nand variability (NDV) analyses include budgets and small-scale distributions for\nthe various projects listed. Little variability analysis will result from the\nchemical and biological program since no suitable data were collected for this\npurpose. Likewise, little variability analysis will result for mean and eddy\ntransports of chemical constituents. With respect to model research, several\nsignificant efforts are underway and significant success has been achieved.\nI am not aware of any predictive modeling research that is underway in the\nnearshore at this time, but research plans may be initiated within the next\nyear or so. There is little test and evaluation of these models due to the\nfact that they are relatively new. Joe Simon's model development of the physical\ncirculation was perhaps one of the first, and his model is at the most 2 years\nold. His first model has had several versions.\nWith regard to publications, we prepared a proceedings of the IFYGL\nsymposium held at the American Geophysical Union (AGU) meeting in April 1974.\nIn April 1973, there were about 20 papers presented at the Great Lakes\nConference and published in the International Association for Great Lakes\nResearch (IAGLR) Proceedings. In August 1974, there were 54 IFYGL papers\npresented at the Great Lakes Conference, and I would expect that next year\nthere may be even more.\n12","Table 6. IFYGL Accomplishments and Deficiencies to Date\nData collections--complete\nSome successes, some failures\nLarge data archive nearly complete\nAnalysis--natural distribution and variability\nBudgets\n- Terrestrial water budget, atmospheric water budget, lake\nbudget, mass balance\nPreliminary analyses complete\nSmall-Scale -\n- Water movements and boundary layer\nVariability analysis partly complete\nSome episode analysis\n- Chemical budget\nStatus of lake surveys nearly complete\nLittle variability analysis\n.\n- Transport\nLittle variability analysis\n.\nModels, simulation\nWater movements\n- Lake-scale circulation\nSeveral developed, limited testing and evaluation\n- Nearshcre circulation\nNone developed\nBoundary layer\n- Mesoscale phenomena and processes\nSeveral developed, limited testing and evaluation\nChemical budget\n- Water quality\nSeveral under development, no testing and evaluation\n.\n- Ecology\nUnder development, no testing and evaluation\nPublications\n- Proceedings of American Geophysical Union Symposium\n- Papers for International Association for Great Lakes\nResearch Conference (54)\n13","Some ideas for questions or objectives for future research are contained\nin table 7. There ought to be better evaluation of lake-scale circulation\nmodels. A second item refers to the nearshore--there have been interesting\nstudies in IFYGL and interesting results of nearshore phenomena. We do not\nfully understand, at this point, the mechanisms of the nearshore jet and\nthe transport and exchange processes. Hypotheses have been postulated, but\nthey cannot be adequately tested with the data collection in IFYCL. The\nchemistry and biology research was a late entry in IFYCL. A lot of research\nis underway, and significant accomplishments are anticipated. The chemistry\nand biology data collections are, however, of insufficient intensity to\nsupport variability analyses and detailed ecological model development.\nThe\nthird item on table 7 recommends more intensive chemical and biological\nexperimental and model research focusing on typical nearshore regions.\nTable 7. Questions and Objectives for Future Research\nTo evaluate lake-scale circulation models,\nDetermine uncertainity, refine models\nApply models to management questions\nTo determine variability of nearshore circulation\nand materials transport,\nImprove understanding of processes\nDevelop and test simulation model\nApply models to management questions\nTo determine variability of C and B properties in a typical\nnearshore region,\nDevelop and test models (1-D, 2-D, 3-D) to simulate\nobserved variability\nApply models to management questions\n14","1.2\nWATER MOVEMENTS: SIMULATION OF LAKE-SCALE CIRCULATION - J. Bennett\nProf. Mortimer's outline had three boxes: estimation of known effects,\nprogress in discovery or improved understanding of mechanisms, and predictive\nmodeling capability. While Prof. Mortimer is inclined to give pluses in most\nof these boxes, I am more skeptical. I think we have just barely scratched\nthe surface in modeling. Most progress so far has been in understanding the\nmodels, which is a far cry from understanding the lakes. I think we have\nelucidated quite a few of the physical mechanisms that should theoretically\nhappen in the lake. We had a fairly foggy view of some of them before, for\nexample, propagation of low frequency waves, and I think they have been shown\ntheoretically and observationally to be important. As far as quantitative\nprediction of these processes, I think we are still a long way off. The\nmechanisms in the model are probably just being understood now, and it took\na long time to do that.\nTo give just a brief review of the modeling that did take place in the\nIFYGL program, there were seven numerical models of Lake Ontario. That says\nsomething already. Some people consider it undesirable to have this many;\nothers, myself included, consider this healthy competition in most cases.\nThe models have been my own, Joe Simon's, Nobuyoshi Baba's, a student at\nPrinceton (the model was for his thesis) The thesis was very interesting,\nand I think it will turn out to be one of the cheapest contributions to the\nIFYGL program since it was done without any support from the program. It is\nthe only model I have seen run for the whole navigation season. It is a 17-\nlevel three-dimensional model which runs for 8 months at a time using typical\nwinds. Another model is that developed by Pandolfo and Jacobs, an air-sea\ninteraction model. Bonham-Carter and Thomas of the University of Rochester\nhave developed a model of the Rochester Bay area. This approach has a lot\nof potential and I think it may be what we will be looking for in the future.\nMost of the applications on the Lakes involve smaller scale shore-based\noperations, and the University of Rochester model is the first step in matching\na small-scale shore-based model to a large-scale lake model. Another model\ndeveloped in the last couple of years was David Paskausky's The other numerical\nmodel, to make it complete, was D. B. Rao's two-layer model. He is using it\nto understand internal waves and seiches in lakes.\nThere are this number of models because a lot of models of natural bodies\nof water can be applied to Lake Ontario.\nThere is a lot of activity in numerical modeling, but I do not think most\nof the progress has been made in that area. Many analytical studies have\nhad much more effect on the design of the IFYGL experiments. The simple two-\nlayer model of Csanady, for instance, had more effect on the design of the\nprogram than any of the numerical models. One of the weaknesses of the IFYGL\nprogram was that there was no thought given to using models to design the\n15","program. Compare that to the Mid-Ocean Dynamics Experiment. A workshop was\nheld a year or two before they designed the program; a group of people got\ntogether at the National Center for Atmospheric Research and played with\nnumerical and analytical models to design the right observing network for the\nprogram and to try to guess what they would measure. In any physical problem,\nit is always a good idea to try to guess what the results are going to be, even\nif it does not turn out that way.\nI have a couple of fairly concrete suggestions for designing field programs.\nFirst, I do not think the Great Lakes research community should devote much\neffort to developing numerical techniques or put too much emphasis on the\ntechnical details involved in modeling. I am not saying that the Great Lakes\ncommunity cannot afford to fund projects that are simply concerned with the\ntechnical details of modeling. There is a huge literature in numerical model-\ning and mathematical modeling in other fields which can be used and we just\ncannot add much. There are a lot of well-known techniques in numerical weather\nprediction (such as matching small-scale models to large-scale models) that\nare not being applied very well to the Great Lakes.\nTechnical details will not be dwelt on because I assume anybody who can\ndo modeling can read those himself. We can take for granted that most people\nin this room could generate a numerical model, and I do not consider that\na big feat anymore. The main difficulty with numerical modeling is lack of\ninsight into the Great Lakes in order to apply them, and this turns out to be\na very difficult problem.\nTo give an example where I think numerical models have some definite\nweaknesses, I am using Bob Pickett's slide of July temperatures during the IFYGL\nprogram (fig. 2) The basic features of this temperature distribution have\nbeen understood for a long time. We do not need IFYGL to tell us, for instance,\nthat in July there is a residual pool of cold water at the bottom of the lake,\nor that the south shore is warmer due to downwelling and perhaps the inflow of\nthe Niagara River is quite warm this time of year, or that there is upwelling\non the north side.\nFigure 3 shows the resultant current field in July at 15 m depth,\nand the dynamic height patterns can be calculated from the temperature field.\nWhat is interesting about them is that they seem to be internally consistent.\nThere seems to be a big cyclonic circulation of the lake. The dynamic height\nmethod seems to work in estimating the currents. If one did not try to predict\nthese currents with a model, the explanation of a big geostrophic gyre would be\nquite satisfactory. Unfortunately, none of the models give this. I think\nthere is something fundamental going on here that we really do not understand.\nIf we have an understanding of the mechanics of the models, then they can help\nus understand basic phenomena like this. It is easy in any model to get east-\nward flow on the south shore, but the trouble is getting the flow to turn\naround and go back west on the north shore. As you know, the prevailing wind\nis from the west and tends to drive the flow in shallow water toward the east.\n16","44.0\n43.5\n76.0\nD\n08\nII\n10\nO\no\nFigure 2. Lake Ontario water temperature (C) at -15m, July 1972.\n77.0\n77.0\n9\n12\n10\n11\nO\nO\nO\n78.0\n78.0\nO\n7\n8\nO\n8 10-12\no\n8\n79.0\n79.0\n7\n07\n80.0\n44.0\n43.5","44.0\n43.5\n76.0\nLake Ontario current (full barb is 1 cm sec-1, at -15 m and dynamic height (mm), July 1972.\n-4\n2,0-2\n77.0\n77.0\n78.0\n78.0\no\no\n-2\n-4\n-6\n0\n2\n79.0\n79.0\n9\n2\nFigure 3.\n80.0\n43.5\n44.0","The question is, why does the wind not have a bigger effect on the current?\nIf you did not know better, you would say the temperature pattern, generated\nby heating and wind, produces currents that are essentially in geostrophic\nequilibrium with no other wind effect at all.\nFigure 4 shows a temperature field from my model for July. It is supposed\nto compare with the first slide. It is not too bad a fit. There may be some\nweaknesses in it, but I see no big problem; warm water is present on the south\nshore, the big cold pool is in the center, and upwelling occurs along the north-\nwest shore. The basic features are roughly correct. Almost any model can be\ntuned to give a pattern that looks more or less like this.\nFigure 5 is the vertically averaged flow. This is stream-function in\nunits of 10 8 cm 3 sec -1 Instead of having the one big gyre that the observed\ncurrent has, it has a relatively large cyclonic gyre and also a smaller\nanticyclonic gyre in the northwest.\nFigure 6 is a graph of the eastward component of the current. The most\nglaring error is near the north shore. It shows all the shore water flowing\ntoward the east, and those measurements by Bob Pickett all show the current\nflowing toward the west. It is not a matter of the observations either because\nother people have measured current even closer to the north shore at other\ntimes of the year, and they also say that the current on the north shore of\nLake Ontario is to the west. This is something we fundamentally do not under-\nstand. However, I still have hope that all these problems can be ironed\nout with the IFYGL data. I hope that the modeling expertise we have built\nup in the last couple of years will eventually explain most of the IFYGL\ncurrent measurements. This knowledge can be used to design a new field\nproblem that will be better.\nAnother suggestion is that serious thought should be given to running at\nleast one model operationally during any field experiments. This is a sugges-\ntion from Joe Simon. His argument is that numerical weather prediction improved\nwhen meteorologists had to make a forecast every day. They found out they had\nblunders which had to be corrected. It is not just a matter of saying \"We\npredict a Kelvin wave over in that cove; see if you can find it.' 11 Operational\nmodeling would provide discipline for the modelers and a means for continual\ntesting of the model. After the field work, modeling can still play an integral\nrole in interpretation of the results.\nI suggest that all the modeling take place at institutions where data\nanalysis is going on; it is helpful to have modeling work hand-in-hand\nwith analysis of the observations.\nI would also suggest that the modeling program should retain versatility.\nIt should not rely on just one numerical model, but it should keep many people\ninvolved in modeling and the analytical and numerical models should be used\nin conjunction with analysis of observations.\n19","6 7\n8\n8\n7\n9\nFigure 4. Lake Ontario simulated temperature field, July 19.\n-100\n-200\n100\nFigure 5.\nLake Ontario vertically averaged flow, July 1972.\n5\nFigure 6. Eastward component of Lake Ontario current, July 1972.\n20","1.2.1 Response - D. B. Rao\nJohn Bennett has preempted me on practically everything I wanted to say.\nI basically have to agree with everything he said.\nIn terms of modeling, the Great Lakes community could conveniently use\nmathematical techniques derived from numerical meteorologists over the last\nthree decades instead of expending efforts on developing new techniques.\nWhen we look at numerical models of the Lakes (John Bennett has described\nseven models or so), they consist basically of two categories: the barotropic\nhomogeneous numerical lake models and the baroclinic numerical models where\nbaroclinicity is either at fixed levels or between movable interfaces. The\nbarotropic models have been successfully used over short-time scales in hind-\ncasting studies of storm surges primarily on the Great Lakes by George Platzman.\nThey have been fairly successful in reproducing what is observed, and perhaps\none might say they are ready for operational use. On long time scales, baro-\ntropic models can be used to look at seasonal circulation patterns and steady-\nstate dynamics.\nThe two most elaborate baroclinic-multilevel models are those of Joe Simons\nof CCIW and John Bennett here. They have been integrated over long time\nperiods. Perhaps one can start using them operationally, although I guess you\ncannot put forth the information for public use like a weather forecast.\nThe models can be run with the idea of understanding. First of all, how\nclose does the model simulate phenomena that have been found in observations?\nOr maybe new phenomena can be discovered. These applications should be looked\nat in terms of analyzing the dynamics of the models, rather than comparing\nwith observations to see how faithfully currents and temperatures at a given\npoint and time can be reproduced. Even though the numerical model may not\nexactly reproduce some observed features, it still gives information for\nanalyzing processes in the lake. Examples are coastal upwellings, nearshore\nprocesses, or generation of internal waves. Also, things like the importance\nof nonlinear dynamical processes and interaction between coastal and interior\nzones of the lake can be understood.\nThe coastal zone is important, of course, from both the biological as\nwell as the waste dispersal point of view. Wastes are injected into the coastal\nareas, and those areas have to be modeled fairly well. Large-scale numerical\nmodels might give some information on how strong the interaction is between the\ncoastal zone and open lake.\nAs Dr. Aubert said, \"How does one go about developing limited-area numeri-\ncal models which probably do not exist at the moment?\" In meteorology there are\nlimited-area models which use large-scale model information. These techniques\nare available.\nFinally, I feel a government institution like GLERL or perhaps CCIW\ncannot only run models on an operational basis, but can see to it that, when\nthese models are put together, they are sufficiently general so that anybody\ncan make specific experiments by getting access to these models. This is\nwhat a general circulation model is supposed to be.\n21","1.2.2 Response - R. Pickett\nI want to comment on a point John Bennett made about using a numerical\nmodel in essentially an operational mode in a son-of-IFYGL experiment. Predic-\ntions would be made and checked during the field program. One of the problems\nin the past was that several years were required for data processing. If\none looks over the lake-scale studies that Prof. Mortimer mentioned, in each\ncase somewhere in the range of 4 to 6 years were required before the data were\nput in a useable form and before a data report was produced. Prof. Mortimer\nalso mentioned that data from studies of Huron and Superior were never published.\nSo, at least to my knowledge, these studies are still unedited recordings on\ntape somewhere.\nJohn Bennett also cited the IFYGL data. This is 1974 and those data were\ntaken in 1972. We are just now reaching the stage where we are integrating and\nediting both countries' data and analyzing the results. That seems to me to be\ntoo long. Certainly if future field operations are planned, more thought should\nbe put into how the data can be made available in a shorter time period.\nWhat are some of the things that can be done? Certainly, we can take only\nthe data we need. For example, we took 6- and 10-minute observations in IFYGL\nand calculated hourly averages. I think with present technology hourly averages\ncan be calculated at the transducers and just the results transmitted. We\nshould be able to get the computer in the process sooner. We could then do\nsome high-speed editing so that observations are immediately verified or thrown\nout. We could display the data over the whole lake in the manner of the\nillustrations by John Bennett. They were done by using a computer-coupled\ncathode ray tube. We could also, if we stage another field year, test our\nanalysis and editing procedures before the field work is begun to make sure\nthey are tuned up and working well. That way we would not go through a\nyear of development to come up with procedures to handle the data after the\nfield work.\nFinally, there seems to be a trend tc put most of the effort into field\nwork; then people and resources drift away so that few are left to crank\nthrough the long-term processing. I think we would do well in planning\nfuture field work to keep enough effort in data processing to squeeze it\ndown to the shortest time period. Until we do, there is no way of getting\nthe kind of feedback we need. As we bump into questionable data now in IFYGL,\nwe cannot find out what happened. Was the current increasing or was the\nsensor drifting at this particular level? What is the most probable\nexplanation? The field people have forgotten or gone.\nSomeone once said, \"Data is like fresh meat---it spoils very quickly.\"\nWe have to compress the period of years that it now takes to process data.\nOtherwise we will never be able to use numerical models in anything approaching\noperational modes.\n22","1.3 DISCUSSION\nAubert. We are running a little ahead of schedule. Even though discussion\nwas not planned in this plenary session, we could entertain discussion for about\n10 minutes. Are there comments on this plenary session up to this point? Do\nI hear any controversial topics?\nHolland. There were two contradictory recommendations made, I think. There\nwas the cost of this real-time business. It would be large, but telemetry is\nnecessary. Prof. Mortimer suggested that telemetry was unnecessary in IFYGL\nbecause there was no real-time use made of the data. This is, in essence, as I\nunderstand his comments.\nMortimer. In the water movement program, yes.\nHolland. If one were to go into real-time prediction, of course, telemetry\nwould be essential. Also the preparation of the system would be essential,\nthe shaking down of all software, all data processing, editing, and analysis.\nDespite automatic procedures, the staffing would be heavy. During the course\nof the thing, you would have to have a team handling the data, so the cost\nwould certainly peak during the operations phase. The cost of data processing\nand analysis would peak very heavily during the operations itself, rather than\nbeing distributed over time as they are in IFYGL. I think some of these things\nare essential in order to assure the success of the field program itself. Part\nof the problem with the IFYGL data is that we did not have test data and\nprocessing procedures in advance of the program and had to develop these after\nthe observations were taken. Then we face these unknowns. We find that we\ndo not have enough information to know exactly what we are doing. There is\nconsiderable merit in the suggestion, just in the interest of guaranteeing a\nsuccessful data collection effort. But I think it has to be understood that\nit is very costly and it introduces a lot of technological uncertainties.\nMaintenance is an example. A 30-hour or a 30-day backup recorder cannot be\nrelied on to serve this purpose. If the communications go out, you have had\nit, so more reliability is needed. Faster trouble shooting is needed and it\ngets to be a much more expensive project.\nBennett. When I suggested that a model be run operationally, I did not imply\nthat all the observations would be real-time. One can run a small-scale\nnumerical model real-time or he can hindcast. Every weekend you could run the\nprevious week with the observed winds, keep track of your prediction model\nthrough the field year, and get verification data anywhere you wanted to whenever\nyou had an opportunity to keep the model in tune. It is a long way between that\nand real-time collection and processing of data.\nHolland. J. Bennett is probably right. The real-time system may be quite ex-\npensive and not worthwhile, but there could be a lot of data collected within\na week or so. If the model were in one big lab like this, the data could be\nused to tune the model on an unofficial basis.\n23","Aubert. Does anybody else have comments on this point? J. Holland mentioned\nseveral points which in IFYGL added to this 2-year lag between observation time\nand data availability which R. Pickett referred to. That is, the data process-\ning procedures were not available, developed, and operational prior to the\nstart of the field year. Part of the reason for this was that the equipment\nwas not fully developed and tested before the operational period started. No\ndevelopmental data base existed on which to develop the data processing\nprocedures. The development and testing of data acquisition systems take a\nlot of lead time and this did not exist for some of the data systems. The\nsystems did not exist at the time they were needed for an orderly development.\nThis resulted in the lag, and the data processing procedures were developed\nafter the fact.\nJ. Bennett, you mentioned the desirability of more intensive testing. I\nguess you are adding the dimension of testing on a routine basis. By routine\noperations, you do not necessarily mean that a new integration would be started\nevery day or every 12 hours. Once a week might be enough.\nBennett. Yes, even every 2 weeks could be useful. You would not want too\nmany iterations. It would be an intellectual exercise more than anything. You\ncould send out the model results for comment to the field investigators every\nmonth or so and ask whether it agrees with what they saw in the lake or not.\nBirchfield. I think that would be particularly useful to view what goes on\nin a lake. Or it could be looked at from an episodic point of view because,\nif a large storm goes by, then some data will hopefully be coming in from some\naspects of the storm that will reflect its passage over the lake.\nHolland. J. Bennett was right. Even if a 2-week time scale is used for\nsimulation, data must be coming in currently, but this would be a less\nexpensive alternative. It certainly costs less to do a 2-week type real-time\noperation than it would to do a day-by-day real-time operation. But it would\nstill be a sterile exercise. You would run your models every week or two and\nyou would not know whether they had any correspondence to the real world or\nnot. This would not be very interesting. So data must be coming in. It\ncertainly would be possible to design a scaled down system to decide what\nparameters you want, what averaging time you want, how quickly you have to get\nthem, and how to size the thing to fit your pocketbook. It might be a month\ninstead of a week that you could afford to handle it in a real-time sense.\nThen you could go out in boats, pick up the tapes and process them, and\ncheck your simulations on that kind of time base.\nBaer. Could I ask a very simple-minded question? I find all the discussion on\nlake-scale circulation, assuming it will be a son-of-IFYGL as I heard it called,\nof equivalent scale, magnitude, and major activities. Are there not things\nthat need to be done that are not so big?\nAubert. Does somebody want to rise to that question?\n24","Mortimer. I think that will be answered by the end of today or tomorrow.\nAubert. Yes. It is not the general point of view that there should be a son-\nof-IFYGL. If you still have that question at the end of the plenary session,\nit should be brought up at a workshop. It is to be hoped that it will be an-\nswered by later presentations.\nMortimer. I want to make one point as a kind of footnote. I am not against\nreal-time telemetry, and I think some instruments might be designed that way.\nWhat I urge is that the instrumentation debate and decisions should sit right\nin the heart of the scientific program planning. They should not be dealt\nwith by some distant agency in Washington. Success in the end depends on\nreliable instruments and on seamanship. For example, in IFYGL there was little\nmoney to service the Texas Instrument buoys. I believe there were only two\nsmall motor boats based in Rochester, N.Y. They could not operate in rough\nweather; therefore they could not get the propane cylinders onto the buoys\nunless the weather was calm because of the buoy design. If the buoy broke\ndown, the weather was rough, and the 30-hour backup tape had run out, there\nwas nothing that could be done about it. of course, Murphy's Law being what\nit is, the most active episodes often occurred when the most interesting\nrecorders had broken down. There may be justification for real-time telemetry\non perhaps a limited number of instruments, but the information cost is much\nless with self-contained recorders of proven design. Now, having said that,\nI also say that the instrument contractors deserve considerable praise. Theirs\nwas a new entry to the oceanographic instrumentation field; when they saw the\ndifficulties, they pulled out all the stops to make things work.\nPinsak. I would like to make one point to clarify our perspective. Is the\nintent to test the model or to test the data? From the trend of discussion,\na presumption seems to be that the data are all good as they come in from the\nobserving system. This, of course, is not true. If the data do not fit\nthe model, there would then be a question as to whether to adjust the model or\nto adjust the data.\nAubert. I think we had better cut off the discussion at this point. The next\nitem on the agenda is Water Movements: Nearshore Circulation. The lead talk\nwill be given by Dr. G. Csanady of Woods Hole Oceanographic Institution,\nfollowed by responses from Prof. J. Scott of State University of New York,\nAlbany, Prof. G. E. Birchfield of Northwestern University, and Prof.\nT. Green of the University of Wisconsin, Madison.\n25","1.4 WATER MOVEMENTS: NEARSHORE CIRCULATION - G. Csanady\nIn the language of dynamical oceanography, the \"coastal boundary layer\"\n(CBL) may be defined as that band of water within which any Ekman drift\nperpendicular to shore reduces to zero. Given the presence of stratification,\nfrictionless fluid theory shows the width of the CBL to be of the order of the\ninternal radius of deformation, the magnitude of which is in most cases between\n5 and 50 km. In the absence of stratification, the width of the CBL may be\nexpected to be determined by frictional effects.\nThe only serious observational studies of the CBL appear to have been\ncarried out in Lake Ontario in connection with IFYGL. We should note here\nthat the CBL extends much further from shore than the littoral drift zone (LDZ)\nwhich has been studied extensively by civil engineers, geologists, and others.\nWithin the LDZ, the momentum of incoming surface waves is rectified by dissi-\npative processes and causes a longshore current. The width of this zone is\nfrom wave breaking depth to shore, or typically a few hundred meters. In what\nfollows, we shall be concerned with the bulk of the CBL which lies outside\nthe LDZ.\nThe IFYGL-related observations have firmly established a qualitative\ndifference between the current regimes of the CBL and those of the deeper,\nmid-lake region. Within the CBL, observed water movements are mostly shore-\nparallel and \"current-like\"; outside the CBL, they vary in direction in a\nperiodic manner, being more nearly \"wave-like.\" The IFYGL observations supplied\na detailed description of nearshore currents and leave no doubt about the\ndistinct identity of a CBL. In Lake Ontario, the width of the CBL is of the\norder of 10 km.\nA particularly important consequence of the distinct flow regime in the\nCBL is that pollutants discharged nearshore remain trapped within it for\nprolonged periods. This has been often noted in connection with effluent and\nriver plumes which generally turn shore-parallel after discharge and has been\ndocumented by specific dye diffusion experiments in Lake Huron.\nLinear theoretical models of wind-driven flow in a stratified fluid have\nyielded the concepts of coastal jet and Kelvin wave. Considering the simplicity\nof these models, they have been remarkably successful in providing an\nintellectural framework for the interpretation of CBL observations. I have\nrecently compiled a more detailed review of the achievements of linear\ndynamics.\nAn aspect of linear dynamics not completely resolved yet is the precise\ninfluence of depth variations on depth-integrated flow or \"transport.\" In\nshallow water, transport is downwind, while in deep water, return transport\noccurs--this much is generally agreed upon. However, the precise effects\nof friction and of the earth's rotation on a flow pattern left over from a\nwind impulse remain unclear. Although the CBL usually lies well within the\ndownwind leg of the topographically controlled transport gyres, a rotation of\nthis flow pattern, or its rapid spin-down by friction, is an important deter-\nminant of CBL behavior.