{"Bibliographic":{"Title":"Evaluation of a remote weather radar display : final report.","Authors":"","Publication date":"1976","Publisher":""},"Administrative":{"Date created":"08-16-2023","Language":"English","Rights":"CC 0","Size":"0000167146"},"Pages":["TL\n521\n.A256\n: No. FAA-RD-75-60, II\nno.75-60\nv.2\nEVALUATION OF A REMOTE WEATHER RADAR DISPLAY\nVol. II - Computer Applications for\nStorm Tracking and Warning\nW. David Zittel\nDEPARTMENT OF OF\nWILL STATES OF Authorized\nLIBRARY.\nFinal Report\nRecataloged\nMAY 1 91994\nDecember 1976\nO.A.A.\nU.S. Dept of Commerce\nDocument is available to the U.S. public through\nthe National Technical Information Service,\nSpringfield, Virginia 22161.\nPrepared for\nU.S. DEPARTMENT OF TRANSPORTATION\nFEDERAL AVIATION ADMINISTRATION\nSystems Research & Development Service\nWashington, D.C. 20590","ATMOSPHERIC\nNOAA\ncomments\nDEPARTMENT\nOF\nNOTICE\nThis document is disseminated under the sponsorship of\nP","521\nA256\nno.75-\nTechnical Report Documentation Page\n60\n1. Report No.\n2. Government Accession No.\n3. Recipient's Catalog No.\nV.2\nFAA-RD-75-60, II\n4. Title and Subtitle\n5. Report Date\nDecember 1976\nEVALUATION OF A REMOTE WEATHER RADAR DISPLAY\nVol. II - Computer Applications for Storm\n6. Performing Organization Code\nTracking and Warning\n8. Performing Organization Report No.\n7. Author(s)\nW. David Zittel\n9. Performing Organization Name and Address\n10. Work Unit No. (TRAIS)\nNational Severe Storms Laboratory\nEnvironmental Research Laboratories\n11. Contract or Grant No.\nDOT FA74WAI-440\nNational Oceanic & Atmospheric Administration\n13. Type of Report and Period Covered\n12. Sponsoring Agency Name and Address\nPhase II\nDepartment of Transportation\nMay 1975 to December 1976\nFederal Aviation Administration\nSystems Research and Development Service\n14. Sponsoring Agency Code\nARD-451\nWashington, D. C. 20590\n15. Supplementary Notes\n16. Abstract\nTests with a Remote Radar Display using R, O coordinates had demonstrated\nthe feasibility of presenting timely warnings for areas in severe storm\npaths graphically. Mathematical derivations and three case studies\nutilizing operationally oriented software are presented with suggestions\nfor future work.\nThis research is divided in two parts. Volume I describes the installation\nof equipment, training of FSS briefers in operation and interpretation, and\ndocumentation of data transmitted during the tests. Volume II summarizes\nresearch efforts at NSSL to computerize further processing of the displayed\ndata for rapid isolation and tracking of selected storms.\n17. Key Words\n18. Distribution Statement\nDocument is available to the public\nWeather radar display\nthrough the National Technical Informa-\nGraphic severe storm warnings\ntion Service, Springfield, VA 22151.\nAutomated computerized weather\nradar data processing\n19. Security Classif. (of this report)\n20. Security Classif. (of this page)\n21. No. of Pages\n22. Price\n129\nUNCLASSIFIED\nUNCLASSIFIED\nForm DOT F 1700.7 (8-72)\nReproduction of completed page authorized iii","°F\nfl oz\nyd2\nmi²\noz\nSymbol\nyd\nmi\nin2\nlb\nin\nin\nyd3\nft\ngal\nft3\npt\nqt\n100\n212\nC\nF\no\n200\ntemperature\nsquare inches\nsquare yards\nsquare miles\nfluid ounces\ncubic yards\ncubic feet\nshort tons\nFahrenheit\nApproximate Conversions from Metric Measures\n80\ngallons\nounces\npounds\nTo Find\nquarts\ninches\ninches\nacres\nyards\nmiles\npints\n160\nfeet\n60\nTEMPERATURE (exact)\n120\nMASS (weight)\n40\n98.6\nMultiply by\nVOLUME\n9/5 (then\n37\nLENGTH\nadd 32)\nAREA\n0.035\n0.16\n0.03\n1.06\n0.26\n0.04\n1.2\n0.4\n2.5\n1.3\n2.2\n1.1\n2.1\n0.4\n3.3\n80\n1.1\n0.6\n35\n20\nhectares (10,000 m²\nsquare centimeters\n40\nsquare kilometers\ntonnes (1000 kg)\n32\ntemperature\no\nsquare meters\nWhen You Know\ncubic meters\ncubic meters\ncentimeters\nmillimeters\nkilometers\nmilliliters\nkilograms\nCelsius\nmeters\nmeters\nO\ngrams\n-20\nliters\nliters\nliters\n-40\n-40\n°C\n°F\nSymbol\ncm2\nkm2\nmm\nm3\n'3\nC\ncm\nkm\nm2\nha\nkg\nml\nm\nm\ng\nt\nI\nI\nI\nMETRIC CONVERSION FACTORS\n10\n6\n3\ncm\n6\n5\n4\n3\n2\n9\nSymbol\ncm2\nkm2\n°C\nm³\nm³\ncm\ncm\nkm\nkg\nm²\nml\nml\nml\nm²\nha\n*1 in = 2.54 (exactly). For other exact conversions and more detailed tables, see NBS Misc. Publ. 286,\nm\ng\nt\nI\nI\nI\nsquare centimeters\nsquare kilometers\ntemperature\nsquare meters\nsquare meters\ncubic meters\ncubic meters\ncentimeters\ncentimeters\nkilometers\nmilliliters\nmilliliters\nmilliliters\nkilograms\nhectares\nApproximate Conversions to Metric Measures\nCelsius\nTo Find\ntonnes\nmeters\ngrams\nliters\nliters\nliters\nliters\nUnits of Weights and Measures, Price $2.25, SD Catalog No. C13.10:286.\nTEMPERATURE (exact)\nMultiply by\nMASS (weight)\nsubtracting\n5/9 (after\nVOLUME\n0.45\n0.24\n0.47\n0.95\n0.03\n0.76\n0.09\nLENGTH\n3.8\n0.9\n*2.5\n0.9\n1.6\n6.5\n0.8\n2.6\n0.4\nAREA\n28\n5\n15\n30\n30\n32)\nsquare inches\ntemperature\nsquare yards\nsquare miles\nfluid ounces\n(2000 lb)\ntablespoons\ncubic yards\nWhen You Know\nsquare feet\nshort tons\nteaspoons\ncubic feet\nFahrenheit\ngallons\nounces\npounds\ninches\nquarts\nyards\nmiles\nacres\npints\ncups\nfeet\nSymbol\nTbsp\nfl oz\nyd2\nmi2\ntsp\ngal\nyd3\nin2\nft3\nft2\nqt\nyd\nmi\noz\npt\nin\nlb\nft\nc\n°F","FOREWORD\nThe Federal Aviation Administration and the National\nSevere Storms Laboratory are cooperating in search of\nimproved methods for severe storm prediction and warning\nfor aviation. Here NSSL Operations staff reports on tests\ninvolving transmission of contour-mapped WSR-57 weather\nradar from NSSL headquarters to a display unit at the Okla-\nhoma City Flight Service Station.\nOur study follows other investigations of the com-\nparative value of various radar systems for severe storm\nsurveillance. Improved signal processing and communication\ntechniques and equipment now permit rapid dissemination of\ninformation concerning storm location, intensity, and move-\nment and offer a new dimension in weather display for\ngeneral aviation.\nATMOSPHERIC SCIENCES\nLIBRARY\nJUL 18 1977\nN.O.A.A.\nU. S. Dept. of Commerce\nV\n77\n2192","","TABLE OF CONTENTS\nPage\nFOREWORD\nV\nLIST OF FIGURES\nviii\nLIST OF TABLES\nxiii\nLIST OF SYMBOLS\nxiv\n1.\nINTRODUCTION\n1\n2.\nBACKGROUND\n1\n3.\nRATIONALE FOR SELECTING AND TRACKING ECHO CENTROID\n2\n(OR WHY LEAVE A PERSON IN THE PICTURE)\n4.\nMATHEMATICAL FORMULATION\n4\n4.1\nIntroduction\n4\n4.2 Echo Shape and Centroid Calculation\n4\n4.3 Echo Motion Calculation\n7\n4.4 Warning Area Calculation\n8\n5.\nREAL TIME SYSTEM AND PROGRAM OPERATION\n14\n5.1 Hypothetical Systems Configuration\n14\n5.2 Hypothetical Software Logic\n14\n5.3 Some Practical Guidelines for Using Echo\n21\nPrediction Software\n5.3.1 Time Weight Constant (TWC)\n21\n5.3.2 Selection of Area and Intensity Criteria\n25\n5.3.3 Splits and Mergers\n26\n6.\nTEST CASES\n26\n7.