\n26","Some especially interesting nearshore phenomena occur early in the heating\nseason during the so-called thermal bar period. IFYGL data provided much\ngreater detail on the nearshore flow structure than available earlier, and these\ncould be interpreted in terms of dynamical concepts with some conclusiveness.\nGiven the distinct identity of the CBL, it is a legitimate conceptual\nmodel to speak of mass transfer between one black box called \"CBL\" and another\nblack box called \"mid-lake.\" From a practical point of view, this particular\nexchange process is of evident importance. Existing evidence shows that under\nstratified conditions the flushing of the CBL is associated with onshore-off-\nshore movements of a thermal front which is the nearshore upwelled or downwelled\nend of the seasonal thermocline. The structure and behavior of this front is\nknown in a gross way and qualitatively, but is poorly understood. Figures\n7 and 8 show an example of nearshore upwelling and associated coastal jet\nobserved during IFYGL. Linear dynamics indicate large isotherm movements near-\nshore, but the theory is so far incapable of describing anything but \"small\"\ndisplacements, small that is, compared to equilibrium thermocline depth. A\nparticularly difficult feature of this problem is that the lake bottom slopes\naway from shore gently, but quite significantly in the sense that the water\ndepth can easily double or triple over the nearshore slanted portion of an\nupwelled or downwelled thermocline.\nIn connection with upwelled or downwelled fronts, one would like to be\nable to answer questions relating to their generation and decay, the factors\nwhich determine how far from shore an upwelled front (say) stabilizes, or how\nlong it takes for such an upwelled front to relax to its horizontal equili-\nbrium position. Also, one would like to know the magnitude of the mass\nexchange between the CBL and mid-lake associated with the development of an\nupwelling, its local disappearance upon the passage of a wave-like front, or\nits eventual frictional decay. These questions involve the effects and\nparameterization of turbulent friction in a strongly statified shear zone, time-\ndependent inertial adjustment to a state of equilibrium characterized by the\npresence of an inclined front, and finite displacements of fluid columns over\na sloping beach, problems all well outside the scope of existing linear theory.\nOne considerable contribution the Great Lakes community could make to\noceanography would be a thorough documentation and understanding of upwelled\nthermal fronts. In the Great Lakes, these fronts are not much more than 5 km\nfrom shore; and the logistics of their study is comparatively simple, certainly\nin comparison with oceanic fronts more than 160 km from shore near the\neastern seaboard of North America or their counterparts in the southern ocean\nthousands of kilometers from major oceanographic facilities. The understanding\nof the generation and maintenance of such fronts is a key outstanding problem\nin oceanography.\n27","20\n25\n30\n35\n10\n15\n5\n12\n9\n6\n10\n8\nFigure 7. Isotherm contours (°C) at Oshawa, Ontario, flag stations, October 10, 1972.\n5\nLAKE SURFACE\n12\n4\nII\n10\n9\nOSHAWA FLAG STATIONS\n8\nIFYGL - 1972\n7\n6\n5\n4\n4\n3\n3\nKILOMETER SCALE\n5\n2\nLAKE BOTTOM\n2\n6\n/\n/\nSTATION No.\no\n15\n20\n25\n30\n35\n5\n10","50\n60\n30\n40\n20\n10\nConstant speed contours (cm sec 1j at Oshawa, Ontario, flag stations, October 10, 1972.\nLAKE SURFACE\n25\n20\n25\n25\n20\n15\n15\n12\n20\n25\nII\n10\n9\nOSHAWA FLAG STATIONS\n8\nIFYGL - 1972\n7\n6\n45\n40\n35\n5\n4\n30\n60\n55\n50\n25\n15\n10\n20\n4\n5\n3\nKILOME TER SCALE\n3\n2\nLAKE BOTTOM\n2\n/\n10\nFigure 8.\n/\nO\nSTATION No.\n10\n20\n30\n40\n50\n60","It should also be pointed out that the current state of understanding of\nturbulent friction in the CBL is unsatisfactory. It is not clear, for example,\nhow far momentum advection by Ekman drift or horizontal momentum transfer by\nturbulent eddies are dynamically important in the CBL. One's best current\nguess is that both of these are similar in magnitude, and both are much less\nimportant than momentum transfer to the bottom by turbulent friction. Without\nfurther detailed studies, one cannot say whether this is so or under what\nconditions it ceases to be so.\nA probably related problem is the explanation of the observed asymmetry\nof \"right-hand\" and \"left-hand\" coastal jets (looking downwind) Observations\nduring IFYGL have shown right-hand jets produced by either westerly or easterly\nwind impulses to be stronger than left-hand ones. As a corollary, time-\naveraged flow or lake \"circulation\" have been observed to be cyclonic.\nRight now there are four proposed mechanisms on the market purporting to ex-\nplain this phenomenon, but none is completely convincing. All explanations\ninvoke some aspect of turbulent friction or of nonlinear momentum advection,\nalthough in quite different ways. This is a relatively happy situation for\nplanning further research: We know a phenomenon exists and we have some\ntentative ideas why, the task being to decide between rival theories.\nTheoretical studies of nearshore frictional effects have suggested the\nprobable existence of a kind of \"secondary flow\" in a vertical plane normal\nto the coast, onshore flow within the top layer being compensated for by off-\nshore flow below, or vice versa. Such secondary flow, superimposed on coastal\njets, could turn out to be of great practical importance in connection with\npollutant dispersal. However, we have not so far been able to relate such\ntheoretical models very well to observation, mainly because of the relative\ncrudeness of observations, or more specifically, the poor accuracy with which\nonshore-offshore components of nearshore currents can be determined. Thus, we\ndo not know the magnitude of the parameters that would realistically represent\nturbulent friction. In at least some of the extant theoretical work,\nspecific assumptions are made regarding the magnitude of the frictional\nparameters, mostly to the effect that \"horizontal\" eddy viscosities are quite\nlarge. What little we know about this problem in the Great Lakes does not\nagree with such an assumption. A crude analysis of frictional effects, based\non empirical information on friction in a mixed layer, leads one to very\ndifferent conclusions from what some of the friction dominated theories\nwould predict. What we clearly need is further fundamental knowledge on\nturbulent friction in the nearshore zone which would enable us to assess the\nprobable importance of the kind of secondary circulation mentioned above.\nAnother point in connection with the long-term average lake circulation\nproblem already referred to above is that in pollution dispersal problems we\nare concerned with Lagrangian properties of the flow. In simple terms, how\n30","does a given water mass get from Niagara to Toronto, or vice versa. It is not\ncertain that averaged data from fixed current meters tell anything at all\nabout the Lagrangian circulation. No long-term experiments have so far been\ncarried out to determine the relationship of Eulerian and Lagrangian average\nvelocities, and we cannot really predict where a batch of pollutants released\nnearshore would end up in a few days.\nIn the design of sewage outfalls or water intakes, it is necessary to\nmodel the dispersal of effluents in the nearshore zone. We have some informa-\ntion on diffusion parameters, but these have so far not been related to the\nspecific flow structure of the coastal zone. The strong coastal jets illustra-\nted\nin figures 7 and 8 above are certain to influence nearshore diffusion in\nimportant ways: When one part of a diffusing batch goes faster than another,\nthe batch becomes elongated and its dispersal may be expected to speed up.\nThere is no systematic quantitative information on similar effects.\nFurther reflection on a number of the above topics leads one to the con-\nclusion that our greatest current need is for fundamental understanding of\nvarious key physical processes operating in the coastal zone. The IFYGL\nobservations were designed essentially to elucidate the large-scale, lake-wide\nresponse of Lake Ontario to such forcing events as a major storm. The results\nof these observations have led to a satisfactory understanding of the first-\norder flow pattern. The next step is not more large-scale observation, but\nwell-focused experiments aimed at such fundamental problems as density fronts\nand turbulent friction. We need a great deal of thought and depth in our\nnext approach, rather than breadth and extensive coverage.\n31","1.4.1 Response - J. T. Scott\nGabe Csanady mentioned experiments in separate scales, and I completely\nagree. Future experiments should be designed along the lines of his final\nremarks. That is, we should concentrate on well-focused experiments on\ndifferent phenomena.\nWe do not have all the results in from IFYGL, so we cannot really\ndetermine all of the interactions between scales. But I have a feeling that\nscales are going to be very difficult to separate for certain phenomena. For\nexample, Csanady showed that large-scale features and waves very definitely\naffected the nearshore circulation. This is the problem in powerplant\nsiting studies where, by law, you have to go out and put up a current meter\nto obtain a year or so of data. You have no idea what is going on in the rest\nof the lake. The same sorts of remarks apply to what he said about upwelling\nand downwelling. These are also to some extent governed by features of the\nlarge-scale circulation.\nI want to comment on what some of the others have said. A lot of those\ncomments fit together. First, Cliff Mortimer's comments about prior instrument\ndevelopment and reliability testing go back to the Texas Instrument experience.\nThey were late in entering the IFYGL program, and there were difficulties in\ninstrument development. This Laboratory (GLERL) could be the mechanism for\nearly instrument package development. Perhaps GLERL could develop a set of\nbasic instrumentation that could be utilized in different programs much like\nCCIW does now. CCIW had a lot of experience with their instrument systems\nbefore IFYGL, whereas we did not. However, I am not disappointed with some of\nthe data from IFYGL.\nAnother comment relates to what Bob Pickett and John Bennett said on\ngyres. I have ideas which I think are slightly different from theirs. My\nideas are that large shore-bounded waves migrate cyclonically around the\nlake. They decay and a new event starts. This results in that mean\ncyclonic gyre basically because the short-term fluctuations are somet imes\nlarger than the mean.\nIf people are talking to each other, a lot of these differences can be\nironed out; but when we are separated, we all develop our own ideas.\nBennett did say that when modelers work together with the data people, better\nresults can be achieved. I agree with him and think that another role of\nthis Laboratory could be that of getting people together like this more often.\n32","I want to go back to something Clifford Mortimer said when using his\ntable; he gave \"pluses\" for \"nearshore circulation\" in column II. Perhaps\neventually we will get one in column III when we have more interaction with\nthe modelers. I am surprised he left us out of column I because IFYGL did\n\"reinvent the wheel\" in many cases. The nearshore circulation and coastal\njets that Csanady predicted in several different theories were rediscovered.\nIn the future we may \"rediscover\" the long waves that both Mortimer and Csanady\nhave been writing about for many years. Perhaps we put a \"tire\" on the \"wheel\"\nof their earlier work.\nBob Pickett made an interesting point, starting some discussion which will\nprobably continue. It was on the operational aspects of real-time data. This\ninterests me very much. I put myself on the side of \"real-time\" information so\nthat we can get a quick look at results. That was one of the problems of the\nTexas Instrument system. They were attempting to build a real-time system\nin a short amount of development time. We did not achieve the desired result,\nbut as Pickett pointed out, we got the data processing time down from several\nyears to about two. The Canadians did better with the older, more fully tested\ntechniques, but real-time capability has more possibilities for the future.\nBeing able to get the data for early spot analysis can be valuable if only for\ninstrument checking, redesign of experiments, and help in modeling. Another\nthing this Laboratory can probably do pretty well would be to develop this\ncapability.\nMy last point relates to what Dave Chandler said on whether to undertake\nsmall or big experiments. I lean toward the large cooperative programs though\nnot necessarily as large as IFYGL. I saw many indirect benefits coming out\nof the IFYGL program, both from a planning point of view and from the point\nof view of the impact in the scientific community.\n33","1.4.2 Response - G. E. Birchfield\nFirst, I want to make a general comment. It seems to me that a signifi-\ncant event in environmental scientific research has taken place recently.\nThat is the establishment of this Laboratory. It is a relatively unusual\nevent in that it is a scientific Laboratory. It has, as part of its program,\nscientific objectives. In the present orientation of the Federal Government\nthis should be given all possible encouragement and support.\nFor a second general comment, I want to refer to history for a brief\nmoment. Prof. Mortimer can perhaps sharpen my numbers. With regard to the\nIFYGL program, one of the first activities in the environmental area which had\nsimilarities with the IFYGL objectives was the voyage of the H.M.S. Challenger.\nThat started out in 1872, or about 100 years ago. My reason for mentioning\nthis is that I think the Challenger was one ship at sea for 3 or 4 years\ninvolved in collecting the first genuine and valuable oceanographic data set.\nIt collected physical, biological, meteorological, and I think, some\ngeological information. I believe it was at least a good 20 years before the\nfinal volumes of analysis of the H.M.S. Challenger's data were published.\nIn that view, the red line extending to 1977 that Dr. Aubert showed us is for\n6 ships and 600 scientists. We have an enormous amount of data here, and we\nare planning to analyze it in very quick order. I then ask the question,\nIs the digital computer enough to compensate for the difference in time and\nscale, to cope with the amount of data we have, and to get what is valuable\nout of the data?\nGoing to the specific area, I would first like to say a few words, with\ntongue in cheek, about John Bennett's comments. It seems to me that all our\nquestions should be solved by numerical models, such as Joe Simon's and\nJohn Bennett's They take the full equations of motion without dropping any\nterms. They put a lot of resolution in the vertical and a lot of resolution\nin the horizontal, and integrate with good initial conditions. Why is not that\nthe end of the story?\nAs Bennett pointed out, there seems to be something they do not understand.\nThe purpose of repeating that comment is to point out that one does not have to\nthink of modeling in terms of numerical models in which you throw in everything\nincluding the kitchen sink and the scouring pads too. Are we not going to be\nable to formulate large-scale, semiquantitative models of the major processes\nthat are going on in the lake due to particular kinds of forcing? Or\nqualitative models? For example, if you look at meteorology, you have synoptic-\nscale models of the Bjerknes school for fronts. Can we construct a model of\nthe response of a lake to surface waves, to the barotrophic response, and\nto the currents and thermoclinal response in a qualitative manner that can be\nunderstood? Is the problem so difficult that we cannot use a simple\nqualitative model of the response of a stratified lake to a particular typical\nwind forcing? I would say we can and that such models would be of value.\n34","We are moving toward developing such semiqualitative models by trying to identify\nthe processes that are important for these models. I would assign these models\na name, something like synoptic lake models--synoptic in the sense that it is\na qualitative physical model of circulation, including the coastal zone.\nSpeaking of the coastal zone, I would like to express some really serious\nreservations about treating the coastal zone as a black box. There are processes\nthat appear to occur in the coastal zone, but that are intimately related to\nwhat is going on in the lake as a whole. We really know very little about the\nexchange processes between the coastal zone and the deep water. The theories of\ncoastal flow that Csanady developed--the baroclinic and barotrophic flows--are,\nas he said, a first approximation to the response. However, because of the\ncharacter of the observations we have at our disposal, we are in a rather serious\nstate of ignorance. One could ask, are the coastal flows dynamically stable?\nAre they baroclinically or barotropically unstable? If they are, what is the\ngrowth time of disturbances? If the growth time is sufficient, does one really\nhave an exchange of mass with the coastal zone through such instability?\nAre\nthese jets or coastal flows stable over a long time compared to the average\nperiod in which cyclones go by? If the time scales on coastal currents are\nlong compared to that period, can we guess that exchange processes have only\nan episodic character when there is a reversal and interruption of the so-called\ncoastal flow?\nI guess what I am really saying is that we do not have a very good three-\ndimensional picture of flow in the coastal zone. We have some good and\ninformative cross sections of the flow. In the field year we had about five\nscattered around a 3,200-km coastline. To really investigate the coastal zone,\nwhich is undeniably very important from the practical point of view, some effort\nis needed to get a three-dimensional picture of the coastal zone. This is\nparticularly true in regard to mixing or exchange of water between the coastal\nzone and deep water.\nThe last thing I want to say is in a somewhat different area that is not\nappropriate under \"coastal water movements\"; but, in looking over the agenda,\nI find no emphasis on other kinds of coastal processes. In particular, I am\nthinking of the transport of sand in the shore zone. Very closely related to\nthat is erosion of the coast. IFYGL was not really involved in that area of\nresearch, and it might be something to consider further.\n35","1.4.3 Response - T. Green\nI find myself, in the position of third responder, agreeing with those\nwho agree. Thus, I have relatively little new to say here, but will check off\na few points. In the first place, I have essentially no IFYGL experience. You\ncan look upon this as either a fresh opinion or naivete. However, the University\nof Wisconsin has been studying coastal currents on Lake Superior for a number\nof years, using ships, aircraft, and moored instruments; I speak from experience\ngathered there. Primarily I have been, and my bias is toward, providing\nbuilding blocks for modelers. I also have a definite bias toward small-scale\nprocesses. This will certainly show in what I say. In my opinion, available\ntheories have far outstripped the field work; we are really in need of some\ndetailed field experiments, i.e., the dense arrays that Gabe Csanady has been\ntalking about.\nTwo people have spoken before me about the problems of real-time data\ntransmission. I agree that this presents a problem, but would hate to\nsacrifice high-frequency current observations because of that. We should\npreserve them, if at all possible. I also am sympathetic with the problem\nof getting poor data from the field. To me, the only way to solve that is for\nthe data taker to be the data analyst. Otherwise, people simply are not\nsufficiently motivated.\nI agree that it is important to separate the deep and coastal scales.\nHowever, even in the coastal zone, there are various scales. A river plume or\na thermal plume has, say, a 1-km scale. In a sense, that subzone is\nthe most important zone because it is closest to the shore. But its scale is\ndifferent from the 10-km scale that we normally associate with the\ncoastal zone. Separating these scales and their interactions will take a\ndepressingly large amount of work. I think my attention would be concentrated\non scales that are even smaller than those of the coastal chain mentioned by\nothers. I am talking about the 200-m to 1-km scale. Here, I think we can\nand should make direct estimates of transports due to turbulence. When\nwe look at this scale, simultaneity is crucial. This is unfortunate, but I\ndo not see how to get around simultaneity and dense arrays of instruments.\nI\nalso think that the longshore direction is important; even on a kilometer\nscale things change so significantly that we probably have to monitor the\nlong-shore changes with as close a spacing as we use with offshore changes.\nI do think the question of inshore-offshore exchange is a crucial fluid\ndynamics question in the coastal zone. I also think it is the most crucial\nfrom a practical viewpoint. This is fortunate; it should motivate both the\nscientists and the policy people to attack the problem. It will not be\ninexpensive; we will need dense arrays in the coastal zone. Isolines of\nconcentration of instruments should probably be circles centered at a point\non shore. We would give some attention to the whole lake, but would\nconcentrate at a point along the shore and decrease our attention outward\nfrom that point.\n36","There are intriguing questions regarding the importance of instabilities\non coastal zone fronts. Then fronts must be important to the exchange\nquestion. It is hard to believe that a temperature gradient on the order of\n5° in 10 or 50 m is not important dynamically. Then the question is whether this\nis a stable or unstable situation, and what is the time scale if it is unstable?\nCoupled with that question is the importance of nearshore upwelling. Again,\nthis is something that must be measured in detail. I think in all of this, as\nJon Scott said, we certainly do not want to reinvent the wheel. Coastal ocean-\nographers are quite a way ahead of us in many respects, and we have a lot to\nlearn from the cue experiment off Oregon and from people working in the Straits of\nFlorida. I do not think we should be reticent in taking these things over\nto the Great Lakes.\nI only have one final comment. We seem to be talking about another IFYGL,\nor whatever it might be called. I think there are three things we can argue\nabout. First, we can argue about the site; second, we can argue about what\nwe will do; and third, we can argue about the organizational framework within\nwhich we are going to do it. I think the latter two of these three things\nwill take a good deal of time. I am not convinced that it would take much time\nto pick a site. I would argue for picking a site relatively early in this\nprocess because some of us have current meters, for example, that we can put\nin that place to start generating time series. If we were to pick a site by\nnext summer, we could start getting some measurements and have some data. Then\nwe could say, when we go to the big program, whether or not the data\ncollected during the program were typical. We would hope that it would be,\nbut we are never sure without a longer time data set at a few points.\n37","1.5 SIMULATION OF AQUATIC ECOLOGY - C. Schelske\nAquatic ecology constitutes such a complex array of scientific disciplines\nthat it may be better to take a slightly different tack than has been taken by\nsome of the previous speakers. I will not attempt to elaborate on a specific\ngroup of research objectives now since these should be an outgrowth of the\nworkshop sessions that will follow. To confuse the situation further, worth-\nwhile specific research objectives could be set since there are large gaps in\nour understanding of ecology in large bodies of water like the Great Lakes.\nBut my approach will be to speak to research problems with respect to the\nGreat Lakes generally and not to specific research objectives.\nDifferences among the five Great Lakes are sufficiently great in a number\nof respects that each lake must be considered separately. Specific research\nproblems can be considered for each lake, but our major thrust should be on\nwhat questions are appropriate to ask and what problems can be studied. As we\nall know, formulating experiments and asking questions with testable hypotheses\nare the difficult parts of this task. Before we can ask questions, design\nexperiments, and make measurements, we must identify the problems with the\ngreatest importance. I have attempted to classify the problems into two main\ntypes: experimental and descriptive.\nAn experimental study is one in which observations are made under at least\ntwo conditions so that one condition can be compared to another. of course,\nin the strictest definition of experimental work, this is called a controlled\nexperiment. I think it is quite obvious from what has been said earlier today\nthat making certain observations or measurements can be considered in the\nexperimental sense, particularly if we formulated a testable hypothesis to go\nalong with that set of measurements. A descriptive study differs from an\nexperimental one in that these measurements are used to determine environmental\nconditions in space and time. Both approaches are important. Some of us\ntalk about surveillance and monitoring, and these might be used interchangeably\nwith descriptive studies. I suspect that these experiments are either so\ncomplex or so obvious that they are not worth discussing in this group.\nI\nwould like to mention, however, that experimental approaches, controlled\nexperiments, have been used in Great Lakes research. There are two examples.\nOne is the work that has been conducted during the past few years at the\nUniversity of Michigan, in the Great Lakes Research Division, on the effects of\nnutrients on phytoplankton production and species composition. These experi-\nments have been conducted in the laboratory with small beakers and in situ with\nlarge plastic bags to simulate nutrient enrichment in nature. More recently,\nCanada Centre for Inland Waters (CCIW) has been working with \"limno-corrals,\"\na slightly different type of experimental system. Our bags were suspended in\nthe water, but the sides of the limno-corral extend from the surface of the\nwater down to the bottom. Different types of experiments and hypothesis can\nbe tested with the two types of systems.\n38","The descriptive study is important, if for no other reason, for purposes\nof assessing water quality. Mentioning descriptive studies, however, produces\nunfavorable reactions from certain members of the scientific community. The\nmost critical reactions are to the effect that such studies are undertaken\nbecause more important scientific questions cannot be formulated. To lessen\nthe impact of such criticism, we need to justify these measurements. They can\nbe justified from a scientific point and also from the point of management. In\nother words, they are needed so that we can assess what has happened in the\nenvironment or so that we can compare present conditions with past conditions\nand present conditions with future conditions.\nWe can justify the rationale for descriptive measurements by categorizing\nthem into three groups of variables: causal, integrative, and descriptive.\nI would like to restrict this part of the discussion, illustrating these var-\niables from the standpoint of eutrophication as it ties in nicely with the\nwater quality work which is a part of this meeting. I have thought quite a bit\nabout this type of rationale, and in addition, it has been considered through\nan International Joint Commission committee, the Research Advisory Board\nstanding committee on eutrophication, which is chaired by Richard Vollenweider.\nTwo of the members of the committee are here--Fred Lee and Al Beeton.\nCausal variables, the first category, are those that stress or force the\nsystem when their inputs are increased and produce what is usually referred\nto as undesirable effects on the system. Phosphorus is undoubtedly the most\nimportant variable in this category if we are considering eutrophication.\nPhosphorus is the principal causal variable in eutrophication processes as\nits supplies limit the growth of phytoplankton. As a consequence, increases\nin the inputs stimulate algal growth, producing accelerated eutrophication\nand some associated changes in the system which may be categorized under\nintegrating variables. We need to know more about other causal factors or\nvariables in the system and their effects on eutrophication even though their\nimportance may seem secondary. Heat, trace elements, organic materials of\nvarious types and with various functions, and even conservative elements may\nplay this secondary role. There is little evidence from existing work to\nevaluate the relative importance of each of these.\nIntegrating variables simply are those which change in a predictable\nmanner as a result of increased inputs of a causal variable. I would like\nto mention only two integrating variables, silica depletion and oxygen\ndepletion, to show how they can be used in extremely different parts of the\nGreat Lakes. One of the reasons I am here today is due to an integrating\nvariable, and as most of you know, that integrating variable is silica\ndepletion in Lake Michigan. Increased phosphorus inputs to Lake Michigan\nhave stimulated the growth and increased the standing crop of diatoms, which,\nin turn, have depleted silica supplies in the lake. In this case then, the\ndecrease in silica concentration is a very good measure of eutrophication\n39","because it reflects the inputs of phosphorus, the causal factor in the\neutrophication process. I might add that, even though some of this work could\nhave been accomplished through descriptive studies, a large part of the under-\nstanding of this particular process has come through experiments conducted\nin large plastic bags.\nIt is perhaps important to note, for the purposes of the meeting, that\nthe success we at the University of Michigan had with this particular research\nprogram and other related programs was due to several factors. These factors\nincluded an organization dedicated to Great Lakes research, ship facilities\nto get on the lakes, and a staff of scientists and technicians with knowledge\nand capabilities that could be applied to specific problems. We were able to\ncarry out these studies because Professor David Chandler and other people at\nthe University of Michigan dedicated a considerable amount of effort to\nproviding facilities and a critical mix of people. Gene Stoermer was available.\nHe knew how diatoms behaved in the environment and, more importantly, could\nrecognize a diatom when he saw it. Sometimes ecologists have a little diffi-\nculty recognizing specific organisms, and we tend to view them as black boxes.\nThe important points are that we had a technical staff that may or may not be\navailable at some organizations and that a lot of expertise is needed to con-\nduct research. It may or may not be necessary to have a large organiza-\ntion to do this research, but it is essential to get a critical mass of\npeople together. In recent years, particularly in oceanographic work, large\nprograms have been conducted successfully either by cooperation among institu-\ntions or individuals, so I an not advocating building up large research\norganizations. There is, however, a definite need for large research\norganizations or research groups, and these will especially be needed if we\nare to attack the types of problems being discussed here today.\nIn addition to silica depletion, oxygen depletion is an integrating\nvariable. The best example of this is hypolimnetic oxygen depletion in the\ncentral basin of Lake Erie. The importance of this integrating variable has\nbeen documented by scientists at CCIW and through a joint Environmental\nProtection Agency (EPA)-CCIW study known as Project HYPO. The causal variable\nin this case was also increased inputs of phosphorus as oxygen depletion is\ndue to the secondary effects of the growth of phytoplankton or to the decom-\nposition of organic matter. Silica depletion occurs in the photic zone in the\nupper layers of the lake. Oxygen depletion occurs in the hypolimnion, the\nbottoil of the lake, as organic matter decomposes. Measurement of silica\nserves no purpose as an integrating variable in Lake Erie since, among other\nreasons, many of the phytoplankten of Lake Erie are not diatoms. Likewise,\noxygen depletion has little use as an integrating variable in Lake Superior\nwhere very little oxygen is consumed due to the low productivity, but silica\ndepletion would be very appropriate because most of the phytoplankcon are\ndiatoms. Integrating variables, therefore, will vary from lake to lake.\n40","The final category is descriptive variables. These would include those\nthat can be justified for such ecological reasons as describing seasonal\nchanges, comparing lakes, and determining long-term changes in the system.\nVariables under this category would include measurements of chlorophy11,\nphosphorus, nitrogen, and carbon and abundance of organisms, including\nphytoplankton, bacteria, zooplankton, benthos, and fish. A number of studies\ncould be cited to justify the importance of studies on the taxonomy and ecology\nof organisms in the Great Lakes. Probably the best known from the standpoint\nof management would be fishery studies.\nIn addition to determining what variables to include under the descriptive\ncategory, another major problem is determining the frequency and extent of sampl-\ning needed for these particular variables. How many samples, at how many\nstations, at how many depths, and at what time interval are important questions\nin our current research and data gathering efforts. It is obvious that the\nfrequency of sampling will vary with the variable. Bacteria, phytoplankton,\nand zooplankton with relatively short generation times must be sampled\nfrequently--weekly, daily, or even hourly--to describe their dynamic properties.\nAt the other extreme are conservative parameters, or variables like chloride,\nwhich have changed on a lake-wide basis in the past 60 to 100 years. But these\nchanges may be too small to be measured on a yearly basis, so a few measurements\nper year may be sufficient for chloride. The sampling scheme then cannot be\nuniform for all descriptive variables and has to be varied according to the\nobjective of the particular study.\nNo ecologist could make a presentation like this without stressing the\nimportance of studying interrelationships in the aquatic ecosystem.\nEven though the effects of eutrophication are most obvious in the primary\nproducers and in certain chemical parameters, at least as I presented them\nhere, studies of eutrophication cannot be restricted to the effects of nutrient\nadditions on phytoplankton. Studies are needed to determine the changes in\nthe system resulting from the initial perturbations at the phytoplankton-\nchemical level and their resultant effects on the food chain. Since all\nprocesses in the system are interdependent, one needs to know which processes\naffect eutrophication.\nI will conclude by reflecting on what I think we have learned from IFYGL.\nSome of this has already been covered by other people. First of all, there\nis an obvious need for advanced planning and lead time. Funding for contracts\nand grants should start at least 6 months to 1 year prior to the initiation\nof the field year research effort so that people would have adequate time to\nperfect and test models. Given this amount of lead time, it would be easier\nto adjust schedules for logistic support if there were unforeseen delays in\nthe scientific program.\n41","There is an obvious need to assemble a critical mass of investigators\nwith appropriate expertise. For large programs, this means identification of\nindividuals or groups with such expertise and the need for coordination of\nefforts. Such expertise must be available when we need it. Particularly in\nthe university community, we cannot support large research staffs on university\nbudgets, so advanced planning or continuing support is needed if large-scale\nuniversity participation is desired.\nWe need to recognize that the Great Lakes differ from one another. This\nis probably obvious, but it is important because it may be feasible to under-\ntake some research problems on some lakes and not on other lakes. As an\nexample, it may be more appropriate to study the materials balance in Lake\nMichigan than in Lake Ontario because most of the tributary inputs are quite\nsmall compared to the input and the output from the Niagara and St. Lawrence\nRivers. Certainly the success or failure of a project undertaken on Lake\nOntario may not be the same if undertaken on another lake.\nFinally, a need exists to refine and improve the advisory process. In\nIFYGL a lot of the work was done through this mechanism, and those advisory\ngroups did a commendable job. On the other hand, I think we should always\nlook for improvements and for other mechanisms. My suggestion will be slightly\ncontroversial. We need to consider new approaches to managing research, so I\nwould like to propose an alternate or additional system from the advisory\nstandpoint. The reasons for proposing such a system include minimizing the\nchances of undertaking trivial problems and maximizing the benefits obtained\nfrom funding. We all realize that limits on funding will always be a limiting\nfactor. Traditionally, science has progressed through the work of individuals.\nA dichotomy developed in this approach about World War II, when people began\nto talk about big science and little science. Ecologists have thought about\nbig science, but probably have not made as much progress as desirable in this\nparticular area. The usual approach to big science has been to take a lot of\nlittle science and put it together. By doing that we may not consider very\nimportant questions. An alternative to this approach might be to utilize\nindividual researchers in a slightly different way. Traditionally, a scientist\nhas always been someone who could gather and synthesize his own data. That\nbecomes very expensive, especially on large lake problems. An alternative\nwould be to put individuals to work on different tasks. These tasks would have\nindividuals working on planning experiments and evaluating data. It would still\nallow them to participate in the overall scientific program. This would result\nin much better designed experiments than we have now. People would be thinking\nabout questions to answer, rather than the data they can collect. In the long\nrun, with proper incentives, the individual probably would get more satisfaction\nout of this particular approach than by working alone on a complex problem with\ninadequate resources. This approach would not eliminate the individual researcher\nas there are many worthwhile problems that can be pursued by individuals. These\nadvisory groups would have to determine which problems could be undertaken by\nindividuals and which would have to be undertaken by larger groups.\n42","1.5.1 Response - A. M. Beeton\nMuch of the work in ecology on the Great Lakes is still in the descriptive\nphase. We really need a lot more imagination than we have had in the past,\nespecially if we, as biologists, are going to take advantage of the kind of\nwork being done by the physical limnologists and the people involved with\nmodeling. So, for example, if we had an understanding of what is going on\n2\nunder 1 m of the open lake in terms of biological interactions, we would\nthen be in a much better position to interpret much of the data that have\nbeen collected in a number of large, lake-wide surveys. Lake-wide surveys\nhave a very long history. Several were conducted on Lake Erie in the 1920's\nand early 1930's and on Lake Michigan in the 1930's. Various surveys have been\nconducted within the past 20 years, and as a consequence, there are reams of\ndata waiting to be interpreted, but we cannot interpret these data because we\ndo not understand some of the basic mechanisms and interactions.\nLooking at some things that we could perhaps paint with a broad brush, I\nam intrigued by the differences in the inshore and offshore conditions in\nbiology and chemistry. I think this is an area in which we can provide a lot\nof data that will fit in very well with some of the things Gabe Csanady was\nmentioning this morning. To get a handle on water quality in the inshore area,\nwe must understand something about the exchange rates between the large mass\nof relatively high-quality water that sits in the middle of most of these lakes\nand the water in the inshore areas. It appears that the quality of the water\nin the inshore area is determined by point source discharges as well as diffuse\nsources and by sediment water interactions, biological activity, and mixing of\ninshore and offshore waters. We do not know how much phosphorus might be tied\nup in the clay minerals that are redistributed by every storm that comes along\nor how much phosphorus may be removed by the organisms and just how much of\nthis is dispersed into the lake by exchange processes between inshore and\noffshore waters. Each of these factors needs to be considered, but if we did\nhave reasonable estimates of exchange rates, then we might be in a position to\nstart getting a handle on the role of biota in removing the nutrients. For\nexample, if you look at some of the conservative properties, such as chloride\nand sodium, and compare their distribution with some of the nutrients, you\nfind that, while there are relatively high levels of some of the conservative\nproperties coming into the nearshore water, they are dispersed rather rapidly\nand concentrations do not differ greatly within one lake. When you look at\nthe major nutrients, concentrations in the nearshore water are often 10 times\nwhat you find in the offshore waters; we cannot explain the lower levels of\nsome of these out in the open lake and higher levels inshore just due to\ndilution alone. The loading rate, the biota, and some other mechanisms are\ninvolved in removing the nutrients as well as recycling those inshore and\ntherefore keeping much higher concentrations inshore. We do not understand\nwhat these mechanisms are. I think this is very important if we are going\n43","to be able to contribute in a meaningful way as biologists to an understanding\nof the Lakes. This certainly fits in with our concept of eutrophication. We\nhave more than adequate evidence of changes in the Great Lakes. This has been\ndocumented extensively. We know that eutrophication has progressed from the\nshore lakeward; this is logical because point sources are along the shore.\nCertainly we have plenty of evidence that eutrophication progressed from west\nto east and from shore lakeward in Lake Erie, and this is what is happening in\nLake Michigan. Our conceptual model has been that inshore-offshore differences\nare pronounced, and we have sometimes talked of Lake Michigan as if it were\nlike a doughnut where you have the inshore waters with somewhat degraded water\nquality and a big mass of higher quality water out in the central lake. That\nis all right for a conceptual model, but actually what we probably have are\npoint source inputs from a number of metropolitan areas where the water is\nprobably degraded. We really need to understand the loading rates in these\nareas and the rate at which these inputs can be dispersed. This is very\nimportant from the biological and water quality standpoint and fits in with\nthe physical limnology kind of thing that we need to tie together.\n44","1.6 SIMULATION OF WATER QUALITY--Carl Chen\nJust a week ago, our group completed a survey of water quality modeling\nas it applies to the Great Lakes. The work was done for the Corps of Engineers.\nOur goal was to define the various methodologies that have been developed to\nevaluate the effectiveness of various waste water management problems. In\nthis study we reviewed several methods of determining water movement and\ntransport, including direct measurement, scale models, steady-state models,\nand time-dependent models. We then evaluated the state-of-the-art of water\nquality models. Some of them are designed for long-term projection of various\nquality parameters, like salinity and phosphorus, and some of them for waste\nheat. We also evaluated a whole slew of water quality simulation models that\nhave been developed and may be applied with some minor adjustments to the Great\nLakes problems.\nBased on this review, we concluded that water quality models have been\nadvanced greatly. There is no need for people to reinvent the wheel since,\nwith modification of some sort, one can apply it specifically to various lakes.\nAs has also been pointed out by others, each lake is different. The approach\nto be taken for each lake might be a little bit different, but the underlying\nconceptual framework is strong enough to make it transferable.\nWhen we talk about water quality models, we cannot talk about water quality\nalone because the water quality is influenced by biology. Right now, the water-\nquality modeling technology has been advanced from the traditional biochemical\noxygen demand/dissolved oxygen relation type of analysis to include more and more\nbiological parameters. This is important.\nEqually important, we cannot talk about water quality models without\nhydrodynamic transport models. The water quality models always require\na\ntransport model to drive them. A transport model is the prime mechanism to\nmove the materials and distribute them in space where they influence the biota\nand the biota in turn influence the water quality.\nI would like to talk about some of the basic concepts in water quality\nmodeling and the approaches being taken to date and will present some questions\nwhich must be answered to improve the models. I would like to throw out some\nideas on how we might model a Great Lake, what kind of transport mechanism we\nmight need, and what kind of biological information we would like to have.\nThe purpose is to simulate discussion and receive input from the audience.\nFigure 9 represents a simplified conception of the interactions which\nbear on the water quality of an ecosystem. The figure shows many simultaneous\ninteractions between biotic and abiotic entities of an ecosystem. Nutrients\nsimulate the growth of phytoplankton, phytoplankton consume nutrients, and\nso on. Basically there are two major types of interactions or processes.\nThe upper half of figure 10 lists the physical processes taking place to\ninfluence the distribution of pollutants. Physical processes include advection,\n45","MAN INDUCED\nNATURAL\nWASTE LOADS\nINPUTS\nBACTERIA\nPHYTOPLANKTON\nWATER\nQUALITY\nDEAD ORGANISMS\nDETRITUS\nZOOPLANKTON\nEXCRETE\nFOOD\nDEPARTMENT\nDE\nBENTHIC\nFISH\nANIMAL\nFOOD\nFigure 9. Definition of an aquatic\necosystem.\n1.\nPhysical Processes\na. Advection between segments\nb. Diffusion between segments\nC. Sedimentation from the segment\nd. External input to the segment\ne. Output to external from the segment\nf. Reaeration\ng. Solar insolation\nBiochemical transformation, uptake, and release associated with\n2.\nthe following:\nBacteria\nO2\nNH2\nNO\nNO\n3\n2\n3\nBiological Oxygen\nCO\nDemand\n2\nDetritus\nNH3,\nPO\nCO2\nAlgae\nZoo-\nFish\n4'\nplankton\nPO\n4\nBacteria\nDetritus\nBenthos\nFigure 10. Important ecological processes for modeling.\n46","diffusion, sedimentation, input (namely waste input or atmospheric input), and\noutput. The other types of processes are chemically or biochemically mediated\ntransformation, uptake and release of nutrients, bacterial degradation, etc.\nWhile the approach may be similar, models can differ in the amount of\nthe biosphere incorporated into a model to describe adequately a water quality\nproblem. There have been models taken up to the zooplankton level in the biota\nthat exclude the computation of oxygen. A problem occurs here, however, be-\ncause, if you do not know oxygen, you do not know if bacteria are going to be\naerobic or anaerobic. The problem is how to increase the parameters such that\nwe can correctly do the biology and the water quality simultaneously. We like\nto have a comprehensive but tractable model.\nHow do we learn enough about processes to do basic modeling? The first\nthing is to go to the laboratory and learn something about what is going on.\nIf we want to study algae, we put algae in a beaker. If we want to study\na chemical interaction, we measure what comes in and what goes out. Through\nthat, we develop two basic principles. The first principle is the conservation\nof mass law; i.e., mass has to be conserved. It might transform from one form\nto another, but mass has to be conserved. The second principle is the kinetic\nprinciple; i.e., when transformations occur, they do so at a certain rate.\nWe like to know how fast algae is growing. How many nutrients are consumed\nfrom water to conserve mass?\nTo apply such principles to prototype simulation, the water body has to\nbe divided into small hydraulic elements such that each one of these hydraulic\nelements can be approximated by the laboratory condition (fig. 11). The natural\naquatic ecosystem can thus be viewed as a series of interconnected hydraulic\nelements. Water and mass can be transferred from one element to another.\nBased on kinetic and mass conservation principles, it is a classical\nsituation to write a so-called mass balance equation. The equation says how\nfast mass in the element is changing due to physical and chemical processes.\nThe following equations can be solved using digital computers:\n1. General Mass Balance Equation for Abiotic Substances\n+ EQ in C in - EQ\nADVECTION DIFFUSION INPUT\nOUTPUT SETTLING REAERATION\nDECAY\nTRANSFORMATION\nUPTAKE\nRESPIRATION\nNH3\nNO\nNO\nBYPRODUCT\nRELEASE\n2\n3\n47","c'o'\nC2\n2\nCn\nA. a continuously stirred tank reactor, CSTR\nOUT Q IN\nC1\nTDS\nC2\nBOD\nC3\nDO\nTEMP\nAj\n=\nA\nALGAE\nQj\nQ\nZOU\nVOLUME\nFISH\nB. an idealized hydraulic element\n= Concentration of some constituent\nC\n= Flow through a face (04, , oj), out (Out) or in (ain)\nQ\nCross sectional area of face i (A) or j (Aj)\nA\n=\nAS\nSurface area\n=\nS\nTDS = Total dissolved solids\nBOD = Biological oxygen demand\nDO = Dissolved oxygen\nFigure 11. An idealized hydraulic element versus a laboratory-\nstirred tank reactor.\n48","2. General Mass Balance Equation for Biota\ng + +\nGROWTH\nSETTLE\nGRAZING\nRESPIRATION DEATH\nwhere\nV = Volume\nC1 = Concentration of constituent 1\nQ = Flow through a face i(a), in (ain), or out (Out)\nE\nDiffusion coefficients\n=\nA\n= Cross sectional area of face i\ndC1/dx; = Concentration gradient of C1\nCin = Concentration of C1 in the inflow\nS1\nSettling rate of C1\n=\nKL\nReaeration coefficients\n=\nAs\nSurface area\n=\nc+ = Saturation concentration of C1\nKd,1\n= Decay coefficient of C1\nKd,2\nDecay coefficient C2\n=\nH3 R Respiration factor of between biota C3 C3\n= Growth rate of biota\nF3,1 = Conversion C1 and C3\n= rate\nH1 = Specific growth rate of C1\nR1\n= Respiration rate of C1\nM1 = Mortality rate of C1\nH2 = Specific growth rate of higher trophic species C2\nF2,1 = Conversion factor between C1 and C2.\n49","The following figures show different approaches to segmenting different\nkinds of water bodies. Figure 12 is the way that has traditionally been used\nto segment a river. Figure 13 is a way to represent a small reservoir. The\nsmall reservoir is usually upstream of a river. It does not receive much waste\ninput. All the water quality influence is in the vertical direction due to\nthermal stratification and overturn. The reservoir is therefore cut into\nhorizontal slices. Some river-run type of reservoirs can be cut into reaches\nand then into horizontal segments (fig. 14).\nHow does one go about segmenting the Great Lakes? One approach was\n2\nHe recognizes that a lake has to be divided into a\nconceived by Canale\nlittoral zone and a central zone as shown in figures 15 and 16. This might\nbe too coarse, but the concept is good. Based on the concept, we can\nenvision what to do with another lake. I will use Lake Erie as an example.\nWe envision that a segmentation shown in figure 17 might be appropriate. Thus,\na more detailed spatial resolution is possible at the nearshore zone where\nthe lakes receive waste water input. By that, we can see the pollution effect.\nWhen it comes to the central lake, the horizontal spatial variation is not\nas big. We can use a bigger hydraulic element. The advantage of this type of\nsegmentation is that it can fit into the currently available computer core\nspace and also that the computer time is not excessive.\nThere are also different philosophies in the development of water quality\nmodels. Some of the modeling technology has been advanced by starting from a\nsimple one-reactor representation and progressing to greater detail. When one\ndiscovers a single reactor representation is not enough, he begins to cut the\nreservoir into two layers, and when two layers are not enough, more, and so on.\nThat is one way to do it. Another way is to go through the literature to\ndetermine the current status of modeling. The model is initially conceived\nas comprehensively as possible. After a comprehensive model is developed, one\nbegins to simplify the model to see how far he can go. I will not say which\nway is better, but these are two approaches. One starts from comprehensive to\nsimple. The other starts simple, then discovers that it cannot do the job, and\nevolves into something complicated. Eventually both approaches may merge at\nthe middle. Meanwhile, those taking the comprehensive route may be accused\nof being too ambitious or just plain unrealistic. The other group, on the\nother hand, may make a bad reputation for the modeling field. They build models\ntoo simple to be real.\n2\nProfessor Raymond Canale, Sea Grant Program, University of Michigan.\n50","b\nInflow\nX\nA\nWATER FLOW\nRain\nP\nEvaporation\nFigure 12. Physical representation of a stream.\nE\nInflow\nQout\nQin\nOutflow\nAy\nCONTROL VOLUME\nV\nSEGMENTED\nSTREAM\nSYSTEM\nOutflow","tributary\ninflow\nevaporation\ntributary\nrain\ninflow\nvertical\nadvection\ncontrol\nslice\noutlet\noutflow\nFigure 13. Geometrical representation of a reservoir.\n52","Figure 14. Segmentation for river run reservoir.\n53","Zone 4\nZone 1\nZone 3\nZone 2\nZone 5\n15. Hypothetical horizontal sectioning of Lake Michigan into\nFigure\nuniform zones (from Canale*).\nR. P. (No date), , A methodology for mathematical modeling of Covering biological\n*\nCanale, production, Report . to the University of Michigan Sea Grant Project\nJan. 1970 - Jan. 1971, 35 pp.\n54","ZONE 1\nOrganic Material\nvertical sectioning of a lake into uniform cells (from Canale*).\nBenthos\nMigrations\nCarbon Dioxide\nExchange of\nOxygen\nSunlight\nZONE 4\nSedimentation\nSedimentation\nExchange of\nOxygen\nZONE 3\nZONE 2\nFigure 16.\nZONE 1\n*","Figure 17. Segmentation for Lake Erie.\n56","Let me say one thing about what we should do in the Great Lakes. The Great\nLakes model must perform computations for a series of hydraulic elements that\ncan be arranged horizontally as well as vertically. The hydraulic element may\nbe a small cell along the shore. It could be a big one in the pelagic zone.\nAny element can accept upwelling and downwelling. It can have horizontal\nadvection which can go both ways. Mass balance computations can be performed\nfor all the important water quality parameters. Transport can either be\ngenerated by a hydrodynamic model or prescribed according to field data. Some\nof the hydrodynamic conditions may be very difficult to compute, but they are\neasy to prescribe. The object of the model will probably be the simulation of\nmean monthly water-quality conditions.\n57","1.6.1 Response - R. V. Thomann\nI think Carl Chen did an excellent job reviewing the nature of water\nquality models and their interaction with biology. I will make three points.\nFirst, what did we learn from IFYGL concerning water quality modeling\nand the interaction between water quality modeling and biological modeling?\nI think we learned that models that have no circulation in them at all, where\nlake-wide averages are taken on a horizontal plane and the model is only\nconsidered in the vertical dimension, hold a lot of promise, at least, for\nmaking long-term planning decisions on the Great Lakes. I think the dynamics\nof nutrient and phytoplankton behavior for such lake-wide situations are well\nadvanced. We are understanding more and more about the behavior of some\nof those lake-wide dynamics. The reason I say this is that the basic analytical\nstructure which Carl Chen just reviewed has now been applied to problem\nsituations that span two or more orders of magnitude in total phytoplankton\nbiomass. The analytical structure that Carl Chen reviewed has already been\napplied, we believe reasonably and successfully, to several different types of\nwater bodies. Some examples are eutrophic estuaries with maximum concentration\nof 200-ug chlorophyll per liter for phytoplankton biomass; delta regions in\nCalifornia with a maximum of 50 to 100 ug; Chesapeake Bay with concentrations on\nthe order of 10 to 50 ug/1; and Lake Ontario with 5 to 10 ug/l. We have now\nspanned almost two orders of magnitude, and applications are underway to model\nLake Huron which would be 1 to 3 ug/l. By the time we finish with these half\na dozen applications of the analytical structure, we will cover environments\nwith almost three orders of magnitude difference. The analytical structure\nhas really held up which I think says something for the ability to utilize the\nlake-wide average for planning purposes. That is point number one.\nThe second point I want to make concerns this whole notion of the\nimportance of circulation to phytoplankton dynamics. In addition to the lake-\nwide model, one of the other outcomes of IFYGL is a first preliminary three-\ndimensional phytoplankton biomass model of Lake Ontario. That work is just\nstarted, but I wanted to show you one preliminary result to illustrate a point.\nThe three-dimensional model is a rough grid five-layer model and looks some-\nthing like what Carl Chen was talking about (fig. 18). It is an attempt to\ndescribe at least some nearshore phenomena. Shore segments extend about 10 km\nout and are about 40 km long. We use the kinetics given by a lake-wide model\nwhich was verified by about 4 years of data.\nThe question of circulation came up and we hassled this back and forth.\nHow do we handle all these complex interactions we heard about all morning?\nWe took a summer and winter circulation pattern and put in some thermal bar\nphenomenon when and where we think it happens. A11 of these phenomena are\nprescribed externally in addition to the waste loads. We then ran the model.\nFigure 19 shows a cross section across the lake comprised of segments 14,\n16, and 17. Seventeen is Rochester Harber. The area is 10 m by about 40 m,\n58","Dos\n51\nee\n01\n49\n50\n52\n48\nKM\nMI\n45\n47\n65\n50 - 150 Meters\n10 O 10 20 30\n30 50\n46\n44,\n10 0 10\n43\n4 - 17 Meters\n41\n42\n40\n64\n39\nSegmentation for Lake Ontario.\n38\n37\n36\n35\n34\n33\n32\n63\n31\n29\n27\n67\n0\n30\n28\nM\nW\n> 150 Meters\nDes\n25\nof\n23\n24\n66\n26\n60\n22\n61\n19\n62\n21\n0\n20\nFigure 18.\n18.\n59\n17 - 50 Meters\n17\n15\n16\n0-4 Meters\n14\n58\n57\n13\n12\n11\n10\n-N-\n56\n9\n8\n7\n54\n6\n5\n55\n3\n1\n53\n4\n2\nan","14\nSegment 14\n30\nSegment 16\n25\n16\nSegment 17\n17\n20\n15\n10\n5\no\n30 F 60 M 90 A 120 M 150 J 180 J 210 A 240 S 270 o 300 N 330 D 360\nJ\nDAY OF YEAR\nFigure 19. Phytoplankton chlorophyll as a function of day of year for\nselected segments of Lake Ontario (0-4 m).\nwhich is a pretty coarse grid. The figure shows phytoplankton in micrograms\nchlorophy11 per liter as a function of time of year. A spring bloom develops\nthat precedes the open lake segment (16) by about 30 days, and there is a\ngradient of about 15 ug/1, which is a reflection of the fact that it is near-\nshore entrapment. Now compare model output to some observed data as shown in\nfigure 20. The figure is for Rochester Harbor. The black dots are the mean\nvalues calculated by the model. The range is what the model calculated during\nJune and the open circles are IFYGL data. The run used the same kinetics as\nthe lake-wide model. The difference is in spatial detail as shown in figure 18.\nLocal circulation, thermal bar effects, and vertical stratification were included.\nThe comparison is remarkably good. There is only one thing wrong with this.\nWe are a little uncertain as to why we did so well on the first shot. For\nexample, this program was not completely finished for this run, so the run does\nnot include any phytoplankton settling. The run also includes an order of\nmagnitude higher concentrations for the phosphorus Michaelis constant (10 ug/1)\nthan what Claire Schelske would normally consider for phosphorus on the basis\nof his Lake Michigan work. In spite of all of that and a simple circulation\npattern, it is really quite surprising that the results are so encouraging.\n60","CHLOROPHYLL a, pg/l\nCHLOROPHYLL a. pg/l\no\n2\n4\n6\n8\n10\n12\n14\n16\no\n2\n4\n6\n8\n10\n12\n14\n16\no\n0\no\nO\n10\n10\n20\n20\n30\n30\n40\n40\n50\n50\n60\n60\nSegment 15\nSegment 17\nDay 150 - 180 (June)\nDay 150 - 180 (June)\n70\n70\n80\n80\nLEGEND\nComputed for the time shown:\nObserved IFYGL data over the\ntime shown: o\nMean\n15\n17\nRange\nFigure 20. Preliminary comparison of Lake Ontario output to observed data\ntwo segments, June 1972.\n61","Point number three is that the long residence time of water in the Lakes\nprohibits any kind of meaningful testing of these kinds of phytoplankton models.\nWe cannot reduce a load and make a prediction and then see how well the model\ndoes. We are kind of describing what we have already observed in a hindcasting\nfashion. Also, there are a variety of processes that we have not even begun to\ntouch; for example, the multispecies model that many people have talked about\nand the problem of nearshore rooted aquatic plants. In spite of these\ndifficulties, the \"success\" to date leads us to think that there is considerable\npromise in the utility of these models for aiding the long-term decision-making\nprocess concerning effects of nutrient removal on the phytoplankton of the lake.\n62","1.6.2 Response - S. Chapra\nI will supplement Carl Chen's and Bob Thomann's presentations by discuss-\ning a scale of analysis which has yet to be addressed today. Although it is\nnot a scientific scale, but rather an engineering or planning scale of analysis,\nit is potentially useful for addressing some of the Great Lakes water quality\nproblems.