\nSUMMARY, CONCLUSIONS AND RECOMMENDATIONS\n53\nACKNOWLEDGMENTS\n56\nREFERENCES\n56\nAPPENDIX A: Annotated Command Structure, June 6, 1975\n57\nAPPENDIX B: Annotated Command Structure, June 16, 1975\n67\nAPPENDIX C: Annotated Command Structure, November 2, 1974\n83\nAPPENDIX D: Computer Program Listing for ECHOPRED\n95\nvii","LIST OF FIGURES\nFigure\nPage\nWSR-57 radar PPI, 100 n mi. range, 20 n mi. range\n1.\n3\nmarks, 1454 CST, April 3, 1964.\nWSR-57 radar PPI, 200 km range, 40 km range\n2.\n3\nmarks, 1132 CST, September 24, 1974. Light to\nmoderate stratiform showers indicated over most\nof the radar scope.\nRepresentative echo samples show the path and order\n4\n3.\nof points to describe echo perimeter.\n4.\nGraphic presentation of error in determining the\n6\necho centroid due to non-uniform perimeter density\nfor a hypothetical circular echo with a 10 km radius.\n5.\nIllustration of use of an approximating ellipse,\n9\nlinear least squares predicted echo trajectory and\nthe RMSE values to calculate graphic warning area\nfor w > 10 degrees.\n6.\nIllustration of use of an approximating ellipse,\n13\nlinear least squares predicted echo trajectory and\nthe RMSE values to calculate graphic warning area\nfor W 10 degrees.\n7.\nSchematic of systems flow chart.\n\"15\n8.\nIllustration showing various commands used in\n16-17\nECHOPRED.\n9.\nIllustration of cone used to locate airports in\n20\nan echo's path.\n10.\nIllustration of response curve for weighting\n22\nfunction, W.\n11.\nIllustration of the effect of varying the TWC for\n2\nan echo whose track is curved in calculating\ndirection, speed, Od' and ot.\n12.\nIllustration of the angular change in centroid\n26\nposition as a function of range.\nJune 6, 1975, 12Z surface analysis adapted from\n13.\n28\nDOC, NOAA, EDS daily weather maps, Weekly Series.\n14.\nJune 6, 1975, 12Z surface analysis adapted from\n28\nDOC, NOAA, EDS daily weather maps, Weekly Series.\nviii","LIST OF FIGURES (cont.)\nFigure\nPage\n15.\nSubsynoptic surface data provided by NWS, FAA\n29\nand NSSL stations, 2000 CST, June 6, 1975, with\nstreamlines superimposed.\n16.\nDivergence field derived from surface data,\n29\n2000 CST, June 6, 1975.\n17.\nWSR-57 radar PPI, 400 km range, 100 km range\n30\nmarks, 1600 CST, June 6, 1975.\n18.\nRemote radar display PPI, 200 km range, 2040 CST\n32\nJune 6, 1975.\n19.\nRemote radar display with computer generated\n32\nwarning areas, echo contours and Victor airways,\nJune 6, 1975. Prediction period is 2110 to\n2200 CST.\n20.\nRemote radar display with computer generated\n32\nwarning areas, echo contours and Victor airways,\nJune 6, 1975. Prediction period is 2110 to\n2210 CST.\n21.\nRemote radar display PPI, 200 km range, 2105 CST,\n33\nJune 6, 1975.\n22.\nRemote radar display PPI, 200 km range, 2115 CST,\n33\nJune 6, 1975.\n23.\nRemote radar display with computer generated\n33\nwarning areas, echo contours and Victor airways,\nJune 6, 1975. Prediction period is 2130 to\n2230 CST.\n24.\nRemote radar display PPI, 200 km range, 2120 CST,\n33\nJune 6, 1975.\n25.\nRemote radar display with computer generated\n34\nwarning areas, echo contours and Victor airways,\nJune 6, 1975. Prediction period is 2135 to\n2235 CST.\n26.\nRemote radar display PPI, 200 km range, 2130 CST,\n34\nJune 6, 1975.\n27.\nRemote radar display with computer generated\n34\nwarning areas, echo contours and Oklahoma State\noutline, 400 km range, June 6, 1975. Prediction\ntime is 2145 to 2245.\nix","LIST OF FIGURES (cont.)