\nAs outlined by previous speakers, the approach has been to study phenomena\non a whole lake during a year, or to resolve smaller space and time scales\nwithin a lake. Time scales of a week or less have been mentioned. Space\nscales on the order of kilometers or less have been addressed.\nThe approach I will discuss would include all the Lakes in one model in a\nmanner similar to the way hydrologists simulate lake levels. However, instead\nof lake levels, water quality problems might be addressed in time scales of\nyears or decades. This approach has been applied to water quality problems\npreviously and stems from a chloride model of the Great Lakes published\n3\nby\nO'Connor\nand\nMueller Their basic idea was that social and economic\nparameters, such as population, could be used to generate time series of waste\nsources to each of the Great Lakes. These sources were then introduced into\na simple transport model which treated each of the Lakes as continuously\nstirred tank reactors (CSTR). In this way O'Connor and Mueller made long-term\nprojects of the chloride levels due to various waste management strategies.\nAs Bob Thomann just stated, many water quality models are expensive to\nrun longer than a few years. I think O'Connor and Mueller demonstrated that,\nat least for a certain class of problem, a \"Great Lakes space-scale\" and a\ndecade time-scale could be effectively used to predict these long-term effects.\nThere have been some other applications of this approach. For instance,\n4\nGustafson modeled tritium levels in the Lakes due to nuclear power plants;\n5\nLerman has modeled strontium-90 in this way.\nIn all cases, simple transport models with simple reaction kinetics were\nformulated. If other substances such as pesticides, total phosphorus, etc.,\ncould be reasonably modeled in this way, we would gain a valuable tool to\nanswer questions about the future quality of the Great Lakes.\n3\nO'Connor, D. J., and J. Mueller (1970), A water quality model of chlorides\nin the Great Lakes, Journal of Sanitary Engineering Division, ASCE, 96,\npp. 955-975.\n4\nGustafson, P. F. (1970), Future levels of tritium in the Great Lakes from\nnuclear power generation, paper given at the 13th Conference on Great Lakes\nResearch.\n5\nLerman, A. (1972), Strontium-90 in the Great Lakes: Concentration-time\nmodel, Journal of Geophysical Research 77, pp. 3256-3264.\n63","1.7 SIMULATION OF ENVIRONMENTAL DYNAMICS OF THE GREAT LAKES - C. H. Mortimer\nIn the final agenda for this workshop, the biological section, including\necological modeling, is listed separately from \"environmental simulation.\"\nI did not realize this before preparing table 8. Without defining \"environ-\nmental dynamics\" too precisely, I intended to talk about interactions between\nhydrodynamic and ecological models, not because I can claim to be a modeling\nexpert, but because the design of an optimum interaction strategy is the most\nimportant task facing us if limnological modeling is ever to have a usefully\npredictive impact on lake management. Ignorance rarely constitutes a bar to\npublic speaking, but the results are often platitudinous. Nevertheless, I hope\nmy five platitudes (table 8) will serve to generate fruitful debate.\nThe first platitude is an attempted one sentence definition, with alterna-\ntive wordings, of the purpose of lake system modeling. The second is a\ntriarchy of three interacting boxes, a triarchy illustrating the application\nof the scientific method to acquisition of understanding of lake systems. Box\nA (top left) represents the way in which limnologists have traditionally worked\nin the past, through development and testing of conceptual hypotheses. This\nbox is a compendium of knowledge, or suppositions, or biases, that constitute\n\"what every limnologist knows.\" Box B (top right) is a fairly recent arrival\non the scene, i.e., computer manipulation of systems of equations which can\nbe deterministic, stochastic, or a mixture of the two. We could spend\nall day of this 2-day workshop defining various types of models and what\nthey do.\nThere is also a third \"model\" or source of knowledge in box C, i.e., the\nreal lake, providing a direct avenue to knowledge through what Claire Schelske\nreferred to as descriptive studies. Classical limnology has been largely\nbased on a combination of descriptive study results and conceptual hypotheses\nderived from those results. Therefore classical limnology is represented\nby the box pair A-C, while box pair B-C represents the recent emergence of\nwhat we might call mathematical ecological limnology. There are various\ninteracting arrows between the boxes. The downgoing arrows lead to improvement\nin experimental design, either from conceptual or numerical modeling. The\nupgoing arrow on the left feeds from the data base to the conceptual hypothesis.\nVia the right-ascending arrow, the data base provides verification for the\nmathematical model, most essential to test the model's worth and to improve it.\nThis leads me to the third platitudinous statement: IFYGL has provided\nan unparalleled data base in spite of what I said about instruments this\nmorning, and this in turn gives an unequalled opportunity for verification\nof a variety of models. The surest method of approach is a pedestrian, bipedal,\nprogressive iteration between modeling, improving experimental design, veri-\nfying the models, and so on. We must walk before we run, and we should\nnot promise too much.\n64","Table 8. Five Platitudes of Environmental (Lake System) Modeling\n1.\nIntroductory platitude: Lake system models needed to assess costs of\nnot taking management actions.\n2. Model triarchy: Three general interacting model classes (or approaches)\nare available and shown here in boxes:\nA\nConceptual hypotheses:\ninteractions_\nComputer manipulation of B\n\"What every limnologist\nsystems of equations (sub-\nneed\nknows\"\nmodels), deterministic\nstrengthening\nand/or stochastic\nimprovement of\nexperimental design\nfood for and\nExperimental interrogation\nmodel\nverification of\nof nature, through limited-\nverification\nhypotheses\naperture windows, yielding\n(essential)\nC\nthe \"data base\"\nC\n3. Environmental modeling under IFYGL: The prime purpose was to use the large\ndata base C--already available from CCIW and other sources and later to be\nprovided by IFYGL--to test and perfect numerically predictive schemes, de-\nsigned in B. In particular, the hitherto unsurpassed physical data base was to\nbe combined with biological data to predict production and species competition\nin given hydrodynamic regimes and with given inputs of radiation and nutrients.\n4.\nAnticipated difficulties: It is too early to assess the degree of success\nachieved under 3, but \"classical limnology\" (box A founded on box C) suggests\nthat, while hydrodynamic modeling is approaching a useful operational stage\ngiven adequate routine updating from C, combination of physical models with\nthe present primitive biological models to form operationally predictive\nwhole-system models will encounter severe complexities and will therefore be\ndelayed. Two examples of anticipated complexities are given (Mortimer, C. H.\n(1974), contribution to North Atlantic Treaty Organization Symposium: \"Modeling\nof Marine Systems,\" Elsevier Oceanography Series), one arising from the\nepisodic nature of mechanical forcing and the nonlinear characteristics of\nlocalized shear-flow instabilities, and one arising from the switches to new\nsets of biological species, which commonly occur when a lake system is\nhighly perturbed and which the environmental manager above all needs to predict.\n5. Lines of future progress: In spite of (and because of) the anticipated\ncomplexities, developments in box B continue to be pursued. Rapid progress\nshould not be promised, but the best hope of advance lies in a successive,\ntwo-step, pedestrian B/C iteration (modeling/verification) with hard work in\nboth boxes. Also A/B interactions can be more fruitfully fostered, avoiding\npresent signs of elitism in both camps, with A calling B \"naive\" and B call-\ning A \"numerically illiterate.\" Field interrogation and window design (new\ninstruments, better resolution in time and space) in C also need strong and\ncontinued support, balanced against support of A or B; for if management\nof nature is an objective of modeling, nature is the best source of clues,\nshort cuts, and tests.\n65","Under the fourth platitude, I will talk about anticipated difficulties in\nboth physical and ecosystem modeling. I was surprised to hear Claire Schelske\nsay that \"big science\" had not included ecology. The International Biological\nProgram provides an example. It also provides examples of counter-productive,\nelitist confrontations, treated in the final paragraph of the table (platitude\n5)\nSome classical limnologists have called systems analysts naive, and systems\nanalysts have regarded classical limnologists as numerically illiterate. While\nthere may sometimes be truth in both these accusations, it is more important to\nrecognize and to strengthen the interactions between boxes A and B, as well as\nthe important modeling verification, i.e., B-C interactions already mentioned.\nI now give (in platitude 4) two examples in which A-C or A-B interactions\ncould be productive and indeed essential for progress. The first is a physical\nexample, a conceptual A-type model of wind-driven motions in a small, stratified\nlake (fig. 21), in which wind drag at the surface moves the surface warm layer\nto the downwind end of the basin. A return current forms just above the thermo-\ncline and, if the resultant shear in that layer exceeds a critical value\nassociated with a Richardson number of 1/4, the flow becomes suddenly unstable\nand large vortices form. The return current acts like a carpenter's plane,\neroding the subthermocline layer by entrainment and intensifying the thermo-\ncline gradient at the downwind end of the basin. The \"shavings\" of mixed water\nare carried toward the upwind end by the return current, yielding the observed\nfan-shaped distribution of isotherms. When the wind stops and its stress is\nremoved, the preexisting and the newly formed layers undergo redistribution\naccompanied by a series of oscillations (internal seiches) to new equilibrium\npositions.\nThe important point about this conceptual model--yet to be verified in\ndetail, but obviously describing observed features--is that the final depth\nand shape of the thermocline depends not only on what happened in the water\ncolumn at that station, but more importantly on events, e.g., flow instabilities,\nelsewhere in the basin. But presently available physical and mathematical models\nof thermocline formation and entrainment are one-dimensional and therefore\nof limited use in predicting day-to-day developments in a lake.\nThe overriding importance of the Richardson number and the nonlinear\nnature of the \"explosive\" change from stable to unstable flow when that number\n6\nfalls below 1/4 is illustrated by an IFYGL example from Lake Ontario (fig. 22).\n6\nBoyce, F. M. (1974), Some aspects of Great Lakes physics of importance to\nbiological and chemical processes, Journal of the Fisheries Research Board\nCanada 31, pp. 689-730.\n66","WIND\na\nb\nNEGLIGIBLE FLOW\nC\nWIND FORCE 7-8\n8.03\no\n1\n2\n3\n4\n5\n6\n7\no\n10.6°\n11.4'\n11.7\n11.95\nm\n11°\n11.5\n11.9°\n10\n20\n40\n7.5\n7.44\nWINDERMERE - NORTHERN BASIN\n60\n7.35\nd\n26 OCT. 1949\nFigure 21. Model of wind-driven motions in a small stratified\nlake.\n67","6\n4\n3\n2\n1\no\n30\n25\n30M\n20\n15\n10\n5\n10\n9\n8\n>\nMOOM\n6\n5\n4\n3\n68","Current meters and thermographs were placed at 10-, 15-, 30-, and 50-m depths.\nThe wind stress (aftermath of Hurricane Agnes) was computed. With rising\nwind stress at the surface, the current velocity began to rise first at 10 m\nuntil the velocity difference between 10 and 15 m depths exceeded a critical\nvalue, at which time the temperatures at those two depths were suddenly\nequalized. It is significant that this equalization took place when the mean\nRichardson number between 10 and 15 m had fallen to about 1/4, at which point\nmixing occurred and momentum was then transferred to deeper layers. Subsequent-\nly the same sequence was repeated between 15 and 30 m. This is a beautiful\nexample, the first of its kind, of the downward transfer of momentum, clearly\nshowing the relationship between shear instability and mixing.\n7\nThe second illuminating example is an extract from Lund's many-year study\nof the spring increase in diatom (Asterionella) population in three neighbor-\ning lake basins with similar nutrient input, but considerable differences in\nmaximum depth (Esthwaite Water, 15 m; southern basin, Windermere, 33 m; north-\nern basin, Windermere, 65 m) . In most years, a simple silica-limited model,\nwith uptake rate proportional to basin depth, fits the observations very well.\nThe diatom population shows a log phase of growth that terminates when the\nsilica concentration has fallen to about 0.4 mg/l Si0 Growth starts earliest\n2\nin the shallowest and latest in the deepest basin because the average light\nexposure of a diatom cell is proportional to the ratio of the depth of light\npenetration to the depth of the water column, which is well mixed to the\nbottom in all three basins at that time of year. This is the simple depth-\ncontrolled, nutrient-limited model proposed by Gran in the 1920's to explain\nthe sequence of spring diatom-population peaks in the Norwegian Sea and later\nused by Riley for U.S. coastal waters.\nThere are exceptional years, however, in which the simple model fails, as,\nfor example, in one case (Esthwaite Water during 1949), as a result of fungal\nparasitism of Asterionella, and in the northern basin of Lake Windermere during\n1948 probably because of lack of an essential but unknown micronutrient added\nto the lake during years of normal or above-normal rainfall, but lacking in\nyears of spring drought. When the spring flood eventually came to Lake\nWindermere in 1948, normal diatom growth was resumed but with a 2-month delay.\nThe failure of the simple model in these two cases must be attributed to bio-\nlogical peculiarities which are not uncommon, in one case due to parasitism\nthat is difficult to predict in a deterministic manner and in the other case\ndue to the result of poorly understood mechanisms of cell nutrition.\n7\nLund, J. W. G. (1950), Studies on Asterionella formosa Hass. II.\nNutrient depletion and the spring maximum, Journal of Ecology 38,\npp. 1-35.\n69","These peculiarities can also be instructive for would-be lake system\nmodelers. During 1949, the year of failure of the simple model in Esthwaite\nWater because of parasitism, other diatoms took over and grew in place of\nthe parasitized Asterionella. This illustrates an important property of\nlake systems, well known to limnologists in box A (table 8), but not\nsufficiently appreciated by systems analysts in box B; namely, the fact that\ngross perturbations of the system commonly remove one set of actor organisms\nfrom the stage, replacing them with a different set. Unfortunately, models\nto predict the effects of perturbations are precisely what lake managers\nneed most, but it seems that this is the type of model which will be most\ndifficult to give him.\nResolution of this difficulty and the interfacing of the very different\ntime and space scales of hydrodynamical and biological models should be one\nof the prime post-IFYGL research targets.\n70","1.7.1 Response - G. F. Lee\nI have chosen to use my alloted time on the topic area of Simulation of\nEnvironmental Dynamics of the Great Lakes to focus on those aspects of water\nquality simulation which I feel should receive attention in the immediate\nfuture. No attempt will be made in this presentation to determine what agency\nor agencies should focus on these problems; instead, the problem areas will be\noutlined and briefly discussed.\nFrom an overall point of view, it is important to emphasize that simulation\nor modeling has a definite place in research on and management of water quality\nin the Great Lakes. The Great Lakes, like many other bodies of water, are\nexperiencing water quality problems due to excessive discharge of chemicals.\nChemical problems can, in general, be compartmentalized into three approaches\ndefining the sources, fate, and significance of specific chemical contaminants\nfor a given part of or the whole of the Great Lakes. Each of these compartments\ncan be formulated into relatively simple models which describe the overall\ntransport and transformations of the chemical contaminants. Further, for each\nof\nthe major chemical species, models can be developed that demonstrate how\nthese species interact with various parts of the aquatic ecosystem and, there-\nfore, how a given concentration of a contaminant could affect water quality.\nModels of this type serve as frameworks to compile existing information, thereby\npointing to areas where additional work is needed. Therefore, such models\nshould be developed prior to initiation of any research on the problem.\nFurther, these models are extremely helpful in defining possible manage-\nment alternatives and benefits to be derived from certain types of management\npolicies for chemical contaminants in the Great Lakes. The work sponsored by\nthe EPA as part of the IFYGL studies on nutrient sources, transport, and\ncycling within Lake Ontario is a prime example of how such modeling efforts\ncan be used for water quality management. The overall objective of these\nstudies was to determine what benefits might be derived from the removal of\n80 percent of the phosphorus from domestic waste waters entering Lake Ontario.\nThe sources of phosphorus have been fairly well defined, and studies have been\nconducted that estimate the amounts of available phosphorus entering the lake\nfrom each source.\nA major modeling effort by Thomann and DiToro of Manhattan College is on\nthe relationships between a concentration of phosphorus in Lake Ontario water\nand the biomass that would develop in the open waters of the lake. When fully\ndeveloped, this model will provide a technical basis for estimating the decrease\nin biomass of planktonic algae that might result from an 80- to 90-percent\nremoval of phosphorus from domestic waste-water sources. Discussed below are\nother areas of water quality modeling which I feel should receive attention in\nthe near future.\nAt present, several individuals are developing what might be called\nconservative element models. These models enable researchers to utilize\ncurrent rates of input, future populations projections, and the mixing character\n71","of the lake in making reasonably accurate predictions of concentrations of\nelements such as chloride. Modeling efforts of this type are relatively simple\nbecause they deal with chemical compounds which are nonreactive in the system\nand, therefore, focus on dilution of the materials added to the lake. The\nability to predict an open-lake concentration of a particular chemical species\nis directly dependent on the reliability of input data, the lake mixing\ncharacteristics, and the hydrology of the lake.\nThere is need for additional work in this area in order to better under-\nstand the nearshore mixing processes. For example, it is often said (without\nany technical basis) that at certain times of the year the thermal bar represents\na significant barrier to mixing between nearshore and offshore waters. However,\nwhen one examines the conservative element composition in these waters, both\nfor periods when the thermal bar is present and when it is not, one finds that\nthe concentrations in both water areas are approximately the same. This\nindicates that the overall rates of transport of chemicals and water between\nthe nearshore and offshore areas are independent of the presence of the thermal\nbar.\nThe open-lake eutrophication modeling efforts, being conducted as part of\nthe EPA Chemistry and Biology Panel activities for the IFYGL studies on Lake\nOntario, are progressing well. In my opinion, the models being developed by\nthe Manhattan College group appear to be of sufficient technical validity to\nwarrant further major efforts along these lines for the other Great Lakes. It\nshould be noted at this time that these eutrophication models are for the open\nwaters of the lake and do not consider nearshore processes. Also, these models\ndo not presently relate total phosphorus flux to the lake's response. They are\nbased on a concentration of available phosphorus in the open lake water. In\norder to determine the relationship between amounts of phosphorus entering the\nlake from both its tributaries and direct wastewater inputs and the amounts of\nphosphorus that will eventually become available in the lake, chemical modeling\nshould proceed simultaneously with eutrophication modeling.\nOf all Great Lakes water quality modeling efforts, probably the most needed\ntoday is development of a nearshore eutrophication model. Such a model would\ndemonstrate the relationship between nutrient input from tributaries or direct\nwaste inputs and the growth of attached algae such as Cladophora. Essentially,\nno significant progress has been made in this area. Yet, this is one of the\nmost significant water quality problems in the lower Great Lakes. At present,\nit is impossible to predict with any degree of reliability what environmental\nimpact reducing the phosphorus input to the nearshore waters of a given region\nwill have on Cladophora. One of the problems that makes modeling of this type\nespecially difficult is that a key aspect governing the growth of these algae is\nthe hydrodynamics of the interface between the organisms' holdfast (i.e.,\nsubstrate) and the overlying waters. Growth of Cladophora is not only dependent\n72","on the concentrations of nutrients in the water, but also on the rate of trans-\nport of these nutrients to the algae. From an overall point of view, I feel\nthat the next major IFYGL study effort on the Great Lakes should bring the\nhydrodynamist, biologist, and chemist together to work on nearshore water-\nquality models emphasizing the Cladophora problem.\nReasonable progress is being made today on oil spill modeling in order\nto predict the rate of transport of oil and gasoline. Further, some efforts\nare being made in modeling of dredged-material disposal practices. Generally,\nhowever, modeling efforts in this area are hindered by lack of information on\nthe short-term, high-concentration toxicity of chemical species to aquatic\norganisms. The acute toxicity data that are available today are generally based\non a 96-hour exposure period. For dredged material disposal, the excessive\nconcentrations rarely persist for 96 hours. Instead, after a few hours, higher\nconcentrations are rapidly diluted below the acute lethal level. Within a\nrelatively short time contaminant concentrations fall below chronic sublethal\nlevels as well. For example, it is known that dredged material disposal in\nopen waters results in a release of ammonia to the water column. In many\ninstances, the concentrations of ammonia will be above the 96-hour LC50\nHowever, because of the intermittent nature of the dumping practice, the\nrelatively high concentrations of ammonia are usually diluted within a few\nhours to below acute toxicity levels and within a day or so to background levels.\nIt is impossible at this time to establish criteria for such a situation\nsince we do not have short-term ammonia toxicity data for various forms of\naquatic life. Data are needed on the relationship between the acute lethal\nconcentrations at various periods of time that match to some extent the normal\nrates of dispersion that occur from a point source. This same type of data\nis needed for industrial and municipal outfalls into the lake.\nThere is a need for models that can simulate (and thus offer some potential\nfor predicting) the environmental impact of chronic sublethal effects of chemical\ncontaminants on Great Lakes waters. The 1972 amendments to the Federal Water\nPollution Control Act require that by the mid-1980's, industries, and quite\npossibly municipalities, demonstrate why they should not achieve a zero\npolluting discharge from their installations. They will have to consider\ntechnical feasibility, economics, social desirability, and environmental impact.\nThe general problem that exists today and will certainly prevail in the future\nis not one of acute lethal toxicity. Instead, it is one of chronic toxicity,\nimpairment of the rate of growth, rates of reproduction, or other vital\nfunctions of aquatic organisms. Even so, there will not be a complete blockage\nof reproduction but probably some impairment, i.e., a 10- or 20-percent reduction.\nIt is highly likely that environmental quality litigation will raise\nquestions about the significance of even a 10-percent reduction of reproductive\npotential of a certain form of fish due to the presence of an apparently\nexcessive amount of certain chemical. Largely as a result of the current\n73","relatively poor understanding of ecosystem functioning, at this time we have\nessentially no ability to answer this type of question. If decisions are to\nbe technically sound, it is mandatory that efforts be made to design ecosystem\nmodels to determine the significance of a certain size waste-mixing zone where\nthere is an impairment of fish reproduction in the fisheries of the lake as\na whole.\nThese results would be applicable not only tc chemical inputs but also to\nheated-effluent discharges from electric generating stations utilizing\nonce-through cooling. Each of the Great Lakes has a large heat-assimilative\ncapacity whereby waste heat could be added to the lake without significant\nimpairment of overall water quality. There is no doubt, however, that\nit would be possible to add sufficient heat to the lake to affect adversely the\noverall lake ecosystem. At this time, we cannot predict with any degree of\nreliability what the ultimate heat-assimilative capacity is for any part of the\nGreat Lakes. Therefore, modeling efforts should be initiated in an attempt to\ndetermine the impact of impaired water quality in one region on the overall\necosystem.\nCurrently some progress is being made on hydrodynamic models of the dis-\npersion from a point source, such as a wastewater outfall. Probably the\ngreatest success of modeling efforts in this area is in connection with pre-\ndicting the size of thermal plumes from electric generating stations. This can\nbe done today with a reasonable degree of reliability. However, virtually no\ninformation is available on modeling of the chemical concentrations and toxicity\nof contaminants in the nearshore zone. There is a need for modeling directed\ntoward examining the physics, chemistry, and biology of mixing zones for waste-\nwater input to the Great Lakes. From an overall point of view, such modeling\nof chemical contaminants is hindered by lack of information on the environmental\nchemistry, physical transport, toxicity, and/or stimulatory capacity of specific\nchemical species in Great Lakes waters.\nThere are several additional study areas. Related to the questions of\nmodeling of the water quality of the Great Lakes is the development of monitor-\ning programs designed to evaluate changes in water quality as a result of man's\nactivities. At the present time, no one has determined the number and location\nof monitoring sites in the Great Lakes in an attempt to detect significant\nchanges in water quality. To do this in a meaningful way, at least crude models\nof the expected response of the lake to various inputs must be available.\nI feel NOAA or some other agency should work on ascertaining the signifi-\ncance of marshes and wetlands to the ecology of the Lakes. Some parts of the\nGreat Lakes have considerable areas which interface with wetlands or marshes.\nIn marine systems, marshes are known to be the primary source of nutrients and\nfoods for larval forms of aquatic organisms. Yet little work has been done\non the significance of wetlands to the Great Lakes aquatic ecosystems.\n74","Another area that should be considered for possible NOAA activity is\ndevelopment of a data storage and retrieval system on Great Lakes water quality.\nThe existing systems simply are either nonfunctional or unreliable. These are\nof little or no value in establishing water quality of the Great Lakes. There\nis also need for periodic critical examination of the data to ascertain whether\nthere has been any change in the water quality of the various Lakes. In\naddition, someone with a high degree of technical competence should review all\ndata going into the storage system in order to ensure their reliability.\nFurther, some permanent record should be kept of the analytical methods that\nwere used to generate the data. Then sometime in the future, someone from\neither within or without the agency can examine the historic data and determine\nwhether or not there had been real changes and whether apparent changes can be\nascribed to changes in analytical procedures and/or sampling techniques.\nAnother aspect of Great Lakes water-quality problem studies which I feel\nneeds considerable attention is the diffusion of phosphorus sources for the Lakes.\nRecently completed studies by W. F. Cowen and myself have shown that only a\nsmall part of the total phosphorus present in the organic and particulate\nforms entering Lake Ontario from tributary sources will likely become available\nin the lake. This means that efforts to control urban and rural storm water\ndrainage in many parts of the Great Lakes basin would result in little or no\nimprovement in water quality because the majority of the phosphorus derived\nfrom these sources is in an unavailable form. The studies of the type conducted\nby Cowen and Lee on the Lake Ontario Basin should be expanded to all of the\nGreat Lakes and include not only tributary but also atmospheric sources of\nphosphorus.\nThere is a great need for a comprehensive look at dredged material dis-\nposal criteria in order to determine what is the actual environmental impact\nof offshore disposal of contaminated dredged sediments on the Great Lakes.\nThese criteria are of great economic significance to the Great Lakes. The\ncurrent ban on open water disposal of dredged materials in the Great Lakes\nwithin the next few years will cost an estimated $230 million for dike disposal\nareas. There are serious questions about whether this expenditure is justified\nfrom an economic, or more importantly, ecological point of view. It is highly\nprobable that dike disposal systems currently being developed may do more harm\nto the aquatic ecosystems in the Great Lakes than have open-water disposal\nsystems used in the past. Efforts in these areas should be closely coordinated\nwith the U.S. Army Corps of Engineers Dredged Material Research Program.\n75","1.7.2 Response - J. Spain\nI will expand on Dr. Mortimer's \"model triarchy,\" perhaps adding a little\nresolution to it. I will also consider what a model is and what it can do\nand consider some basic questions about modeling and its relationship to\nresearch from my point of view. I have also detected what Dr. Mortimer called\n\"an elitism\" in that some people tend to scoff at the idea of modeling and\nperhaps think of it as a separate endeavor. I would rather think of it as an\nintegral part of the research process.\nRelating models to research (fig. 23), what we are interested in is, of\ncourse, the real system and what makes it tick. From the real system, we derive\nreal data; from the real data, we derive a descriptive or conceptual model.\nThe classical feedback loop has pretty much dealt with these components: real\nsystem, real data, conceptual model; or it has gotten to the level of a mathe-\nmatical model about which, after it has been put on paper, people say, \"Yes,\nthis is fine, seems to look right, etc.\" Again, going back through this loop\nthat has been research up until recent years, we are now capable of testing\nmathematical models in a very impersonal fashion. The value of the computer\nis that it provides an impersonal evaluation of your model. If it is no good,\nthe computer is going to tell you by providing simulation data. Now, you have\nsomething to compare with the real data. It is usually not going to match up\nwith real data the first time. We are going to crawl before we walk. So what\ndo you do then? You modify conceptual models, mathematical models, and simula-\ntion. Predictive ability is sort of a spinoff. I think too many people are\nthinking in terms of using simulation as a predictive and decision-making tool.\nMaybe we will get this kind of spinoff in the future, but the real value of\nsimulation is that of providing tests for the conceptual models that have been\ndeveloped over the years. The computer can also keep track of multiple, com-\nplex-coupled nonlinear interactions which are characteristic of ecosystems\nand which confuse the human mind.\nWhat is the role of simulation? It is part of the research feedback loop\n(fig. 23) and everybody should be involved in it. It is wrong to have one guy\ngoing out and collecting the data and another guy doing the simulation. Ideally,\neach researcher should be involved in both parts of this process one way or\nanother because it provides an impersonal evaluation of scientific concepts.\nIf the simulation does not match up with the real system, you should say\n\"hurray.\" That is the best thing that could happen. Do not be downhearted or\nsay, \"Let's throw it out. If it works completely and matches up with the real\nsystem, your job is done and you have not learned anything. You have the old\nconcepts; but you had them before. If it does not work, then you know some-\nthing. The old concepts were not totally correct. Now you have to find out\nwhy it is the simulation data do not match up with the real data. For example,\nthere must be something lacking in our understanding of the whole process of\ncirculation of lakes. If the simulation does not match up, it means there is\nsomething wrong with the conceptual model. A concept may be completely lacking\n76","REAL\nREAL\nSYSTEM\nDATA\nDESCRIPTIVE\nCOMPARISON\n(CONCEPTUAL)\nEVALUATION\nMODEL\nSIMULATION\nDATA\nMATHEMATICAL\nSIMULATION\nMODEL(S)\nPREDICTION\nDECISION-MAKING\nMANAGEMENT\nFigure 23. Research feedback loop.\nor there may be some kind of nonlinear interaction taking place that we are\nnot aware of . On the other hand, maybe the model is no good. That is a\npossibility, but, assuming you have checked the model mathematically, it is\ndoing what you think it is doing and the real data are correct, then if the\nsimulation does not work, there is something wrong with the conceptual model!\nPeople should not be very concerned that we are starting out with very\ncrude models. Eventually we will add to them. Look what happened with the\nmyoglobin molecule. First they worked out a model with 10-angstrom resolution.\nAll they got was some big hunky thing. All they knew previously was that there\nwas some ordering of the molecule. Gradually, as the resolution was improved,\nwe were able to see a sharper and sharper picture of the chemical molecule.\nHere we are working with a mathematical model. You have to start with a\ncrude, low resolution thing and work toward the high resolution models of the\nfuture. You must crawl before you walk and walk before you run.\n77","1.7.3 Response - B. Eadie\nI want to emphasize some of the points that I feel are really important.\nPrimary is the one brought up by Prof. Mortimer this morning that biological\nmodels and chemical-biological interaction processes are poorly understood in\ncomparison with physical processes. The reason is that no strong theoretical\nfoundation for biology and nonequilibrium chemistry exists, especially in\naquatic systems.\nProf. Lee's analysis of our weakest points in nearshore chemistry is\nalso well taken. Effects of river plumes and water mass entrapment on\nbiological systems are incompletely understood. The sublethel effects of toxic\nmaterials are something the oceanographers have begun to look at, and I think\nwe can take a key from the beginnings of their research. They are beginning\nto look at sublethel effects of petroleum hydrocarbons and pesticides on some\nfish and smaller organisms in the ocean.\nLooking back at Prof. Mortimer's diagram, what we need to do is in box\nC, Experimental Interrogation of Nature. That is where, in the biological\nand chemical areas, we have to expend our greatest effort if we are going to\nbetter understand the system. We do not understand the formalizations or the\nfunctional relationships which exist between biological uptake and some chemical\nspecies. What mechanisms are triggered by certain perturbations in the\nbiological system? What causes a bloom to occur? What we will have to look at\nis something on a higher frequency scale and a smaller spatial scale than was\nattempted in IFYGL.\nAs Dr. Aubert mentioned this morning, the natural distribution and\nvariations in the chemical parameters were not well understood in IFYGL,\nalthough analysis has just begun, primarily because they were not collected on\na fine enough grid and time scale to get good relationships. What we need to\ndo in the study of environmental dynamics is to examine the system at a higher\nfrequency in a spatial scale we can handle.\n78","1.8 DISCUSSION\nAubert. I want to comment on the interpretation of Environmental Dynamics as a\ntitle used for the plenary and work group sessions. I conceive of a model\nhierarchy with environmental dynamics being the most inclusive level. Environ-\nmental dynamics involves the interactions of all relevant processes. If a model\ndiagram were drawn, environmental dynamics would be at the top. Four other\nitems are roughly equal but one level below environmental dynamics: water\nmovements, aquatic ecology and water quality, lake-atmosphere interactions,\nand water levels and flows. Somebody else would probably set up a different\nhierarchy. A hydrologist says hydrology includes limnology; a limnologist says\nlimnology includes hydrology. It depends on the viewpoint as to how a structure\nis set up, but that was my rationale in preparing the outline. No holds are\nbarred in the area of environmental dynamics, nor should there be restrictions\nin the other areas. In some respects, aquatic ecology may be almost as broad\nas environmental dynamics, but I think aquatic ecology emphasizes biological\nand chemical aspects more than physical aspects.\nMortimer. Will there be an opportunity for the groups to get together aside\nfrom the group sessions?\nAubert. In the workshop sessions.\nMortimer. Could two work groups merge if they felt so inclined?\nAubert. Yes. Again, in concept, this was structured to maximize output and\nalso to reduce groups to manageable size. While people have been assigned to\nwork groups, it is not meant that everybody must remain in that one work group.\nSome amount of floating between work group sessions might make some sense\nin that there is clearly overlap between the groups; however, I cannot suggest\nhow one might float in order to participate in all relevant discussions. Like-\nwise, the chairmen may feel that for part of the workshop session it might be\ndesirable to schedule a joint work group session. That decision\nwill be left to the chairmen.\n79","1.9 LAKE-ATMOSPHERE BOUNDARY LAYER PROCESSES OF LARGE LAKES - M. Estoque\nFor the purpose of my presentation, I will assume that the problem for\nthis group is the prediction of boundary layer processes over large lakes.\nThe term boundary layer processes is understood to refer generally to the\nturbulent fluxes of momentum, heat, and moisture. You might ask whether or\nnot there is, indeed, a problem. We know, of course, the approximate behavior\nof the boundary layer over lakes; there is a problem only if one wants to pre-\ndict the magnitudes of the processes more accurately than we can at present.\nThe accuracy of this prediction at present is not high, in general within a\nfactor of five, but somewhat better in cases when the thermal stratification of\nthe atmosphere is near neutral. Why are boundary layer processes over lakes\nimportant? They are important because they are the mechanisms which generate\nthe surface water currents and transfer heat and moisture between the lake and\nthe overlying air.\nThe behavior of the boundary layer over large lakes depends primarily on\nthe prevailing synoptic conditions and the lake surface temperature. Synoptic\nconditions over the Great Lakes change due to the passage of cyclones and anti-\ncyclones. The change is more or less regular, with a periodicity of about 1\nweek. On the other hand, the lake surface temperature changes much more slowly.\nThe changing synoptic conditions, in conjunction with the lake temperature\ndistribution, produce corresponding thermal stability changes in the boundary\nlayer. One can classify the thermal stability conditions into three categories:\nunstable, neutral, and stable. The unstable condition generally occurs during\nthe winter season when the lake surface temperature is warmer than the surface\nair associated with the large-scale prevailing flow. The neutral condition\noccurs when the surface air has the same temperature as the lake surface. The\nstable condition occurs in late spring and early summer when the lake surface\nis colder than the surface air. In general, the intensity of the boundary layer\nprocesses over the lake is largest under unstable conditions and least under\nstable conditions.\nIt might be of interest to give an indication of how much the large-scale\nsynoptic condition can be modified by Lake Ontario. This is done with the aid\nof a numerical simulation of a thermally unstable case. The synoptic condition\nwhich is simulated is that which occurs during the period immediately following\nthe passage of a cyclone slightly north of the lake; the period is, therefore,\ncharacterized by a veering of the wind from westerly to northwesterly over the\nlake. In order to simplify the numerical integrations, Lake Ontario was\nreplaced with a rectangular lake of about the same size. The lake surface\ntemperature is assumed to be uniform and 20°C warmer than the large-scale\nprevailing surface air. The simulated distributions of air temperature,\npressure, and velocity (fig. 24-27) correspond to distributions after the\nsynoptic wind has veered from westerly to northerly. Figure 24 shows the\nair temperature distribution near the Earth's surface. A warm pool of air is\n80","288\n280\n288\n296\nFigure 24. Temperature distribution in the surface boundary layer.\n1005\n1009\n1013\nFigure 25. Surface pressure distribution.\n81","Figure 26. Surface wind distribution.\n-10\no\n10\n-5\n20\no\n-5\no\nFigure 27. Vertical velocity (cm sec-1, distribution at a height of 1 km.\n82","generated; the center is south of the lake center, close to the southern coast-\nline. This feature is a result of the warming of the air as it moves southward\nacross the warm lake. This warm pool is reflected as a low pressure area at the\nsurface as shown in the surface pressure distribution (fig. 25). Figure 26\nshows the surface wind distribution. Notice the strong winds which have been\ngenerated over the lake. Notice also the strong horizontal convergence along\nthe southern shoreline. Associated with this convergence line is a region of\nupward motions. This is shown in figure 27, which shows the vertical motion\nfield at about 1 km above the Earth's surface. On the basis of these diagrams,\none concludes that the lake could strongly modify the prevailing synoptic flow\npattern during thermally unstable conditions. It is, therefore, impossible to\ndetermine accurately the boundary layer processes over the lake under these\nconditions by considering only the undisturbed synoptic-flow pattern. One has\nto take into account the fact that the lake can strongly distort the synoptic\nflow pattern, thereby producing a mesoscale disturbance whose boundary layer\nis different from that inferred from the undisturbed synoptic condition.\nThe problem can be summarized with the aid of figure 28. We envision\nthe problem to be prediction of the boundary layer processes from the specified\nundisturbed synoptic-flow, the lake surface conditions, and the surrounding\nland-surface conditions. The crudest solution would be to consider only the\nundisturbed synoptic flow without taking into account the mesoscale distortions\ninduced by the lake and to use empirical-physical techniques. In figure 28,\nthis method of solution can be indicated schematically by arrows which proceed\nfrom the given boxes (synoptic flow, lake surface conditions, land surface\nconditions) to the predicted box (boundary layer processes) through lines\n1, 2, 3, 4, and 11. In this case the feedback loop is not considered; i.e\nlines 9 and 10 are disconnected from the lake and land surface conditions.\nThe ideal solution should consider the feedback loop. As indicated in the\npreceding paragraph, the surface fluxes produce a modification of the atmosphere\nthat results in a mesoscale disturbance. The associated boundary layer of\nthe mesoscale flow may, in turn, produce changes in the surface lake and land\nconditions, thus altering further the original boundary-layer processes. This\ncomplicated chain of events can be taken into account only by incorporating the\nfeedback loop. This implies that lines 5, 6, 7, 8, 9, and 10 (fig. 28) should\nbe taken into account.\nIt is appropriate to assess the current state of knowledge of the physical\nprocesses which are required for the prediction of a boundary layer. In figure\n28, these processes are those which are involved in empirical-physical models\n(line 4) and the mesoscale-physical models (line 5). Let us consider first\nthe current state of knowledge concerning the empirical-physical determination\nof turbulent fluxes. A common method for doing this is the so-called bulk\naerodynamic method which requires the use of drag coefficients. The magnitudes\nof these coefficients are not accurately known. And one of the important\n83","6\n7\n8\nCONDITIONS\nSURFACE\nLAND\n9\n11\nFEEDBACK\nLOOP\nFigure 28. Schematic flow diagram of the boundary-layer prediction problem.\n10\n3\nProfiles, Exchange Coefficients)\nEMPIRICAL-PHYSICAL MODELS\nBOUNDARY LAYER PROCESSES\n(Surface Fluxes, Vertical\nand Precipitation)\n(Including Clouds\nPHYSICAL MODELS\nMESOSCALE FLOW\nCONDITIONS\nSURFACE\nLAKE\nV\n5\n2\n4\nSYNOPTIC FLOW\nEXTERNAL\n(Including\nRadiation)\nFORCING","studies under IFYGL is to determine the values of these coefficients under\nvarious meteorological conditions over Lake Ontario. One can get an indication\nof the current state of knowledge of these coefficients by examining figure 29.\nThis diagram indicates the accuracy of determining the turbulent moisture flux\nby using a drag coefficient. An approximate value of the coefficient would be\n1.23 X 10 -3\nHowever, such a value would seriously underestimate the turbulent\n.\nflux at large values of UAq. The corresponding accuracy of determining the\nturbulent heat flux is indicated in figure 30. Again, large errors in deter-\nmining the heat flux with the aid of a constant value of the drag coefficient\nare expected. In addition to empirical relationships between turbulent fluxes\nand the mean flow, one can also establish empirical relationships between other\nboundary layer quantities. An example of such a relationship is between the\nRichardson number and the bulk Richardson number (fig. 31). The Richardson\nnumber is often used as a parameter for determining empirically turbulent\nfluxes. In concluding the discussion of empirical-physical relationships\nbetween turbulent fluxes and the mean flow, one can say that current relation-\nships are rather reasonable. However, in order to predict the boundary layer\nprocesses more accurately, one should formulate more accurate relationships.\nThe current relationships are erroneous under highly unstable thermal stratifi-\ncation and strong winds. More research must be done in order to formulate\nsatisfactory relationships under these extreme conditions. Hopefully, investi-\ngations under IFYGL might provide improved relationships.\nWe discuss next the current state of knowledge concerning the physical\nmodeling of mesoscale flow (line 5 of fig. 28). This is normally done numeri-\ncally with the so-called primitive equations. What are the current weaknesses\nin physical models of mesoscale flow? The most serious weakness is the descrip-\ntion of the effects of subgrid-scale eddies in terms of grid-scale quantities.\nFor practical purposes, the minimum grid distance which can be used for\nnumerical integrations of mesoscale model equations is probably on the order\nof 10 km. Therefore, the effects of eddies smaller than 10 km in scale should\nbe understood. Another weakness in the modeling of mesoscale flow is the\nincorporation of terrain effects, variations in elevation, and roughness. The\ndescription of the latter is especially difficult because it involves the effects\nof trees, buildings, and similar inhomogeneities of the Earth's surface. Finally,\nthere are weaknesses related to the formulation of lateral boundary conditions\nand initial conditions.\nTo summarize the important points in predicting the boundary layer over\nlakes, we need first to specify the synoptic-scale flow. The synoptic-scale\nflow is predicted on an operational basis by the National Weather Service. The\naccuracy of the prediction is reasonably accurate. We can, therefore, assume\nthat the specification of the synoptic flow is not an important obstacle in the\nboundary-layer prediction problem. The interaction between the synoptic flow\nand the lake is also an important factor which should be taken into account\n85","32\n28\n24\n20\n16\nwq - = 1.23x10-3 UA9\n12\n8\nCCIW NIAGARA 1971\n4\nBOMEX (POND) 1971\n0\n16\n20\n0\n4\n8\n12\n10-3 (pg/cm2S)\nFigure 29. Relationships between evaporation\nand large-scale parameters (from Elder*).\n*\nElder, , F. C. (1973) , Some results of direct measurement of Bowen ratio over\nan open lake surface, presented at the 16th Conference on Great Lakes Research,\nInternational Association for Great Lakes Research, Sandusky, Ohio.\n86","2.8\nw'T = 1.23 x 10-3 UAT\n2.4\n1.6\n.8\n0\n-.8\nCCIW NIAGARA 1971\nBOMEX (POND et Al 1971)\n-1.6\n-1.6\n0\n.8\n-.8\n1.6\n2.4\n10-3 UAT°C cm/s)\nFigure 30. Relationships between sensible heat flux and\nlarge-scale parameters (from Elder, 1973 ).\n*\nElder, F. C. (1973), Some results of direct measurement of Bowen ratio over\nan open lake surface, presented at the 16th Conference on Great Lakes Research,\nInternational Association for Great Lakes Research, Sandusky, Ohio.\n87",".2\naT/az\nRi\no\nRb= 5g (T5-Tw)\n-.2\n-.4\n-.6\n-.8\nUnstable\nStable\n-.08\n-.04\n0\n.04\nRb\nFigure 31. Relationship between the Richardson number and\nthe bulk Richardson number (from Donelan, 1974*).\nfor the lake distorts the synoptic flow, and the distortion produces a boundary\nlayer over the lake which could be very different from that deduced purely from\nthe undistorted synoptic-scale flow. Finally, the deficiencies which have to\nbe overcome in order to achieve an accurate prediction of the boundary layer\nflow are as follows:\n(1) The specification of turbulent fluxes and other boundary processes in\nterms of the mean flow, especially under highly unstable (thermally)\nand strong wind conditions.\n(2) The specification of subgrid-scale mixing processes.\n(3) The incorporation of varying terrain elevation and roughness.\n(4) The formulation of lateral boundary and initial conditions.\nIn addition to the above items, there are certain observational deficiencies\nwhich have to be remedied. Observational data are needed in formulating\nempirical flux relationships as well as in testing models. IFYGL may provide\nadequate observational data over Lake Ontario; however, over the surrounding\nland areas, we may not have adequate data.\n*\nDonelan, M. (1974), Over water atmospheric boundary layer profiles under\nvarious conditions of wind speed and stability, presented at the International\nAssociation of Meteorology and Atmospheric Physics First Special Assembly,\nMelbourne, Australia, January 14-25, 1974.\n88","1.9.1 Response - D. D. Hougton\nI will expand considerations for the lake-atmospheric boundary layer\nprocesses beyond that discussed by Dr. Estoque. Dr. Aubert referred to enlarg-\ning the concept of environmental dynamics to include the total environment. I\nwould propose the same be done here for the boundary layer. Firstly, we need to\nconsider the planetary boundary layer over entire watershed areas, such as was\nshown on the map for the Lake Ontario study, instead of just over the lake.\nSecondly, attention needs to be given to a layer much deeper than the surface\nboundary layer. Dr. Estoque alluded to this deeper layer, but referred\nexplicitly to the 10-m surface layer only when discussing fluxes. With these\nenlargements, the important man-lake interactions can be studied, particularly\nwith respect to air pollution. Man's environment must include the air we\nbreath as well as the water we drink.\nAn important point that Dr. Estoque made was that the boundary layer-lake\ninteraction was not all one way. Many oceanographers and limnologists deal\nwith the atmosphere as the forcing function for the water. But, for the\natmospheric boundary layer, it can also be the other way around. Atmospheric\ntemperature gradients and changing static stabilities due to water temperature\ncan be important factors in determining the boundary layer mixing process.\nIf we consider air pollution, the lakes may act as an active sink.\nThe important thing to note is that we are talking about a turbulence\nphenomenon for nonhomogeneous situations. If the atmosphere near the lake\nedge is examined, local circulations and other inhomogeneities are common,\ncausing important deviations from homogeneous and isotropic turbulent mixing.\nPerhaps the adjustments in the lowest 10 m are relatively rapid and non-\nhomogeneous conditions can be handled locally. For the layer from 10 m to\n2 km, it is not so clear how the boundary layer responses can be handled.\nTherefore, I would encourage further effort to get turbulence and flux data\nfor the layer from 10 m to 2 km so that better studies can be made. Experience\ngained in the recently completed Global Atmospheric Research Program Atlantic\nTropical Experiment suggests that a tethered balloon system might be sufficient\nto provide actual heat and moisture flux determinations under various synoptic\nconditions.\nDr. Estoque indicated that we already know vertical fluxes to within one\norder of magnitude, based on simple turbulence models, and that this matter\nis under control. I feel that one order of magnitude is not good enough;\nparticularly for cases of extreme instability, a more accurate determination\nof magnitude is needed.\nAs mentioned earlier, the Lakes are a sink of atmospheric air pollution.\nThis makes it important to determine the rate at which the atmosphere is bringing\nthe pollution to the surface of the water. This needs to be explored with\nemphasis on the near lake-edge areas. Sometimes the lake-edge circulation\n89","becomes closed so that pollution advected away from a city may be recycled back\nto the area, leading to enhanced pollution danger.\nFinally, with reference to scales, Great Lakes boundary-layer effects\ninvolve mesoscale phenomena with a depth on the order of several kilometers.\nAdd to this a significant diurnal variability, and we are in a scale range\nwhere the dynamics and interactions are poorly understood. We need a much\nbetter data base for proper study of these processes. I endorse Ted Green's\nconcept of having a local concentrated observing setup nested within a broader\nscale network. This will provide needed mesoscale data.\n90","1.9.2 Response - J. Holland\nI will enumerate a few things that I think, as a result of our experience\nin IFYGL, still need to be done on the Great Lakes. IFYGL did not properly\ncover the stable season in the spring when the lake is cooler than the air.\nThe instruments were not installed early enough, they were not working well\nenough, and the vertical resolution and vertical coverage were not adequate; i.e.,\nthe airplanes did not fly low enough, the towers were not high enough,\nand we had no tethered balloons so the interval from about 10 to 100 m above\nthe lake where a big transition occurs between lake surface conditions and the\nfree atmosphere was not observed. We had some evidence of negative fluxes of\nmoisture and heat, but this is an important case theoretically and dynamically\nand from the standpoint of pollution (although it contributes little to the\nlake-air transport of heat and water vapor because they are very near zero\nduring that season) because this is one that needs more work. It should not\nbe done soon. We need adequate platforms and instrumentation for tackling this\nproblem before a new experiment is done.\nWe had a peculiar year, as every year is peculiar. One of the peculiar\nthings about the year of IFYGL was Hurricane Agnes. Because the June-July\nperiod was so severely perturbed by that, there is some question as to how\nmuch we can generalize any of the results obtained in IFYGL during that time\nof the year to other years; obviously, we need data in other years.\nWe certainly should get much farther along in understanding what we learned\nfrom IFYGL before we finalize the design of a program to check the applic-\nability of IFYGL results under other large-scale or seasonal conditions.\nSimilarly, Lake Ontario is a peculiar lake, but we called this the field year\nfor the Great Lakes. It was the field year for Lake Ontario and for generali-\nzation of IFYGL results to the other lakes; it will again be necessary to\ntest these results on the other lakes, although maybe not with a project the\nmagnitude of IFYGL. But certainly when the IFYGL data have been analysed to\nthe point where we can say what we learned from IFYGL, we should anticipate\nwhat will be found in the other lakes based on generalization of the IFYGL\nresults. These should then be tested by suitable observations on other lakes,\nmaybe not all the other lakes, but certainly lakes which are different in\nimportant respects. Lake Michigan, for example, is elongated in the meridional\ndirection instead of the zonal direction; therefore gradients along the axis\nof that lake perhaps cannot be as readily neglected as they can be in Lake\nOntario.\nWe need information on the nearshore atmospheric gradients. Nearshore\nlimnological gradients have been a matter of very great interest and have led to\nimportant discoveries in IFYGL, but we had poor coverage on the atmosphere in\nthe nearshore region and the shore region. We had good coverage on homogeneous\ninstrumentation nicely exposed all over the lake. On the shoreline, we had few\nstations. There was no attempt to standardize the exposure of the instruments.\n91","Each one had some peculiar local effects. There were not enough to establish\nthe boundary conditions of the lake and the interplay of the meteorology between\nthe lake and the land. What happens in the nearshore region in the atmosphere\nis still a mystery as we try to analyze the IFYGL meteorological data. This is\na subject that needs more work, and again it needs thought and suitable instru-\nmentation and platforms before extensive field work is done.\nThe Center for Experiment Design and Data Analysis in the Environmental\nData Service of NOAA will be working with IFYGL data to get mean values and\nconfidence limits on some of the exchange coefficients and to determine whether\nsome of the nonlinearities of the coefficients can be found. The evaporation\ngraph that Mariano Estoque showed suggests the possibility that high evaporation\nrates must have a bigger coefficient. Also, one of the previous speakers\n-3\nshowed a graph in which a drag coefficient of 3 X 10 was used, and limnologists\ntend to use numbers like this. Meteorologists use numbers more like 1 X 10 ,\nand we do not know the effects of the intermittency of these phenomena on the\nmean values of these coefficients. One of the things we learned in IFYGL was\nabout the intermittency of the fluxes. We knew that most of the energy flux\nfrom the lake to the air occurs in the fall season and that most of that occurs\nin a few episodes of a few days each, and we found in IFYGL that, within those\nfew days, most of it occurs in a few hours. We also found that it occurs in\na small percentage of the lake area. When a big cold outbreak occurs over the\nlake, it turns the lake over and pushes all the warm water over to the downwind\nend of the lake. Essentially all of this flux is taking place in a short time\nin a small space in the lake. What this means in terms of exchange coefficients\nin Estoque's model is that, in addition to the mesoscale perturbation of the\natmosphere, the perturbation of the lake also is going to effect these exchanges.\nWe may or may not be able to learn from IFYGL what we need to know about these\nnonlinearities, these intermittencies, in order to evaluate the feasibility of\nmodeling means over months, seasons, or years, or whether these have to be built\nup statistically from the probability distribution of the widely differing\nstates that occur.\nAnother thing that will interfere with mesoscale modeling is the possible\nimportance of small mesoscale or large microscale, that is, the kilometer-\nscale structure that may occur. We know from satellite pictures that lines,\nclouds, and streaks tend to occur and that these things have varying widths and\nintensities, so this is going to make some difference in the exchange coefficients\nthat are used. There is apparently structure on all scales, and IFYGL was not\nable to cope with the observational requirements of this kilometer scale. It\nwas very good for lake scale and maybe down to one-third lake-scale, but we\ndid not instrument to 5-percent lake-scale in IFYGL. There is systematic\nbehavior on this scale, which is essentially the scale of the nearshore transition\nzone.\nThese are areas for further work which I propose would need a high priority\nas a result of deficiencies in the IFYGL program.\n92","1.10 SIMULATION OF GREAT LAKES WATER LEVELS AND FLOWS IN CONNECTING\nCHANNELS - D. D. Meredith\nThe Great Lakes are the earth's greatest expanse of fresh water. Due to\nthe extensive demographic and industrial development of the Great Lakes region,\nthe hydrologic conditions of the lakes influence the economy and growth for a\nmajor region of both the United States and Canada. The water levels of the Lakes\nand the flows in the connecting channels influence many of the activities on\nand around the Lakes. Commercial navigation requires certain minimum lake\nlevels and flows in the connecting channels to provide the necessary minimum\ndraft. Hydroelectric power generation requires minimum flows in the channels\nto maintain power capacity. Shoreline property may sustain inundation and\nerosion damage from direct flooding during high water periods. In order to\nachieve the maximum benefit from the Great Lakes, they must be managed in the\nmost efficient way to achieve the objectives of those who enjoy their use.