\nPage\nFigure\n34\nWSR-57 radar PPI, 200 km range, 40 km range\n28.\nmarks, 2249 CST, June 6, 1975.\n35\nJune 16, 1975, 12Z surface analysis adapted from\n29.\nDOC, NOAA, EDS daily weather maps, Weekly Series.\n35\nJune 16, 1975, 12Z 500 mb analysis adapted from\n30.\nDOC, NOAA, EDS daily weather maps, Weekly Series.\n36\nSMS-3 Satellite pictures, June 16, 1975.\n31.\n37\nWSR-57 radar PPI, 400 km range with 200 km range\n32.\ndelay to first gate, 40 km range marks,\nJune 16, 1975.\n38\nSubsynoptic surface data provided by NWS, FAA,\n33.\nand NSSL stations, 1800 CST, June 16, 1975, with\nstreamlines superimposed.\n38\nDivergence field derived from surface data,\n34.\n1800 CST, June 16, 1975.\n39\nRemote radar display PPI, 200 km range, 1830 CST\n35.\nJune 16, 1975.\n39\nRemote radar display PPI, 200 km range, 1835 CST\n36.\nJune 16, 1975.\n39\nRemote radar display with computer generated\n37.\nwarning area, echo contour and Victor airways,\nJune 16, 1975. Prediction period is 1835 to\n1935 CST.\n41\nRemote radar display PPI, 200 km range, 1900 CST,\n38.\nJune 16, 1975.\n41\nRemote radar display with computer generated\n39.\nwarning area, echo contour and Victor airways,\nJune 16, 1975. Prediction period is 1900 to\n2000 CST.\n42\nRemote radar display PPI, 200 km range, 1915 CST,\n40.\nJune 16, 1975.\n42\nRemote radar display with computer generated\n41.\nwarning areas, echo contours and Victor airways,\nJune 16, 1975. Prediction period is 1915 to\n2015 CST.\nX","LIST OF FIGURES (cont.)\nFigure\nPage\n42.\nRemote radar display PPI, 200 km range, 1920 CST,\n42\nJune 16, 1975.\nRemote radar display PPI, 200 km range, 1930 CST\n42\n43.\nJune 16, 1975.\n44.\nRemote radar display with computer generated\n43\nwarning area, echo contour and Victor airways,\nJune 16, 1975. Prediction period is 1930 to 2030 CST.\n45.\nRemote radar display PPI, 200 km range, 1945 CST,\n43\nJune 16, 1975.\nRemote radar display with computer generated\n43\n46.\nwarning area, echo contour and Victor airways,\nJune 16, 1975. Prediction period is 1945 to\n2045 CST.\nRemote radar display PPI, 200 km range, 2000 CST,\n43\n47.\nJune 16, 1975.\n48.\nRemote radar display with computer generated\n44\nwarning area, echo contour and Victor airways,\nJune 16, 1975. Prediction period is 2000 to\n2100 CST.\nRemote radar display PPI, 200 km range, 2015 CST,\n45\n49.\nJune 16, 1975.\nRemote radar display with computer generated\n45\n50.\nwarning area, echo contour and Victor airways,\nJune 16, 1975. Prediction period is 2015 to\n2115 CST.\nRemote radar display PPI, 200 km range, 2030 CST\n45\n51.\nJune 16, 1975.\nRemote radar display with computer generated\n45\n52.\nwarning area, echo contour and Victor airways,\nJune 16, 1975. Prediction period is 2030 to 2130 CST.\nNovember 1, 1974, 12Z surface analysis adapted\n46\n53.\nfrom DOC, NOAA, EDS daily weather maps, Weekly\nSeries.\nNovember 2, 1974, 12Z surface analysis adapted\n46\n54.\nfrom DOC, NOAA, EDS daily_ weather maps, Weekly\nSeries\nxi","LIST OF FIGURES (cont.)\nFigure\nPage\n55.\nNovember 3, 1974, 12Z surface analysis adapted\n46\nfrom DOC, NOAA, EDS daily weather maps, Weekly\nSeries.\n56.\nNovember 1, 1974, 12Z 500 mb analysis adapted\n47\nfrom DOC, NOAA, EDS daily weather maps, Weekly\nSeries.\n57.\nNovember 2, 1974, 12Z 500 mb analysis adapted\n47\nfrom DOC, NOAA, EDS daily weather maps, Weekly\nSeries.\n58.\nNovember 3, 1974, 12Z 500 mb analysis adapted\n47\nfrom DOC, NOAA, EDS daily weather maps, Weekly\nSeries.\n59.