\nTo manage the Great Lakes water levels and flows in connecting channels,\nthe inflows to the Lakes or the outflows from the Lakes or both inflows and\noutflows must be controlled. In order to control the lake levels and flows in\nconnecting channels, we must have a regulation policy which incorporates the\nbasin's hydrology, regulatory works, and political and management issues.\nThe purpose here is to present a brief review of the basin hydrology and\nthe procedures used to determine optimal regulation plans.\nFrom the conservation of matter principle, a water balance equation can\nbe written for each lake as follows:\nAS = P + R - E + I - 0 G,\nwhere AS is the change in amount of water stored in the lake, P is precipitation\non the lake surface, R is runoff into the lake from the surrounding land area,\nE is evaporation from the lake surface, I is inflow from the upstream lake, 0 is\noutflow from the lake through its natural outlet, D is diversion into (+) or out\nof (-) the lake, and G is ground water flow entering (+) or leaving (-) the\nlake. All variables are expressed in the same units and for the same period of\ntime. Obviously, any variable may be equal to zero for a lake where it is not\npertinent. The change in amount of water stored in the lake, AS, is a positive\namount when supplies exceed removals and is a negative amount when removals\nexceed supplies.\nThe Tides and Water Levels Section, Marine Sciences Branch, Canada\nDepartment of Energy, Mines, and Resources; National Ocean Survey, NOAA, U.S.\nDepartment of Commerce; and Detroit District, U.S. Army Corps of Engineers,\nmaintain water level gages on the Great Lakes, rivers which connect the Lakes,\nand channels in which water is diverted into or out of the Lakes. Change in\namount of water stored in a lake is calculated from the area of the lake and the\nmeasured change in the elevation of the water surface over a period of time.\nThe amount of inflow from the upstream lake, outflow from the lake through its\n93","natural outlet, and diversions into and out of the lake are determined from\nthe water level records and rating curves which give the relationship between\nthe amount of flow past a point and the surface elevation of the water at that\npoint.\nPrecipitation, evaporation, runoff, and ground water terms in the equation\nabove are sometimes combined into a single term which is called the net basin\nsupply (NBS) to the lake. The water balance equation can then be written as\nfollows:\nAS = NBS + I - o + D.\nThe value of the NBS term in the equation can be determined as the sum of the\nprecipitation, evaporation, runoff, and ground water contributions to the lake,\nor it can be determined as the residual after the value of the other terms have\nbeen determined.\nThere have been numerous studies of the hydrology of the Great Lakes and\ntheir subbasins. Buetikofer and Meredith (1972) prepared an annotated\nbibliography of studies made prior to 1972. The latest description of the\nhydrology and hydraulics of the Great Lakes system, including a discussion\nof factors which affect the water supply and the response of the system to its\nsupply, was prepared by the International Great Lakes Levels Board (1973).\nThis study, as with most of the studies on Great Lakes regulation, is oriented\ntoward use of the NBS as the hydrologic input to each lake. Each historical\nNBS value was computed as the residual after the value of the other terms in\nthe latter equation above had been determined. In addition to the historical\nstudies, attempts have been made to develop models to forecast the NBS for use\nin regulation of the Lakes. These studies are summarized by Meredith (1970)\nand the International Great Lakes Levels Board (1973).\nJones and Meredith (1972) determined monthly values for precipitation on\neach lake, evaporation from each lake surface, and runoff into each lake from\nsurrounding land areas for the calendar years 1946 through 1965. The former\nequation is not satisfied when the precipitation, evaporation, runoff, river\nflow, and change in storage values are substituted into it (Jones and Meredith,\n1972). This indicates that either ground water, about which we know very\nlittle in the Great Lakes, should be considered, or there is some other\nexplanation for this discrepancy.\nThe consideration of the thermal expansion of water would change the values\nof the AS in the former equation and would have the effect of decreasing AS for\nmonths when the temperature is increasing and increasing AS when the tempera-\nture is decreasing. Recent results indicate that, for some months of the year,\nthe temperature effects on lake levels are on the same order of magnitude as\nthe NBS of the lake (Meredith, 1975a) A step-wise multiple regression\nanalysis indicates an apparent influence of upstream lakes on the precipitation\nin downstream basins (Meredith, 1975b). For example, a statistically signifi-\ncant relationship was derived which indicates that the precipitation in the\n94","Lake Erie Basin during June is a function of the evaporation from Lake Superior\nin May and June (Meredith, 1975b).\nLittle data exist concerning ground water contributions to the Great\nLakes. The usual assumption is that it is insignificant and can be ignored\n(International Great Lakes Levels Board, 1973). About the only data are from the\nwork by Haefeli (1972) concerning the northern shore of Lake Ontario. Ground\nwater contributions may become important in terms of water quality, especially\nif there is the practice of disposing of wastes into the ground water.\nAnother component that will come into play a little more in terms of lake\nlevels is the increase in consumptive use of water around the Great Lakes.\nIncrease in the amount of water that is taken out of the Great Lakes and not\nput back in will have a long-term gradual effect on the levels. This effect\nwill probably be minor when compared to other factors affecting lake levels.\nThe development of a regulation plan is dependent upon the data used\nand the procedure used to formulate the plan. The International Great Lakes\nLevels Board study (1973) used the monthly NBS values for the period from\nJanuary 1900 to December 1967 as the \"study period\" for the regulation study.\nThe NBS values were determined as residuals after the other terms in the latter\nequation were determined. These historical NBS values were used to develop\noperational regulation plans. Additional testing of the regulation plan was\nconducted, using 68 years of data generated by a multivariate model.\nThe river flows used were those values developed by the coordinating\ncommittee on Great Lakes Basin Hydraulic and Hydrologic Data (International\nGreat Lakes Levels Board, 1973).\nThe current regulation plans in operation on Lake Superior and Lake\nOntario were based on hindsight. A regulation plan which would benefit one or\nmore interests was established somewhat arbitrarily, and the effects determined\nby computing the resulting levels and outflows that would occur with this\nregulation plan if the historical sequence of NBS values were to occur again.\nIf the regulation plan did not satisfy the objectives of criteria for regula-\ntion over the critical period, adjustments were made to the regulation plan\nand the adjusted regulation plan was tested. This process was repeated until\na regulation plan satisfied the objectives or criteria over the critical\nperiod (International Great Lakes Levels Board, 1973).\nThe International Great Lakes Levels Board (1973) used dynamic programming\nand a successive approximation technique to develop trial regulation plans\nusing the January 1900 through December 1967 historical sequence of NBS values.\nThese trial regulation plans were then tested by using synthetic sequences\ngenerated by multivariate models.\nNeither the current regulation plans nor the best of the trial plans were\nable to satisfy the criteria during a test run using the critical period of\n1968 through 1973, during which time the Great Lakes Basin received extremely\nlarge amounts of precipitation (International Great Lakes Levels Board, 1973).\n95","Even the use of synthetic sequences in the testing of trial plans does not\nindicate how the Great Lakes system will respond to the plans under more extreme\nconditions. Most attempts to generate synthetic sequences of flow variables\nare aimed at preserving the statistics which are used to define the historical\nsequence. If the historical sequence of values does not contain the most\nextreme events that can occur, then any synthetically generated sequence will\nmost likely not contain the extreme events either because the multivariate\nmodel is designed to preserve the characteristics of the historical sequence.\nI know of no case in which a synthetically generated sequence of events contained\na critical period which was more extreme than the critical period of the\nhistorical sequence.\nThe International Great Lakes Levels Board (1973) used a deterministic\napproach to optimization. The optimization was performed by using a particular\nsequence of flows; whether that sequence of flows was the historical sequence\nor a synthetically generated sequence does not matter. Lake Superior regulation\nplans have been developed by using a nondeterministic approach to optimization\n(Su, 1971; Su and Deininger, 1974). The inflows are treated as stochastic\nrandom variables, and this stochastic nature is incorporated directly into the\noptimization technique. However, this technique results in excessive computation\ntimes when applied to the entire Great Lakes system. A greatly simplified\nexample for a four reservoir case required 161 minutes of computation time\n(Su, 1971).\nMorris (1974) proposed a modeling procedure for utilizing all the relevant\ninformation in a multiobjective decision-making scheme to develop an optimum\noperating policy of the Great Lakes system. A multiyear linear-screening\nmodel is postulated to provide an initial regulation plan, and a simulation\nmodel is to be used to evaluate proposed alternatives.\nOther studies for determining optimal operating rules for multiple-purpose,\nmultireservoir systems might also be readily transferable to the Great\nLakes system. One such approach is a two-dimensional dynamic programming approach\n(Rood, 1974). .\nMuch of the concern with flows in the connecting channels has been with\ndetermining the effects of ice retardation or with flow conditions at regulatory\nworks. The ice retardation problem has been analyzed, using a hydrologic\nresponse model with stage-fall discharge relations, rather than using the\nhydraulic routing techniques to determine the flow conditions (International\nGreat Lakes Levels Board, 1973; Quinn, 1971, 1973).\nTo sum up, Great Lakes hydrology has been briefly discussed and sources\nof more complete information indicated. Procedures for determining plans for\nregulation of lake levels due to changes in lake water volume have been\ndescribed, but changes in lake levels due to waves, tides, wind, and pressure\ncannot be controlled by regulation.\n96","The Great Lakes system is subject to natural regulation. The approaches\nand techniques used in the past have resulted in regulation plans which provide\nfor more efficient use of the Great Lakes than if there were no regulation.\nHowever, there needs to be further studies on lake regulation plans. There are\npowerful optimization techniques which could be used on this problem. There\nmust be some procedure devised to allow for testing of regulation plans for\nextreme conditions which are worse than have ever occurred in the past.\nWe are just beginning to understand Great Lakes hydrology. We are just\nbeginning to approach the problem from other than a lumped parameter model.\nThe application of conceptual models in the study of Great Lakes hydrology will\nbe another improvement in our knowledge and understanding.\nThe decision processes of any regulation plan require some knowledge or\nassumption of future water supplies to a lake. Forecasts of weather would\nenable the extension of hydrologic forecasts. Current skill in forecasting\nweather and related phenomena can only be measured in terms of a few days. The\nInternational Great Lakes Levels Board (1973) reports that, with 4-month perfect\nforecasts, benefits on the Great Lakes can be increased by one-third.\nWe have just begun; there is much to do.\n97","REFERENCES CITED\nBuetikofer, L. B., , and D. D. Meredith (1972), Annotated bibliography on\nGreat Lakes hydrology, Water Resources Center, University of Illinois\nat Urbana-Champaign, Urbana, Illinois, Research Report No. 56.\nHaefeli, C. J. (1972), Groundwater inflow into Lake Ontario from the Canadian\nside, Inland Waters Branch, Department of the Environment, Ottawa,\nCanada, Scientific Series No. 9.\nInternational Great Lakes Levels Board (1973), Regulations of Great Lakes water\nlevels, Report to the International Joint Commission, Washington, D. C.\nAppendix A--Hydrology and hydraulics\nAppendix B--Lake regulation\nAppendix C--Shore property\nAppendix D--Fish, wildlife and recreation\nAppendix E--Commercial navigation\nAppendix F--Power\nAppendix G--Regulatory works\nJones, D. M. A., and D. D. Meredith (1972), Great Lakes hydrology by months,\n1946-1965, Proceedings of the 15th Conference on Great Lakes Research,\nInternational Association for Great Lakes Research, pp. 477-506.\nMeredith, D. D. (1970), Modeling of the Great Lakes water system, Water Resources\nBulletin 6, pp. 55-64.\nMeredith, D. D. (1975a), Temperature effects on Great Lakes water balance\nstudies, Water Resources Bulletin 11 (in press).\nMeredith, D. D. (1975b), Predictive models for Great Lakes hydrology, Water\nResources Bulletin 11 (in press).\n.\nMorris, E. G. (1974), Modeling of the Great Lakes water levels, M.S. Thesis,\nMassachusetts Institute of Technology, Cambridge, Massachusetts.\nQuinn, F. H. (1971), Quantitative dynamic mathematical models for Great Lakes\nresearch, Ph.D. Thesis, University of Michigan, Ann Arbor, Michigan.\nQuinn, F. H. (1973), Effects of ice retardation on Great Lakes research,\nProceedings of the 15th Conference on Great Lakes Research, International\nAssociation for Great Lakes Research, pp. 549-555.\nRood, 0. E. (1974), Optimal operation of a reservoir system, Ph.D. Thesis,\nUniversity of Illinois at Urbana-Champaign, Urbana, Illinois.\nSu, S. Y. (1971), Optimal operating policies for multiple-purpose, multi-\nreservoir systems, Ph.D. Thesis, University of Michigan, Ann Arbor,\nMichigan.\nSu, S. Y., and R. A. Deininger (1974), Modeling and regulation of Lake\nSuperior under uncertainty of future water supplies, Water Resources\nResearch 10, pp. 11-25.\n98","1.10.1 Response - F. Quinn\nOn the overall aspects of lake hydrology, a series of models relating\nto the water quantity are available. They are basically mass continuity models\nwhere the flows are routed through the system. The model inputs consist of\nevaporation, precipitation, and runoff, either individually or considered as\nthe lumped term, NBS, into the lake, that Dale Meredith mentioned. Also, the\nhydrologic response models that are now being used encompass the regulation\nplans for Lakes Superior and Ontario. As many of you are probably aware, both\nof these lakes are completely regulated by man. The middle lakes in the\nsystem, Michigan, Huron, St. Clair, and Erie, are all governed by their natural\nresponses. Therefore models for the system consist of the operational regula-\ntion plans for Superior and Ontario and the natural response models for the\nrest of the system. Several areas need a much larger input as Dale Meredith\nhas brought out. One is on the interactions between precipitation and runoff.\nIt has been pretty well documented just by looking at time-series relationships\nbetween precipitation and lake levels that the lake levels lag the precipitation\nby about 2 years. This is a result of the precipitation-ground interaction in\nwhich the basin reacts similarly to a sponge. However, there is a problem of\nquantifying this and using it as a type of predictive model to determine the\ninputs into the hydrologic response models. In addition, as has been mentioned\nin most of these hydrologic studies in the Great Lakes area, ground water\nhas been completely neglected. The reason for this is that no one has any\nideas as to what its contribution is. A study is going on now in relation to\nIFYGL which may give some insight into this problem.\nOne of the things that should be mentioned about the hydraulic models is\nthat they also serve as a major input in water quality models. If you want to\nknow how much of a constituent is coming into one lake from another on a\nvolumetric basis, you have to know how much water is coming through that system,\nwhere that water is coming from, and where the water is going. This is some-\nthing that the hydrologic models provide. In addition, there are hydraulic\ntransit models and steady-state models of the connecting channels. At the\ncurrent time, I have models for both the Detroit and the St. Clair Rivers.\nThese models have as inputs the upstream and downstream hydrographs for example,\nfor the case of the Detroit River, hydrographs for Lakes St. Clair and Erie are\nused. With these as the forcing functions, river flows can be computed on an\nhourly basis at about four or five sections in the Detroit River. The impor-\ntance of this stage-discharge relationship has been brought out earlier. Wind\ntide and seiches on Lake Erie can cause flow variations coming into the lake of\nbetween 2,200 and 9,100 m 3 D -1 of water. This type of variation can occur over\na 12- to 14-hour period. If what is happening biologically or chemically\nis being monitored during any of this time, it becomes necessary to know how\nmuch water is coming in at any particular time. We have used models in several\nstudies. One study for the Great Lakes Regional Office of the International\n99","Joint Commission used the Detroit River model in conjunction with its chloride\ndata to compute loadings and to determine the effect of river flow variation on\nloading and on the representativeness of the sampling as reflected in Detroit\nRiver loading.\nLooking toward the future, one of the prerequisites for a better hydraulic\nmodel will be more accurate discharge information. It comes back to the same\nthing; a model was formulated using the equations of continuity and motion, and\nnow we want to calibrate that model. But, to date, all we have is discharge\ninformation which was made basically at one point during a limited time and\nusually many years ago. What we need to do is to devise a measuring system and\ngo out and provide additional and better discharge measurements for the river\nmodels. To illustrate the importance of the river model, for Lake Erie\napproximately 70 percent of the water which comes into the lake enters via the\nDetroit River, and about 70 to 80 percent of the water that leave the lake\ndischarges through the Niagara River. Therefore you can see that slight errors\nin the computation of the river flows can create considerable error in the\nwater balance and consequently in the chemical models.\nLooking toward the big picture many years in the future, I can see, and\nthis is a suggestion which has been brought up several times by Dr. L. Bajorunas,\nthat one of the important aspects of the ecology of the Great Lakes may be in\nterms of water quality regulation. The water quality of the Lakes varies with\ntime, and during various times of the year pollutants may be in different\nareas of the Lakes. Therefore all the models which are being derived, those in\nthe biological and chemical realm, those in lake circulation, and those in the\nhydrologic realm, must be combined into operational models using operations\nresearch to regulate the Great Lakes. The long-range view of what might\nconceivably come to pass indicates that all these models may be amalgamated\ninto a large scheme which will provide, in addition to the current lake-level\nregulation, perhaps a more important regulation as far as the future of the\nGreat Lakes in terms of water quality.\n100","2. WORKSHOP DISCUSSION SESSION\n2.1 Guidelines for Work Groups - E. J. Aubert\nFirst of all, I will comment on the work group makeup. Each of these\ngroups has a chairman, a scientific secretary, and various members. I have\nalready mentioned that some movement between working groups is desirable.\nToo much motion is probably going to be chaotic; exactly how the line is drawn,\nI leave to you. The handout material (Appendix 1) contains a sheet entitled\nWorkshop Group Membership.\nIn my introduction this morning, I mentioned five purposes of the work-\nshop that I consider relevant. Whether these are all compatible in one work-\nshop, I am not sure. The first objective is to identify future Great Lakes\nenvironmental research initiatives, i.e., major programs required to provide\na satisfactory state-of-the-art in environmental simulation and prediction to\nsupport the decision process for Great Lakes activities. This includes predic-\ntion, simulation, and those experimental studies necessary to support the\nmodeling effort as well as the environmental description\nThe second objective is to provide an opportunity to the Great Lakes\nresearch community to discuss and recommend future Great Lakes environmental\nresearch initiatives.\nThird is to consider possible United States-Canadian joint research\ninitiatives.\nFourth is to identify logical follow-ons to IFYGL.\nFifth is to explore the priority environmental research needs of NOAA,\nviewing NOAA operating units as users of environmental information. They\nactually are producers, but you can look at it from the point of view of\ndeveloping research products to backup the operating environmental units.\nThe National Weather Service is one of the major operating units of NOAA, but\nit is not the only one that was considered. Sea Grant and Coastal Zone Manage-\nment are others. Amor Lane, one of the NOAA representatives, could comment\non this point. He has responsibilities at a program management level in Sea\nGrant, Coastal Zone Management, and Project Independence, which, translated\ninto everyday language, means energy-environment problems. This is a Department\nof Commerce activity.\nThe purpose of our plenary session was to set the perspective for work\ngroup sessions, and it is perhaps clear that not all of the topics that are\nrelevant to some of these work groups have been covered, either by the principal\nspeakers, or by the responders. This gap was brought out clearly relative to\nan operational problem with surface wave prediction, which has not been\nmentioned. The National Weather Service considers wave prediction to be an\nimportant forecast problem, and I hope that one of the work groups will discuss\nthis. In fact, since National Weather Service people are here, I hope they\nbring up such problems. I hope it gets discussed from the point of view of the\nstate-of-the-art and what future research is meaningful. Wave prediction\ncould be included in one or two workshop sessions. It could be in the Water\n101","Movements group, which is where it was in IFYGL. Clearly, the action of waves\nis not only related to the stress on the lake but also to the condition of the\nlake. So one must know something about the lake thermal structure in order to\nknow how the surface stress is going to affect waves. Another aspect of surface\nwaves concerns the scope of the boundary layer. As I would view the boundary\nlayer, it includes not only the atmosphere down to and including the top skin\nof the lake but also the upper layer of the lake. As Mariano Estoque and others\npointed out, the stress is very much dependent upon the stability conditions.\nThe effect that a certain synoptic situation has on the surface waves is very\nmuch dependent upon the lake-atmosphere boundary layer and the situation it is\nin relative to a synoptic weather situation. Surface waves as a topic fall\nbetween the Water Movements and Boundary Layer work groups. If time can\nbe found to get together, it would be appropriate to consider this jointly. I\nsuggest that the co-chairmen meet on this topic tomorrow morning for an hour\nin joint session. Is that reasonable? Do you have a suggestion?\nBaer. There are two or three other things that I think go along with waves,\nfor example, storm surges, which present serious practical problems. Also, I\nhave not heard ice mentioned. I presume that most of the ice is caused by cold\natmosphere. I just wonder if all of that could not go in one session.\nCsanady. Yes. But you see from experience with past planning sessions in\nIFYGL that, unless you have somebody representing some of these constituents\nin the work group, there is not going to be much significant discussion because\nwe do not have the competence or the interest to go into these things. There\nmight be more interest in the atmospheric group in waves, but even then I would\nsuggest you attack this separately. Waves and the littoral zone transport,\ncoastal erosion, and so on, seem to hang together and require the calling of\nanother group whose prime interest is in this field and who could make a useful\ncontribution. We could then either say yes or ignore it. This is what\nhappened last time.\nAubert. What you all say makes sense. My point is that the work group sessions\nshould not be restricted by the scope presented in the plenary sessions. How-\never, the work group sessions will be limited by the interests and capabilities\nof the people attending. Unless these topics are raised, they will not be\ndiscussed. Since there are more people in the work group sessions than there\nare people who have made presentations at the plenary session, it will be up\nto this broader membership to bring up these other topics. How well these\nadditional topics get discussed is not known. They may be inadequately covered.\nBaer. Could you put it all some place so the interested people would know\nwhere to go instead of picking up a little bit all over the place?\nAubert. We mentioned a few topics, but there are more. You cannot separate\nthem all. I do not think you have a reasonable question. I cannot leave\nthis without mentioning something I saw in print from Will Pearson; it must\nhave been about 10 years ago, but is relevant to this mix of purposes for this\n102","workshop. I do not know which people in the marine environment try to cross\nmarine animals to generate new species, but this guy was trying to cross an\nabalone with a crocodile. Somebody put this together and came out with an\nabadile. Somebody else put it together and came out with a crocobalone.\nMaybe these purposes are not completely compatible. Be that as it may, we\nwould like to achieve the maximum possible from the competence that we have\ngathered. The plan is, then, that we will reconvene in our working group meet-\nings at 8:00 p.m. and again in the morning. We will reconvene here in plenary\nsession after lunch. Are there any questions?\nCsanady. What do you expect in the plenary session? A detailed presentation\nof what we want to do in 1977?\nAubert. Not what you want to do, but what you think are priority research\nproblems. Also something about how you may go about it if you have done that\nmuch discussing. Problems, I think, ought to be defined with some thought to\ntheir relative importance within the scope of your discussions. While the\nquestions \"What have we learned? Where to go from here?\" only appeared in the\nProf. Mortimer introductory plenary session this morning, they apply to all of\nthese topics and not just IFYGL. What have we learned from all of the research\non the Great Lakes? People who are not IFYGL participants have been invited\nto this workshop, and many of you have pursued other Great Lakes research, too.\nFrom what you understand the need is--and we do have some people here who ought\nto speak up on need, at least from these NOAA groups--what are the important\nresearch problems that ought to be attacked? Include something about how, but\nyou are not going to develop a research plan during the work group sessions.\nIdeally, we will come out with the proper research questions to be asked and,\nideally, some of the objectives one might want to pursue.\nHolland. This is not for a FY 1977 field program? Looking over time, what should\nbe done during FY 1977? This is for field work that might be in 1980, or\nanalysis of data collected in 1981 or whatever.\nAubert. I am looking at it from the point of view of something new. I would\nsee IFYGL, as a formal program, terminating in 2 to 3 years. That is as far as\nIFYGL goes in the budget process. It will no longer be a line item. Whether\nthere are future Great Lakes research initiatives involve a lot of decisions\nthat go higher up than this Laboratory--the budget channel and that sort of thing.\nHolland. It could be the start of a 5-year program? It does not have to be a\n1-year program carried out in 1977?\nAubert. No, I would view this as a major effort. It could be a 5-year effort.\nIf there are research objectives of importance, they should be identified. I\ndo not view the objective of this workshop as telling our Laboratory in detail\nwhat we ought to be doing in the various projects we are now working on. We have\nan on-going program. There are other considerations that will go into that.\nI am asking you to identify a logical follow-on to IFYGL, if there is one.\n103","Mortimer. Can you say something about the interagency arrangements? I believe\nyou had one conference last year. Is there another interagency conference\nplanned for this year? How are these kinds of programs in the EPA and the\nAtomic Energy Commission going to weld together with the NOAA effort?\nAubert. You are referring to the First Federal Conference on the Great Lakes\nheld here in Ann Arbor in December 1972. This was sponsored by the Interagency\nCommittee on Marine Science and Engineering (ICMSE). Dr. Robert White is the\nchairman, and each of the major agencies or departments has a representative\non ICMSE pertaining to the marine environment in the Great Lakes. A second\nconference is now in the early planning stage.\nMortimer. The scale of Great Lakes research now calls for interagency programs.\nEven IFYGL was funded by multiple agencies. The biological work was funded\nby EPA, and there was fairly considerable National Science Foundation funding.\nAubert. Your question was broad, so I cannot answer it in a few words. A\nsecond ICMSE conference is planned.\nMortimer. Should we think about interagency programs or about NOAA only?\nAubert. Clearly, there is overlap in the mission of several of the U.S. Federal\nagencies pertaining to environmental research.\nMortimer. There are also the principal users. The EPA is a major user for\nmonitoring and enforcement, and the Atomic Energy Commission is the user for\npower generation and dispersal and fate of radioactive materials.\nAubert. I guess the answer is yes. I will tell you what I plan to do relative\nto interagency information as a result of this conference. We could not invite\nmore people; obviously the room is full. We did not want to expand the\nobjectives and purpose of this meeting to invite all the other U.S. agencies\nthat have environmental missions and research programs on the Great Lakes\nbecause the scope of this workshop would be so broad that we could never get\ndone in a day and a half. But recognizing that suggestions might come out\nthat clearly overlap the other agency missions--namely, fish, pollution, power--\nI will make the other Federal agencies aware of the results of this meeting,\ngiye them a copy of the proceedings, and explore any ideas that might have some\ninteragency merit with them.\nMortimer. You want us to think about science and not politics.\nAubert. I prefer that you do that; and yet consider science from the point of\nview that it is problem-oriented, not knowledge for the sake of knowledge, but\nknowledge for better management of the Great Lakes.\nMortimer. Call it strategic research.\nHess. We should not try to design another IFYGL for another couple of years.\nThe major field activity should be more spread out than that.\nAubert. That comment came from a high authority; the situation being what it\nis, it is better not to design another IFYGL.\n104","Hess. There will be some arrangements about interagency coordination. You are\nasking us what it is, and I do not think either one of us knows right now.\nIt would be a waste of effort to come up right away with another major field\nactivity like IFYGL. Address yourself to the problems, but do not try to\nput them all into one big bag to try to solve them like that. Let's have\nsomething that is evolutionary.\nAubert. I did not fully answer your question, Prof. Mortimer, but I think it\ncould take half an hour. There is a second Great Lakes conference, sponsored\nby ICMSE, which is in the preliminary planning stage. The date has not yet been\nset. The first planning session took place yesterday and Dr. Bajorunas from\nour lab attended. The Atomic Energy Commission at Argonne, Illinois, has the\nlead. EPA was the lead agency at the earlier conference. I gather from\nDr. Bajorunas that yesterday's meeting did not result in any clear direction\nof where they were going or when. More planning will be necessary to get that\npoint, but ICMSE has requested that another conference be held. Any other\nquestions? This session is adjourned.\n105","2.2 RECOMMENDED RESEARCH INITIATIVES\n2.2.1 Water Movements - G. Csanady, Chairman\nThis work group addressed the research initiative problem under \"what\"\nand \"how.\" We have several recommendations under each.\nOur first recommendation is that we should fully exploit the present\ndata base. This is, of course, something we have already agreed upon, but\nnevertheless we would like to put on record the strong suggestion for a broad-\nbased exploitation of existing data from IFYGL and earlier Great Lakes studies.\nBy broad-based exploitation, we also mean to include the interrelation of\neach individual's work with the work of others. This kind of activity is\nonly now beginning to start as data become widely available. We are only\nrecently in a position to take advantage of what other people collected during\nIFYGL. Having looked at our own data, we should now look at everyone else's\nto exploit them and make whatever scientific advances we can. Also in this\ncategory is the verification of numerical models. Models of physical processes\nmust be tested against existing data in a broad kind of way.\nThe next point relating to future research is the high priority we place\non work on nearshore-offshore exchange processes. A concentrated study of time\nand space scales of flows nearshore, and specifically of the structure and\ndynamics of fronts, is required. The interchange of momentum, heat, and\npollutants across fronts has relevance to research on the Great Lakes as well\nas to general oceanography. The effects and parameterization of friction\nnearshore are also important. Coastal irregularities and their effects on\ngeneral circulation, the coastal entrapment of materials, and flushing processes\naround bars, bays, and prominences all come under this heading of nearshore-\noffshore exchange.\nThe next recommendation in order of priority is a further study of\nlarge-scale long-term lake circulation or, if you like, circulation climatology.\nWinter circulation is of special interest in this context. Some field data\nextend into winter, but most do not, and this leaves an important gap in\ncurrent knowledge.\nOur fourth recommendation concerns vertical mixing processes. The surface\nmixed layer and its interaction with the atmospheric boundary layer, including\nthe overturning periods in the fall and spring, is of considerable practical\nand scientific interest and relates in an important way to the physics of\nturbulent friction in a stratified fluid.\nOther problems that have been raised here, such as wave studies and fore-\ncasting, beach erosion, beach movement, and ice movement, should be considered\nby a more competent panel.\nTurning now to the \"how\" of the program, one question is: \"Is a son-of-\nIFYGL desirable, and what would be the scientific purpose of such a program?\"\nWe agreed that coordination of scientific work would be beneficial. When\na number of investigators work together on physical problems relating to the\n106","Great Lakes, generally speaking, more is likely to come out of a coordinated\neffort than the sum of the parts. There is certainly a favorable point in\nthinking of a successor to IFYGL. Concerning the logistics of such a program,\nthe concept of again having a core program, with auxiliary programs arranged\nby individual investigators, is recommended. The core program should provide\nthe necessary background body of data much as it did in IFYGL.\nSouthern Lake Michigan is a good place from our point of view. It is\nscientifically interesting, reasonably simple, and accessible. This choice\nalso seems, although we are not the ones to say, to be politically wise.\nAs for details about how to carry out such a future program, the one\npoint we all agreed upon was the need for long-term, careful planning well in\nadvance of field operations. This is to define clearly the scientific problems\nto be attacked and to evaluate previous achievements.\nWe also discussed instrumentation. If we are going to use instruments\nwhich are relatively new, they should be tested and used by the people who are\ngoing to use them in the field well before a major deployment. In any coor-\ndinated program, it is essential to be able to rely on the instruments. In this\nproblem category, we also talked about using whatever technical achievements\nwould be available to us, including satellites, possibly blimps, and any other\ntechnological improvements or advances in the state-of-the-art. Although one\ncannot be too specific at this stage, it is desirable to develop instruments\ncapable of profiling temperature and current velocity. Such instruments have\nto be developed, and the whole project has to be attacked well in advance of\nits execution. A number of years are required to develop and test instruments.\nThe coastal chain, as a way of looking at the shore zone, has been very useful\nand will no doubt be used again, but it has a fair weather bias and other\nshortcomings. It would be desirable to learn from what we have not been able\nto do by this technique. To develop new techniques, it is necessary to start\nplanning fairly soon even if field work is done in 1980.\nThe instrument array in such a core experiment would perhaps be similar\nto\nthe one suggested by Ted Green. That is, we would probably have a central\narray somewhere between Chicago and Milwaukee, if we can agree with the other\ngroups that this is a desirable area.\n107","2.2.2 Aquatic Ecology and Water Quality - C. Schelske, Chairman\nOne of the approaches used by this group was to review what had happened\nduring IFYGL and to discuss what IFYGL provided in the way of understanding\naquatic biology. There was a dichotomy in our group between what the aquatic\nbiologists thought was important and what the modelers thought was important.\nI will try to represent both of those views, but, since some of the modelers\nwere not present at the end of the session, the modeling input may be limited.\nWhat do we feel IFYGL provided? In the way of understanding the biological\nprocesses, the IFYGL program was geared more toward understanding climate than\nunderstanding weather, if we can apply that analogy. Coordination needed for\npredictive modeling was lacking, mainly because the biological-chemical design\nwas added to the original program at a late date. More time was needed for in-\nteraction and development of programs. Good data were obtained from IFYGL,\nbut we feel the next step should be for models that will predict weather and\nnot climate. Weather, in this case, is the sum of the processes in the nearshore\nzone, where the frequency of the phenomena is much greater, and the phenomena\nare more varied than in the open lake. Another reason is that most of the\nsevere management problems are in the nearshore zone. The IFYGL design,\nparticularly from the biological point of view, was fairly well restricted to\nthe offshore waters. The people who worked nearshore has little help from the\nphysical modeling point of view, and that is an essential element for future\nstudies.\nWhat do we do with existing data, IFYGL, as well as other data sets? Five\npoints related to this question were identified.\nFirst is the need to identify different existing data sets. This becomes\nmore and more important with the passage of time. As more and more new investi-\ngators come into the system, the need to identify sources of data becomes more\ncritical. Eventually studies that have been done may be lost. There are\nunpublished results that should be identified and collected in some organized\nform.\nSecond is the need to develop a means to make this type of information\navailable to investigators. This is something people could do without actually\ngoing out and collecting data, and it might be a profitable way to spend money\nfrom the standpoint of government agencies. Many university people have students\nthat would be interested in certain aspects of this problem.\nThe third point is that funding must be provided for the analysis of\nthese data sets.\nFourth is the need to determine what kind of biological samples are\navailable that have not been analysed and whether they have been stored properly.\nSome samples have gone to the Smithsonian Sorting Center, but other samples may\neventually be discarded. Ideally there should be a museum that would curate\nsamples, or maybe a Federal laboratory with a museum in it that would perform\nthis function.\n108","Fifth is the need for a continuing effort on the problems addressed above.\nThese problems will continue as long as science, so we might as well face them\nnow. It may be significant that Dr. Beeton, who is probably the senior member\nin our group from the Great Lakes point-of-view, felt that this was very impor-\ntant and everyone agreed with him. One of the reasons for making this point\nis that sometimes the perspective of experience is needed to realize the\nimportance of factors such as continuing programs.\nRegarding new research initiatives, we had trouble with specific research\ninitiatives, but neatly avoided the issue by proposing a title for the new\nresearch initiatives. It is \"Biological and Chemical Processes As Influenced\nby Materials Input and Transport in the Nearshore Zone.\" I will outline this--\nthe where, what, why, and how of these initiatives. \"Where\" is, of course,\ndefined as the nearshore zone. Specific nearshore zones will be considered\nlater. The next question is why we picked this title. One reason already\nmentioned is that many practical problems are in the nearshore zone. The scale\nand frequency of measurements that would be needed would be a function of the\nspecific problem of interest. It is difficult to define research objectives\nwithin 1 day. We know there is a gap; in fact, we might say there is a lack\nof biological and chemical knowledge about the nearshore zone, particularly\nwith regard to modeling. Available predictive models, as I understand it,\ncannot address certain significant problems in the nearshore area. If that is\nnot correct, will somebody correct us. Therefore, there is need for information\nof a descriptive nature and data on significant processes in the nearshore\nzone prior to mounting large-scale modeling efforts. Finally, how do we do\nthis?\nFirst would be the comparison of different nearshore areas either within\none lake or between lakes. It is important to find out whether all of the\nlakes behave similarly or whether differences exist. The study should be\ndesigned so the inshore zone could be compared with the offshore zone. Any\nof\nyou who have ever been to a meeting where the inshore was discussed know\nquestions always arise as to the distance from shore to the offshore waters.\nThese studies should be designed so that offshore stations will definitely\nrepresent the open lake and therefore assist in defining the boundaries of the\nnearshore zone for a particular region. From this program, offshore water within\nlakes can be compared if more than one nearshore region is sampled, and if not,\nwe can certainly compare offshore waters among the lakes that are selected.\nSecond, there is a need for coordination among the aquatic ecologists\nand people working with water dynamics and water circulation.\nWhat will we obtain from this extensive study of the nearshore zone? First\nof all, there will be descriptive knowledge of the lakes. Prof. Mortimer\nchided me for possibly minimizing its importance, but I think all biologists\nrealize there is a need for descriptive knowledge and that there is a gap in\nthis area for the Great Lakes. We feel that the descriptive parts, at least,\n109","will be a byproduct of the experiments that are undertaken as part of the\nprogram. The needed but unspecified experiments will define mechanisms and\nprocesses in the environments under study. Although the experiments are not\ndefined, it is apparent that there is need for coordination due to the problems\nof scale. For instance, to coordinate with people who are studying nearshore\nwater transport, we have to be certain the appropriate data can be obtained for\nthe area under study since water transport is a significant input for any\nnearshore models as it is needed for materials transport. Our people also felt\nstrongly that we should have actual water transport data that were obtained\nwhile ecological data were being collected, rather than a water transport model.\nOur specific experimental design is very general, but this is to be\nexpected at such an early stage of a long-term program. We have proposed a\n5-year plan: 1 year for planning, 1 year for a feasibility study for testing on\na limited scale, and 3 years to run the actual experiments. A number of people\nstressed that we need data for more than 1 year. As some of you know, phenomena\nsuch as Hurricane Agnes occur frequently in the Great Lakes, resulting in\natypical years for ecological purposes, so we need data for more than 1 year.\nTime is needed for feasibility studies. I feel strongly, and I think\nmost of the panel agree, that feasibility studies should be carried on until we\nare actually prepared to do the experiments. This may mean, in some cases,\nthat studies never get beyond that point of feasibility, implying all projects\nshould not be continued throughout a funding cycle. That viewpoint could\nreflect a personal bias on my part.\nIt may not be possible to study more than three sites. Four of the five\nGreat Lakes have been selected, leaving out Lake Erie. Eventually one must\ndecide whether, for comparative purposes, to select one site in each of the\nmultiple lakes or to select multiple sites in one lake. Several options for\nsites were discussed. In Lake Ontario there were two sites--the Rochester, N.Y.,\narea and the Oswego, N.Y., area. Again, the scale of the study area is not\napparent at this point, so possibly one site could extend from Rochester to\nOswego.\nIn Lake Michigan, three or four sites were selected. If we take the lead\nof the previous group, we would end up with three. Possibly there is an area\nnear Chicago as well as an area near Milwaukee. We could not decide whether\nthis should be one or two areas, and I think the previous group put the\nmiddle of the study area halfway between Chicago and Milwaukee, resulting in\nonly one study area. There should be another area on the eastern shore of\nLake Michigan somewhere between Benton Harbor and Muskegon. Muskegon is almost\ndirectly across the lake from Milwaukee, and Benton Harbor is roughly at the\nsame latitude as Chicago, so this would provide east-west comparison. Then an\nunpolluted area of Lake Michigan should be included as well; it would have to\nbe in the northern part of the lake.\n110","Dr. Beeton proposed that we assess the land effect as it relates to water\nquality. He has collected data from Beaver Island. This is another way to\nstudy the nearshore--by selecting an island site in the middle of the lake with\nclean water and little pollution from land.\nIn Lake Superior, there are two logical study areas. One is the Keweenaw\ncurrent which flows along the south shore from the Keweenaw peninsula to\nWhitefish Bay. It has been recognized by a number of investigators, and it is\na discrete water mass. Then, of course, the area around Duluth is one that has\nbeen affected by man. We also felt there might be some justification in propos-\ning a study site in Lake Huron, particularly south of Saginaw Bay, since the\narea is being studied extensively this year as part of the Upper Lakes Reference\nStudy. The ongoing work will provide background data, but these studies are\naimed more toward the open lake than toward the nearshore area.\nA number of important items were discussed that have not been covered\nyet. I will conclude by presenting a shopping list of six or seven items.\nFirst is the need for high frequency sampling in the nearshore zone. Perhaps,\na nearshore experiment might last only 3 or 4 weeks, but sampling would be\nintensive during that period. Another way of undertaking high frequency\nsampling is to study a square meter of the lake, as Dr. Beeton mentioned yester-\nday in his response.\nSecond is the need to develop instruments, particularly instruments for\ncontinuous monitoring, so that data can be obtained without using a ship to\noccupy a station.\nThird is the important problem of pathogens that was outside the competence\nof our group. Pathogenetic organisms are released into Great Lakes waters, but\nlittle is known about their fate in the environment. Cooperative studies might\nbe arranged with agencies who have public health responsibilities; these studies\ncould be conducted simultaneously with the proposed program.\nFourth, a lot of people were concerned with sediment-water interchange, and\nI am sure it will come up again today.\nFifth, there is a great need not to neglect the study of the atmospheric\ncontribution of pollutants, even if we are talking about a localized zone. The\npreliminary phosphorus budget for Lake Huron is one-third the nutrient contribu-\ntion from the atmosphere, another third is from the two major inputs which are\nLake Michigan and Lake Superior, and the final third is from Saginaw Bay. This\ngives some idea of the importance of the atmosphere, and, of course, there are\nalso toxic or hazardous materials in atmospheric inputs.\nThe final thing which possibly should be stressed is that the term\n\"biological and chemical processes\" refers to studies at every level of the\necosystem, including phytoplankton, zooplankton, benthos, fish, and bacteria.\nThese processes include the function and quantification of various biological\ncomponents. Although this observation is apparent to most ecologists, we stress\n111","it because it may not be essential for presentations from other groups. We\nhave to be concerned with interactions, not only the biological interactions\nbut also the chemical interactions, among these different biological groups.\n112","2.2.3 Lake-Atmosphere Interactions - M. Estoque, Chairman\nThis group interpreted differently from the two previous work groups what\nit was supposed to do. Instead of considering general research initiatives,\nmembers of the group considered specific problems on lake-air interaction\nprocesses. This was done first by calling upon each member to suggest specific\nproblems. A list of the problems was made, and then each problem was discussed\nin detail in order to define it clearly. Finally, we assigned priorities to\nthe problems on the basis of socio-economic value, scientific merit, and\nresource (manpower and money) required for solution of the problem. As expected,\nwhen assigning priorities, each person was plugging for his own suggestion.\nPriorities were assigned by consensus among members of the group.\nThe specific problems which were suggested are summarized in table 9. The\nfirst two problems come under the general category of wave studies. The first\nitem under wave studies is concerned with the physical processes which are\nresponsible for the generation, growth, and breakup of waves. Investigation of\nthis item requires observational as well as analytical studies. The second item\nunder wave problems is the applied problem of wave prediction by semi-empirical\ntechniques. The empirical relationships will be formulated with the aid of\npressure distributions or other large-scale synoptic descriptive parameters\nwhich are observed over the land surrounding the lake. The next problem\nconcerns the prediction of surges. The problem should be restricted to surges\nwhich are induced by subsynoptic-scale weather disturbances. The next problem\nis the determination of the effects of waste heat disposal on the quality of\nthe environment, in both air and water. The next problem is concerned with\ncoastal erosion--the effects of wind stress and waves of coastal erosion under\nsevere weather conditions. The next problem involves the prediction of the\nspace distribution of ice and the physical characteristics of ice on lakes. The\nprediction of ice distribution needs an understanding not only of the physical\nproperties of the ice but also of the atmospheric conditions which tend to break\nup or melt the ice and transport them in the lake. The next topic is concerned\nwith the evaporation from the lake; this is one of the problems being studied\nunder IFYGL. The problem is the determination of the amount of evaporation\nfrom the lake, primarily for the purpose of using it for the analysis of the\nlake water budget. There is still some question in my mind about how the group\nwanted to approach this problem. I feel that what was intended was a determi-\nnation of the rate of evaporation from the synoptic conditions by empirical\ntechniques. The next topic concerns precipitation, also in connection with lake\nwater-budget calculations. When it was originally proposed, this problem was\nto be concerned only with determining the amount of precipitation from synoptic\nobservations by empirical methods. But as the discussion progressed, it\ngradually evolved into the more complicated problem of calculating precipitation\nby physically modeling the mesoscale disturbance generated by the lake. The next\n113","Table 9. Summary of Priorities Assigned to Various Problems\nRequired Effort\nSocio-Economic\nand Resource\nScientific Merit\nValue\n2\n1\nWaves (Physics)\n3\n1\n3\n1\nWaves (Empirical)\n1\n2\n2\nSurges\n3\n1\n1\nWaste Heat Disposal\n2\n1\n3\nCoastal Erosion\n3\n1\n1\nIce\n1\n2\n2\nEvaporation\n3\n3\n1\nPrecipitation\n2\n1\n1\nAir Pollution\n3\n1\nFluxes (T, M, Q)\n3\n2\n1\n2\nVertical Structure\n2\n1\n2\nNearshore Boundary Layer\n3\n1\n3\nSubmesoscale Processes\n2\n2\n2\nLake Effect Storms\n1\n2\n2\nWater Pollution Drift\nproblem is concerned with air pollution, with special emphasis on the role of\nlake-induced air circulation in transporting and dispersing pollutants over the\nlake and vicinity. In addition, the problem includes the transport of gaseous\nmaterial from the atmosphere into the lake. I had not expected this transport\nto be substantial so was happy to hear from the previous speaker that this\ntransport is important. The next topic is concerned with the basic problem of\ndetermining the fluxes of momentum, heat, and moisture from the lake surface.\nAgain, this is one of the important problems being investigated under IFYGL, but\nthe group feels that the IFYGL program will not completely solve this problem.\nDetermination of fluxes during highly unstable conditions, including strong\nwinds, and also as a function of fetch from the shore will probably not be\nsolved by current IFYGL studies. The group would like analogous studies over\nother lakes. Such studies will show whether empirical relationships formulated\nwith Lake Ontario observations are valid for other lakes. The next problem has\nto do with the determination of the vertical structure of temperature, moisture,\nand wind in the lower planetary boundary over the lake. This experiment was\nalso planned during the IFYGL field program, but I believe the plans were not\ncarried out adequately. The next problem involves the determination of the\nnearshore atmospheric boundary-layer - structure. This is important because the\nshoreline, which separates the land from the lake surface whose characteristics\nare sharply contrasting, will produce strong horizontal gradients under\ndifferent atmospheric conditions. Therefore, it has been suggested that this\n114","structure is an important consideration in the analysis of diffusion processes,\nespecially in relation to the pollution problem. The next problem is concerned\nwith eddy transports associated with disturbances of scales from 1 to 10 km,\nscales which were not observed during IFYGL. As one may recall, the observa-\ntion stations during the IFYGL field program were relatively far apart; there-\nfore, it was not possible to observe effects of disturbances of these scales\nin transporting various quantities, such as momentum, heat, and moisture. The\nnext to the last problem is the description and prediction of lake effect\nstorms. Finally, the last problem is concerned with pollution of lake surface\nwaters. Although the problem involves the water instead of the air, the group\nfeels it is appropriate for us to suggest it because the solution depends upon\nan adequate knowledge of lake-air interaction processes.\nAfter discussing the above problems, we assigned priorities. Three factors\nwere considered in assigning priorities. The three factors were the following:\n(1) the socio-economic value of the problem; (2) its scientific merit; and\n(3) the amount of effort and resources (money and scientific manpower) required\nfor conducting the research needed to solve the problem. Priorities in table 9\nare indicated by numbers from 1 to 3. In terms of socio-economic and scientific\nvalues, 1 means the highest value. From the point of view of effort and\nresources, 1 indicates the least amount of effort and resources required.\nThe highest priority problem would be that which has a 1 in all three columns.\nBut no problem had 1 in all three columns because, as expected, there is a\ntendency for a high rating in socio-economic value to go with a low rating in\nscientific merit. It has been suggested that the table may be used for deter-\nmining whether a university, private research organization, or government\nlaboratory is best suited to do a particular problem. The basis for this\nsuggestion is the notion that a university is best suited for undertaking a\nproblem with high scientific merit (basic research), while a private research\norganization is best suited for a problem with high socio-economic value\n(applied research). Moreover, a government laboratory would be ideal for doing\na problem which requires a large amount of effort and resource. For example,\na good problem for a university is one with a rating of 3-1-1; for a private\nresearch organization, 1-3-1; and for a government laboratory such as GLERL, 1-1-3.\nIn conclusion, the collection of problems which have been presented by my\ngroup looks like a rerun of the Boundary Layer and the Lake Meteorology programs\nof IFYGL. This similarity did not emerge by design, but rather by chance.\n115","2.2.4 Environmental Dynamics - C. H. Mortimer, Chairman\nThe environmental dynamics group had, we believe, one of the most difficult\ntasks of all. One of the difficulties was defining what is meant by environ-\nmental dynamics. This we took to include the physical dynamics of water, air,\nand solid substrate; the dynamics of chemical transformations and transports;\nthe dynamics of biological production; and the interactions between all these.\nAt the same time, although we did not consider it in detail, socio-economic\ninteractions should be kept in mind. These categories cover almost everything!\nPerhaps surprisingly, we did come up with a consensus on a number of points,\nand although we were instructed to consider science rather than policy, most of\nthe things we agreed on are in the policy area.\nWe agreed on the following:\n(1) The nearshore zone, defined hydrodynamically as 10- to 15-km wide,\nshould be the principal focus for post-IFYGL investigations and modeling\nbecause this is the zone of maximum physical activity, maximum chemi-\ncal and biological variance, and maximum human use.\n(2) The proposed concentration of field work, instrument arrays, and\nmodeling efforts in the nearshore zone must not lose sight of the\nfact that the physical and biochemical driving forces are developed\non larger whole-basin, drainage basin, and meso-atmospheric scales.\n(3) Active support for analysis of IFYGL data should continue (and this\nwas strongly emphasized) for several years to exploit this unique\nbase for progress under (2) and to plan the proposed nearshore zone\nstudy, including essential modeling testing activities listed below.\n(4)\nThe distinctive mission of GLERL should be development of the\nscientific basis and support, with the appropriate in-house interdis-\nciplinary expertise and facilities, for a post-IFYGL effort directed\nto the nearshore zone and for the Great Lakes mission-oriented pro-\ngrams or needs of NOAA.\nAmong the NOAA programs, we referred specifically to the Sea Grant Program and\nthe Coastal Zone Management Program and to interaction with the International\nJoint Commission.\nThe International Joint Commission was specifically mentioned because the\nResearch Advisory Board has been very active over the last year or so and has\ncreated a number of standing committees in water quality and physical and\nbiological fields. The Upper Great Lakes Reference Study will be coming to an\nend within the next 2 or 3 years and will be reported upon. If Canadian\ncooperation in the post-IFYGL activity or in other Great Lakes research activities\nis to come about, as we recommend, then the International Joint Commission\nprovides a convenient and proper vehicle. Recommendations of the International\nJoint Commission Research Advisory Boards and its standing committees and\nexistence of the new International Joint Commission headquarters office in\nWindsor will facilitate Canadian-United States collaboration smoothly and\n\"legally.\n116","Turning more specifically to the distinct mission of GLERL, while part of\nits research effort will be directly applied to problem solving, the distinctive\nfunction of GLERL should be development of fundamental understanding of natural\nand perturbed systems through:\n(1) Observation, experiment, and monitoring.