\nSubsynoptic surface data provided by NWS, FAA,\n48\nand NSSL stations, 1200Z. November 2, 1974,\nwith streamlines superimposed.\n60.\nRemote radar display PPI, 200 km range, 1225Z,\n50\nNovember 2, 1974.\n61.\nRemote radar display PPI, 200 km range, 1230Z,\n50\nNovember 2, 1974.\n62.\nRemote radar display PPI, 200 km range, 1240Z,\n50\nNovember 2, 1974.\n63.\nRemote radar display with computer generated\n50\nwarning areas, echo contours and Victor airways,\nNovember 2, 1974. Prediction period is 1240 to\n1340Z.\n64.\nRemote radar display PPI, 200 km range, 1251Z,\n51\nNovember 2, 1974.\n65.\nRemote radar display with computer generated\n51\nwarning areas, echo contours and Victor airways,\nNovember 2, 1974. Prediction period is 1251 to\n1351Z.\n66.\nRemote radar display PPI, 200 km range, 1300Z,\n51\nNovember 2, 1974.\nxii","LIST OF FIGURES (cont.)\nFigure\nPage\n67.\nRemote radar display with computer generated\n51\nwarning areas, echo contours and Victor airways,\nNovember 2, 1974. Prediction period is 1300 to\n1400Z.\n68.\nRemote radar display PPI, 200 km range, 1310Z,\n52\nNovember 2, 1974.\n69.\nRemote radar display with computer generated\n52\nwarning areas, echo contours and Victor airways,\nNovember 2, 1974. Prediction period is 1310 to\n1410Z.\nLIST OF TABLES\nTable\nPage\n1.\nComparison of results of using a TWC of both\n23\n30 and 5 minutes on data sampled at both\n2-3 minutes and 4-6 minutes for tracking two\nechoes.\n2.\nEcho centroid positions, direction and speed\n31\npredicted, and the RMSE values for June 6, 1975.\n3.\nEcho centroid positions, direction and speed\n40\npredicted, and the RMSE values for June 16, 1975.\n4.\nEcho centroid positions, direction and speed\n49\npredicted, and the RMSE values for November 2, 1974.\nxiii","LIST OF SYMBOLS\nFourier coefficient\na\nsubscript denoting arrival time\nFourier coefficient of nth harmonic\na 'n\ncoefficient of general equation of ellipse\nA\n1\ncoefficient of raotated ellipse\nA\ncoefficient of linear least squares\nA\nX\nequation for X (t)\ncoefficient of linear least squares\nAy\nequation for Y(t)\nFourier coefficient\nb\nsubscript denoting beginning point time\nFourier coefficient of nth harmonic\nb n\ncoefficient of general equation of ellipse\nB\ncoefficient of linear least squares\nBx\nequation for X(t)\ncoefficient of linear least squares\nB.\ny\nequation for Y(t)\nFourier coefficient\nC\nFourier coefficient of nth harmonic\ncn\ncoefficient of general equation of ellipse\nC\n1\ncoefficient of rotated ellipse\nC\nFourier coefficient\nd\nsubscript denoting distance\nFourier coefficient of nth harmonic\nd\nn\nbase of natural logarithms\ne\nsubscript denoting ending point, time\ngatelength\nG\nindex counter\ni\nxiv","LIST OF SYMBOLS (cont.)\nk\ntime constant in weighting function\nK\nconstant of general second order equation\nfor ellipse\nl\nsubscript denoting last point, time\nperimeter of echo for arc length function\nL\nM\ntotal number of discrete line segments to\ndescribe echo perimeter\nslope of straight line\ndenotes number of harmonic\nn\nnumber of centroid entries for echo tracking\nN\nnumber of points , used in LLS equation\ncoefficients of linear line for discrete\narc lengths\nP b\nbeginning points along the echo path for\nwarning area\nP\nending point along echo path for warning area\ne\nP\nlast centroid position entered\no\nR\nrange between two points\nR1R2\ndistance to two consecutive gates\narc length parameter\nS\ndiscrete arc length\nS\ni\ntime\nt\nbeginning time\ntb\nending time\nt e\ntime of ith centroid entry\nt\ni\ntime of last centroid entry\nt Fl\nt\nsame as tl\nn\nweighting parameter\nW\nXV","LIST OF SYMBOLS (cont.)