\n(2) Model development, verified at each stage by data produced from new\nand, of course, existing data banks.\n(3) Some of the research themes should be basic, strategic, and long-term\nin nature, i.e., strategically selected to provide research support\nand research output needed by identified users, by other components\nof NOAA, and by other agencies involved in environmental management\nand decisions.\n(4) The emphasis should be on natural science, rather than on social\nscience, although we cannot ignore the social and legal aspects of\ninstitutional design which will be needed to put some of the scien-\ntific recommendations into effect.\n(5) In the planning and implementation of GLERL programs, including the\nproposed nearshore investigation, cooperative activities should be\nencouraged with the academic community, other Federal agencies, and\nresearch groups, both United States and Canadian--the latter case\nthrough the International Joint Commission as appropriate.\nMost of our debate was concerned not with these points of consensus, but\nrather with examples of activities to be undertaken. The following possible\nactivities within the GLERL mission were mentioned:\n(1) Designing a sampling network through a space-time analysis to develop\nthe optimum spacing and frequency required to understand natural dis-\ntributions and to follow significant trends.\n(2) Assembling and critically reviewing existing and emerging physical,\nchemical, and biological data for the purpose of model testing, model\ndevelopment, and design of effective long-term strategies. This\nwould include a critical review designed to detect and analyze the\nsignificance of long-term trends.\n(3) Intensively studying inshore-offshore exchanges and partition of\nenergy.\n(4) Studying the mechanics of upwelling and subsequent whole-basin\nresponses, generation and decay of nearshore currents, and transport\nof material, nutrients, toxins, and organisms.\n(5) Standardizing and intercalibrating methods of measurement and\nanalysis.\nWithin the framework of increased understanding of relevant physical,\nbiological, and chemical mechanisms operating principally in the nearshore zone,\nthe scientific basis for action on a number of present or emerging problems\n117","was discussed. These problems, identified by individual panel members as\nimportant, were energy management in the Great Lakes, recreational planning\nfor the Great Lakes, natural resource inventory and utilization, and water\nquality criteria. Under water quality criteria, the following subtopics\nwere proposed by individual panel members: chronic sublethal effects of\npollutants on lake ecosystems (not enough attention is being given to these\neffects, which probably have long-term significance) ; development of a scien-\ntific basis for evaluation, under the present U.S. laws, of tradeoffs between\ncosts, technical feasibility, and the social desirability of zero pollution\ninput; evaluation of stream loadings for both nutrients and toxic substances;\ncriteria and techniques for disposal of dredgings, involving water quality\ncriteria and recreational considerations; and shore and beach processes, including\nerosion and material transport. The Cladophora problem was also identified as\neutrophication effect in the Great Lakes of maximum public visibility.\n118","2.2.5 Water Levels and Flows - D. Meredith, Chairman\nDuring the workshop, the research objectives and information needs related\nto lake hydrology and regulation were identified.\nA conceptual hydrologic model of the entire Great Lakes system is needed\nthat will be responsive to existing and anticipated user needs. In order to\nmove toward this objective, we recommend the following actions:\n(1) Replace the present empirical relationships for computing monthly\nconnecting channel flows. This should include the implementation of\nthe equations of continuity and motion and a time scale required for\nthe development of an improved conceptual hydrologic model and other\ndiscipline needs.\n(2) Make simultaneous discharge measurements in the connecting channels\nto calibrate and tune the model.\n(3) Investigate the application and adaptation of existing rainfall\nrunoff, snow accumulation, and ablation models in order to define\nbasin runoff more accurately.\n(4) Obtain the necessary parametric input data (wind, temperature,\nradiation, etc.) to support a more complete conceptual hydrologic\nmodel.\n(5) Extend and refine hydrologic models to use remote sensing data.\n(6) Investigate the magnitude and flux of ground water in the system.\nThe first phase should be a limited investigation to determine its\nrelative order of magnitude, including bank storage. If warranted,\na full investigation would require a comprehensive data collection\nand analysis program for use in conceptual models.\n(7)\nDevelop conceptual models of lake ice information, growth, and decay.\n(8)\nInvestigate ice retardation in the connecting channels and the impact\nof ice on hydrologic and coastal zone processes. This will require\na data base which includes lake heat budgets, areal ice distribution,\nand geochemical composition.\nThe present status of lake regulation, based on trial and error procedures,\nhas proven to be inadequate. For example, both present and proposed regulation\nplans for Lake Ontario failed last year. We recommend the use of operations\nresearch techniques and stochastic inputs to derive improved regulation plans\nfor the Great Lakes. This will require the following:\n(1) The use of probabilistic and stochastic models to generate supplies\nto the Lakes. Models for both lumped and individual parameters\nshould be considered.\n(2) Sensitivity and optimization studies coupling the parametric inputs\nwith the constraints and criteria by which the optimum is defined.\n(3) Inclusion of water quality, water levels, dynamics, and other\nenvironmental effects in regulation criteria.\n119","PRIORITY RESEARCH INITIATIVES - E. J. Aubert\n3.\nYou may not have had an opportunity to digest everything fully, but this\nsession is now open for discussion. One topic for discussion may be areas of\noverlap between the panels or areas where there are differences or agreements\namong the recommendations. Any comments you have are appropriate at this time.\nScott. Much was said about the shore zone and I think we ought to define it.\nWe have to define what it is along the lines of what Prof. Mortimer was saying.\nI agree with Cliff. It's in quotes right now 'shore zone.\"\nChapra. The Aquatic Ecology and Water Quality work group had some discussion\nas to what was the \"shore zone\" in terms of its biology and chemistry. I\nwould like to point out that it might be defined differently from a physical as\nopposed to a biological or chemical point of view. This should be kept in mind\nwhen designing a field program to insure that the zone is described with a\nsensitivity to all important perspectives.\nMortimer. What is wrong with 15 km?\nAubert. Does anybody need a shore zone wider than 15 km?\nScott. I was thinking in terms of a water quality and biological definition\nas well as a physical definition, rather than an arbitrary boundary somewhere\nbeyond the coastal jet.\nMortimer. What drives the coastal zone biologically, chemically, and physically\nis, of course, the whole lake, including the regime of the regional atmosphere\nand the drainage area. So I do not think we should regard this nearshore study\nas being geographically defined in that way. I think there is going to be an\narray of observations concentrating in the area of immediate interest, but\nalso, in some cases, taking account of whole-basin motions.\nScott. I agree with the designation of scale. Maybe 10 km is a little small.\nMortimer. But we should give some order of magnitude. To some people, the\nshore zone is the beach zone where the waves break, causing shore erosion and\ntransport of solids. I would want to go to where the Kelvin waves become\nunimportant, and that is 10 to 15 km offshore.\nCsanady. I think 15 km would be fine.\nBeeton. In the Aquatic Ecology work group, we thought that we should have\nsampling out into the open lake, as Claire Schelske indicated. I do not know\nexactly where that would be. It might be 15, even 20 km, depending upon the\nlake; in order to study the perturbations that occur in the nearshore zone, we\nneed reference levels. So if we study processes out in the open lake and the\nsame kind of thing in the nearshore zone, then we might start to get a handle\non some of the things that are going on.\nAubert. Several people have referred to the sampling diagram proposed by\nTed Green. It shows the total lake being monitored, with the sampling intensity\ngreatest in a particular segment of the nearshore zone. I think several of us\nwere looking to that sort of a grid consistent with what you are saying,\nDr. Beeton.\n120","Mortimer. Such a pattern was followed somewhat in IFYGL with a concentration of\nobservations in the Niagara bar, if you recall, but not on this fine a scale.\nThe question of site has been raised. We did not consider that in the Environ-\nmental Dynamics work group. The question of Canadian cooperation has also been\nraised. This is a fairly long-term planning decision, and until we get a reac-\ntion from our Canadian colleagues, the question of sites has to be shelved, at\nleast for the time being.\nAubert. We can examine alternatives, but it is certainly premature to make a\ndecision at this point.\nBaer. I want some clarification on the same question I asked yesterday morning.\nYou said, if at the end I still had the question, to ask it again. It appears\nto me that both the Water Movements and Aquatic Ecology panels are speaking about\na son-of-IFYGL, a large-scale, massive program to cover a great area. I am\ntalking about things on a smaller scale, not a repetition of IFYGL.\nAubert. Why not have the chairmen of those two panels speak to your question.\nCsanady. I think the expression was used, but, when the Water Movements group\ngot down to details, the experiment turned out to require a relatively concen-\ntrated array. In our mind, it is one array; and in the Aquatic Ecology group, it\nis several arrays of relatively small dimensions. These arrays are to be\nsupported in order that they may be put into the broader picture, by lakeward\nmeasurements of decreasing intensity like the T. Green diagram which seems to be\na pretty much agreed-upon scheme.\nBaer. How would the interactions and cost relate to the original program?\nCsanady. In terms of expense?\nBaer. In terms of expense, number of institutions required, and other things\nof this nature.\nCsanady. I think the scope would be less than IFYGL, but that is my feeling.\nFrom what we have discussed, individual members might feel otherwise.\nSchelske. What was the IFYGL cost; for instance, the ships, buoys, and this\nsort of thing?\nAubert. I cannot cite numbers for specific parts of it, but our estimate of the\ntotal program is something like $30 million between the United States and Canada\nover a 7- to 8- year period, all of which has not been completed yet. A cost\nestimate depends on what is included. That amount was not earmarked on the U.S.\nside in a budget item called IFYGL. A lot of existing resources were directed\ntoward this cooperative effort, and I think on the Canadian side the major in-\nput was the redirection of existing resources. If they had not been working on\nIFYGL, they would have been working on some other Great Lakes activity. The\nnumbers could be added differently. Four or five large vessels were involved,\nwith many supporting ones.\nSchelske. You are concerned about the cost. That may be the wrong way to\napproach it. Unless it is a fairly large study, the returns may not be\nmaximized for the amount of money spent. I think we learned from IFYGL that\nwith a large number of people working in the same place utilizing common\n121","facilities, we got more for our money. So I think the size of the program has\nto be determined after the requirements are defined; if it cost $10 million,\nit is important to find $10 million. With more planning, one could get a better\nhandle on the needs. The Aquatic Ecology group felt that this study would not\nbe a son-of-IFYGL, but maybe a cousin to it. In terms of working in the Lakes,\nit is entirely different. We want to study local areas. IFYGL covered a whole\nlake, and the size of the areas that we propose for study may have been covered\nby only one station in IFYGL. It is an entirely different problem.\nAubert. The monitoring program would undoubtedly be quite different.\nSchelske. In the kind of program we are proposing, there might not be an\neffective way to use the Researcher.\nBaer. The point I wanted clarified was, what is the ultimate scale of research?\nI was not trying to find whether it should be big or little at this stage. I\nwas trying to determine your recommended optimum scale. Finances, people, and\ndollars all amount to the same thing at this stage.\nAubert. I think it could be put into a range between 0.1 and 1.0 of IFYGL.\nThe scale must be significant. The required data acquisition systems are not\nsuch that one could dip a thermometer in the lake and expect to come back with\nuseful data. It will require instrument development. The discussions referred\nto more than 1 year of monitoring--3 years of monitoring. IFYGL had primarily\n1 year. The distribution of monitoring stations would be altogether different\nso that the types of systems that might be deployed would have both similarities\nand differences. The time scale of monitoring also would probably be different.\nHolland. The program will be of the pilot type through the second year, with\nfield work for 3 more years. You are going to need time to complete the\nprocessing and analysis of the data, so it will not be a 5-year program. It might\nbe a 7- to 8-year program. Even with good advance planning and quick turn-\naround of the data processing, it still takes time to digest, analyze, interpret,\nand integrate. Another point is that we have not reviewed ongoing programs, and\nI would hate to see the implication left that, if one major thrust is identified\nfor high priority, it consumes all the effort so that essentially nothing else\ngets done during this time. We have not addressed this question, but I think\nour assumption is that other efforts on the whole lake or in the middle of the\nlake or other problems will be taken on their merits and not swallowed up by\nthis program. This is a new initiative over and above other things that may\nbe ongoing.\nAubert. Prof. Mortimer can speak to this latter point because the Environmental\nDynamics panel discussed this question from the point of view of whether we\nare talking about single-agency funding, multiagency, commercial, or a host\nof other potential sources of funding.\nMortimer. Speaking for myself, I am sure that people responsible for university\nprograms would be willing to modify them in order to participate in a program\nof this kind. There would be nothing more important that we could do for the\nGreat Lakes than participate. The question of other agencies needs consideration.\n122","Our panel did not review that matter because it would be getting into politics.\nThere are, for example, two other agencies developing missions on Lake Michigan,\nand it would be sensible planning to find out what they are doing and, where\npossible, to fuse the work together. I am not a fiery proponent of large-scale\nscience, but in projects of this kind, particularly if we have Canadian coop-\neration, we can do something better together than we could do it separately.\nMost people would have a great interest in participating and would drop every-\nthing they are doing for those years. The availability of funding would be the\nkey as well as the availability of research vessels which are expensive. We\ndo not need another Researcher, particularly if it is only going to work 41/2 days\na week as it did in IFYGL.\nSchelske. I want to reemphasize that the proposed biological program is one\nthat would be important from the practical point of view. Again efforts have\nto be concentrated in the nearshore zone where there are tributary inputs,\nmunicipal water intakes, and a whole range of ecological and sociological problems.\nWe need to know about short-term response as well as long-term response.\nAubert. Many of the topics that Clifford Mortimer mentioned from the Environ-\nmental Dynamics panel are obviously long term. IFYGL was a project with a\nplanned start and end. This Laboratory has a long-term mission which is not\nof a project duration, although I cannot say what the duration may be. The focus\nof this workshop was placed on a next major initiative in the Great Lakes. There\nappears to be unanimity for this focus to be concentrated in the nearshore.\nMany of the activities that have been mentioned are of a longer duration than\n5 to 7 years; they are decades. We cannot wait that long to produce useful\nanswers, but neither can we work simultaneously on all of these problems. Some\nsort of a priority listing must be established. If at some future time the\npriority shifts or when some are finished, the list can be modified. Work can\nonly be done on problems at the top of the priority list. Another point is\nthat the activities include topics that do not all come under a single project\nstructure.\nCsanady. In connection with either a long-term or a short-term approach, I\nwant to call attention to a common failing of these programs. When people come\ninto them and have to design a program on short lead time, an early decision\nis often made that state-of-the-art instrumentation will be used instead of\ntrying to solve the problem the best possible way. Let's take things off the\nshelf, put them in the lake, and see what we get. This works sometimes, but\nit would make much more sense to allow enough time to develop the most sophisti-\ncated instrumentation techniques this age is capable of.\nAubert. The Water Movements panel report alluded to that. This means more lead\ntime is required before deployment of any new major instrumentation or data\ncollection system.\nBeeton. I do not see how NOAA can develop an effective program in the Great\nLakes without getting a handle on what is going on in the nearshore zone so\nthat you can logically plan a longer term program. It would be very wise,\n123","whether you call it a small IFYGL or something else, to have a project like this\nwith one spin-off being the definition of the long-term program that you should\nbegin developing.\nSloss. It seems then that one of the first priorities would be to define what\nis the nearshore zone for a particular process. Some of them may extend farther\nout than others, and it would be a logical type of pilot study to see just what\nkind of an area must be looked at.\nScott. We have now heard the individual groups, but not too many people \"float-\ned\" between them as was planned. Some mechanism should perhaps be developed\nwhere the right scales are used for the processes that were mentioned. I talked\nto Don McNaught and Fred Lee and they complained that the wrong scales were\nused. The first station was out 1 km from shore; a lot more detail was needed\nnear the shore. If the physical people set up a station pattern without con-\nsulting other interested parties, as was apparently done in some cases in IFYGL,\nI think we are again making a mistake. Maybe GLERL will be the mechanism for\nthis needed intergroup cooperation.\nBirchfield. I am not sure how this applies to all of the panels, but it seems\nto me that one must always make a conscious effort to make planning more than\nsimple instrument development. The tendency in IFYGL was to do only the latter.\nAs someone mentioned yesterday, the Mid-Ocean Dynamic Experiment had a workshop\nlasting an entire summer in which to develop dynamical models that would act\nas a focal point for ideas on how to gather the observations for that experiment.\nIt seems to me that, since the focus is on the coastal zone here, some sort of\ndevelopment of dynamical models should be started right away in that area, for\nexample, numerical models. There is none now.\nAubert. This could be lumped under experimental design, I presume.\nMonahan. To do it a little more evenhandedly, I think evolution of the models\nshould go hand-in-hand with the evolution of the instrumentation, and needless\nto say, both of them should precede the actual field data collection.\nAubert. Instrument development and numerical model development both need to\ncome early in order to be available when they are needed.\nCsanady. The way they handled it in the Mid-Ocean Dynamic Experiment was with\na scientific council. Maybe you should establish such a council if you decide\nthis Laboratory will support such an effort. One good and relatively cheap way\nto start is to set up a scientific council and maybe have a workshop.\nAubert. That is a mechanism that should certainly be considered.\nComment. I think we need a more specific mechanism to learn about the IFYGL\nresults. No one has mentioned how they are going to feed into any of this\nactivity. What are the plans for scientific discussion of IFYGL results?\nAubert. I think what we are talking about here is more of a scientist-to-\nscientist interaction. I will comment as to what now exists. The IFYGL pro-\ngram is divided into panels. Lloyd Richards is the Canadian co-chairman and I\nam the United States co-chairman of the Joint Management Team. We are meeting\n124","the latter part of this month to review the status of the outlines for the\nfinal International Scientific Reports from each IFYGL panel, which should be\npublished within the next 1 to 3 years, then to identify potential delays,\nand finally to resolve these delays to insure that the Scientific Reports are\nprepared. That perhaps is a narrow objective from the overall point of view,\nbut these reports are considered to be the final IFYGL product. The Inter-\nnational Scientific Reports will summarize all of the hundreds of articles and\nreports that will have been published in scientific journals and the reports\nof various agencies and institutions. The scientist-to-scientist interaction\nin the IFYGL plan is within the panels, with the panel co-chairmen defining\nand producing the scientific reports. Lloyd Richards and I have to insure that\nworkable plans are developed by the panel co-chairmen consistent with all of\nthe conflicting constraints from other programs. What you are suggesting goes\nbeyond our plans.\nCsanady. One problem in IFYGL was cross-panel communications. Those panels\nhave large walls around them.\nAubert. Cross-panel meetings need a defined context. Lloyd Richards and I are\npushing from the context of the International Scientific Report series, the\nproduct which will wrap up IFYGL in 3 years. We believe it is important, and\nif we do not give it continuing attention, it will never happen.\nMortimer. Another product should be a data catalog.\nAubert. That is already included in the plan. People are working on it. It\nis just a matter of time before the archive will be generated and an archive\ncatalog will be available.\nMortimer. I think there should be a final workshop meeting someday, perhaps at\nthe same time as an IAGLR conference. A whole day can be spent on IFYGL wrap-\nup. It would be helpful in about 2 years time.\nAubert. An IFYGL symposium was held last April in Washington at the American\nGeophysical Union meeting. We had 1/2 day and 11 invited papers. At the IAGLR\nconference in August 1974, IFYGL had 54 individual presentations. More papers\nwere submitted, I think. Most of these papers will be in a special proceedings\ndue to be published in the spring of 1975, but that still is not the interaction\nyou are talking of. Somebody gives a formal presentation followed by a minute\nof discussion. Of course, one can then seek the individual out for personal\ninteraction. Do you have something to suggest?\nComment. One thing that would be helpful would be for someone to write a review\npaper on all the publications.\nAubert. One of the final International Scientific Reports will be an overview\nof IFYGL. Lloyd Richards and I are listed as co-authors, but we may get help.\nThis final International Scientific Report will not be published for 3 years.\nCsanady. Another point that I already raised in a group meeting was that, when\nyou get these data availability catalogs, it would be helpful if an individual\nset of data was called something other than GS1500 235MB.\n125","Aubert. If you cannot learn the system, you will have to ask for assistance\nfrom an expert.\nHolland. You do not have to learn the system; just take the listing, find\nwhat you want in it, and ask for it. The filing system must have some such\nmethodical labeling system, and you have to give the person who works in the\nfiles the identification he needs to retrieve the item you want. He has a\ncatalog. If you give him the name, he has to look up the number.\nSchelske. Before the meeting breaks up, there is one thing I would like to make\na statement on. I am a little concerned about some of the items that appear on\nthis list, in particular, zero pollution discharge. That has broad implications.\nA lot of people have worked very hard to get that kind of law on the books. We\nalso talk about energy management. These are almost philosophical questions.\nIf we are going to do that, it is fine with me, but I think we also ought to\nextend that list and include items like no-growth policies and zero population\ngrowth. All those are related. There have been tremendous advances in terms\nof controlling pollution from this one law on zero pollution discharge, and if\nwe now say this is a scientific question that has to be studied, that is\nobvious. But on the other hand, we are making an issue out of something that is\nalmost philosophical.\nAubert. I think a rebuttal from the Environmental Dynamics panel chairman is\nneeded here.\nMortimer. These are examples of pressing national questions or questions some\nway down the road for which this Laboratory will provide part of the scientific\nbasis for rational decisions. I think the important thing to stress here is\nthat the Laboratory should not express opinions on environmental politics, but\nshould provide a sound scientific basis for rational decisions, if such are\npossible. The zero pollution law, as defined, involves a decision, or so I am\ninformed by a panel member, on what is socially desirable and what is techni-\ncally practical. There will be tradeoffs between zero pollution, which is,\nof course, unattainable because you have diffuse sources as well as point\nsources, and what is socially desirable. For wise decisions on pollution\ncontrol, or wise decisions on the use of the Great Lakes as heat sinks,\na sound scientific basis is needed. These points were raised by the panel only\nas examples of questions for which a sound scientific basis is badly needed.\nSchelske. I would agree with that, but my criticism then is, why do you dis-\nregard zero population growth?\nMortimer. Because no panel member raised that particular question.\nAubert. Thank you for attending this workshop. I hope you got as much out of\nit as we did. All who attended will get a copy of the transcribed tapes after\nreview by the principal speakers and responders.\n126","D. E. Meredith, Chmn.\nF. Quinn, Sec.\nL. Bajorunas\nJ. Derecki\nWATER LEVELS\nAND FLOWS\nR. Assel\nL. Baer\nE. Peck\nC. Mortimer, Chmn.\nB. Eadie, Sec.\nC. Herdendorf\nC. Gunnerson\nR. Ragotzkie\nENVIRONMENTAL\nD. Chandler\nE. Aubert\nG. F. Lee\nA. Pinsak\nDYNAMICS\nW. Carey\nW. Hess\nJ. Judd\nR. Kolf\nA. Lane\nAPPENDIX A. WORKSHOP GROUP MEMBERSHIP\nS. Bolsenga, Sec.\nM. Estoque, Chmn.\nLAKE-ATMOSPHERE\nC. F. Jenkins\nINTERACTIONS\nJ. Fletcher\nA. W. Green\nD. Houghton\nT. Donahue\nC. Emanuel\nJ. Holland\nC. Snider\nM. Mull\nP. Liu\nC. Schelske, Chmn.\nS. Chapra, Sec.\nAND WATER QUALITY\nH. Vanderploeg\nAQUATIC ECOLOGY\nS. Tarapchak\nE. Stoermer\nJ. Schmidt\nR. Sweeney\nR. Thomann\nA. Beeton\nD. Friis\nJ. Spain\nA. Vogel\nC. Chen\nG. Csanady, Chmn.\nWATER MOVEMENTS\nG. E. Birchfield\nR. Pickett, Sec.\nG. K. Rodgers\nJ. Bennett\nE. Monahan\nD. B. Rao\nJ. Saylor\nT. Green\nJ. Scott\nP. Sloss","APPENDIX B. LIST OF WORKSHOP ATTENDEES\nJohn Armstrong\nBrian Eadie\nUniversity of Michigan\nGLERL\nAnn Arbor, MI\nAnn Arbor, MI\nRaymond Assel\nConstantinos Emanuel\nGLERL\nEnvironmental Research Laboratories\nAnn Arbor, MI\nBoulder, CO\nEugene Aubert\nMariano Estoque\nGLERL\nUniversity of Miami\nAnn Arbor, MI\nMiami, FL\nLedolph Baer\nJoseph Fletcher\nEnvironmental Monitoring and Prediction\nEnvironmental Research Laboratories\nRockville, MD\nBoulder, CO\nLeonas Bajorunas\nDavid Friis\nGLERL\nEnvironmental Research Laboratories\nAnn Arbor, MI\nBoulder, CO\nAlbert Beeton\nBert Green\nUniversity of Wisconsin-Milwaukee\nUniversity of Michigan\nMilwaukee, WI\nAnn Arbor, MI\nJohn Bennett\nTed Green\nMassachusetts Institute of Technology\nUniversity of Wisconsin\nCambridge, MA\nMadison, WI\nEd Birchfield\nDavid Gregorka\nNorthwestern University\nGreat Lakes Basin Commission\nEvanston, IL\nAnn Arbor, MI\nStanley Bolsenga\nCharles Gunnerson\nGLERL\nEnvironmental Research Laboratories\nAnn Arbor, MI\nBoulder, CO\nWalter Carey\nCharles Herdendorf\nOhio State University\nOhio State University\nColumbus, OH\nColumbus, OH\nDavid Chandler\nWilmot Hess\nOhio State University\nEnvironmental Research Laboratories\nColumbus, OH\nBoulder, CO\nSteven Chapra\nJoshua Holland\nGLERL\nEnvironmental Data Service\nAnn Arbor, MI\nWashington, DC\nCarl Chen\nDavid Houghton\nTetra Tech, Incorporated\nUniversity of Wisconsin\nLafayette, CA\nMadison, WI\nGabe Csanady\nC. Fred Jenkins\nWoods Hole Oceanographic Institute\nGLERL\nWoods Hole, MA\nAnn Arbor, MI\nJan Derecki\nJohn Judd\nGLERL\nState University of New York-Albany\nAnn Arbor, MI\nAlbany, NY\nRichard Kolf\nThomas Donahue\nUniversity of Michigan\nOffice of Sea Grant\nWashington, D.C.\nAnn Arbor, MI\n128","APPENDIX B. LIST OF WORKSHOP ATTENDEES (CONTINUED)\nAmor Lane\nJames Saylor\nMarine Resources\nGLERL\nRockville, MD\nAnn Arbor, MI\nG. Fred Lee\nClaire Schelske\nUniversity of Texas-Dallas\nUniversity of Michigan\nDallas, TX\nAnn Arbor, MI\nPaul Liu\nJim Schmidt\nGLERL\nCanada Centre for Inland Waters\nAnn Arbor, MI\nBurlington, Ontario\nDale Meredith\nJon Scott\nState University of New York-Buffalo\nState University of New York-Albany\nBuffalo, NY\nAlbany, NY\nEd Monahan\nPeter Sloss\nUniversity of Michigan\nGLERL\nAnn Arbor, MI\nAnn Arbor, MI\nClifford Mortimer\nCharles Snider\nUniversity of Wisconsin-Milwaukee\nNational Weather Service\nMilwaukee, WI\nDetroit, MI\nMax Mull\nJames Spain\nNational Weather Service\nMichigan Technological University\nSilver Spring, MD\nHoughton, MI\nEugene Peck\nEugene Stoermer\nNational Weather Service\nUniversity of Michigan\nSilver Spring, MD\nAnn Arbor, MI\nRobert Pickett\nRobert Sweeney\nGLERL\nState University of New York-Buffalo\nAnn Arbor, MI\nBuffalo, NY\nArthur Pinsak\nStephen Tarapchak\nGLERL\nGLERL\nAnn Arbor, MI\nAnn Arbor, MI\nFrank Quinn\nRobert Thomann\nGLERL\nManhattan College\nAnn Arbor, MI\nNew York, NY\nRobert Ragotzkie\nRoss Tocher\nUniversity of Wisconsin\nUniversity of Michigan\nMadison, WI\nAnn Arbor, MI\nD. B. Rao\nHenry Vanderploeg\nUniversity of Wisconsin\nGLERL\nMadison, WI\nAnn Arbor, MI\nG. Keith Rodgers\nAlan Vogel\nCanada Centre for Inland Waters\nUniversity of Michigan\nBurlington, Ontario\nAnn Arbor, MI\n129"]}