\nsubscript for coefficients of linear least\nX\nsquares equation\nindependent variable\nX\nlast position of X(t)\nX\nl\nparametric function of arc length for X\nX (s)\nparametric function of linear least squares\nX(t)\nequation for X\n(X,Y), (x' Y')\ndenotes points in Cartesian coordinates\nslope of echo track\nXM\nsubscript for coefficients of linear least\ny\nsquares equation\ndependent variable\nY\nlast position of Y(t)\nY\nl\nparametric function of arc length for Y\nY (s)\nY (t)\nparametric function of linear least squares\nequation for Y\nAa\nangular change\nAt\nchange in time\ndistance error in echo tracking\nE d\ntime error in echo tracking\nE\nt\nangular difference between consecutive radials\nO\nRMS estimates of distance error\no\nd\nRMS estimates of time error\no t\nsmallest angle to rotate approximating\nW\nellipse to eliminate cross product terms\nxvi","EVALUATION OF A REMOTE WEATHER RADAR DISPLAY\nVOLUME II\nComputer Applications and Techniques for Storm Tracking and Warning\nW. David Zittel\n1. INTRODUCTION\nThis report extends tests of the remote radar display described in\nVolume 1 and examines the feasibility of the display as a graphics termi-\nnal. A storm tracking program has been combined with an echo contouring\nscheme to produce graphic warning areas based on size and motion of storm\necho areas.\nA detailed description is provided of the mathematical techniques\nemployed and a quasi-real time software program is outlined. In addi-\ntion, three case studies utilizing the above logic are presented.\nFinally, a summary of results and suggestions for improvements and future\nwork are discussed.\n2. BACKGROUND\nInformation regarding storm motion and growth tendencies are now pre-\nsented to the FSS pilot-briefer in two forms. First, a numerical coding\nin the hourly radar report transmitted by teletype (RAREP) indicates the\npast tendency of storm pattern growth or decay. Motion, in polar coordi-\nnates, of both the pattern and individual cells are included.\nSecondly, a plain language summary provides a \"layman's\" geometric\ndescription of the storms with geographical references to outline the\npresent and projected coverage. Severe Storm Warnings and special advi-\nsories to airmen (SIGMETS and AIRMETS) carry information on hazardous\nflight conditions. At a few locations, these messages are augmented by\nfacsimile machine replica of the Plan-Position Indicator (PPI) with\na\nappropriate annotations (Bigler, 1969).\nThe reliability of both types of advisories varies directly with the\nspatial and temporal variance of radar echo patterns. Information con-\ntained in the RAREP is usually a sterile summary of the radar scope dis-\nplay, condensing details observed and coded during a specific 15 minute\nperiod. Plain language summaries and advisories may include information\non the position and movement of fronts and squall lines, and observations\nof recent severe weather events.\nDuring periods when echo coverage and/or intensity change rapidly\nspecial observations supplement hourly reports. During periods of severe\nweather, the National Weather Service radar scopes are monitored con-\nstantly, but because the flow of information is restricted by communication\n1","facilities and the heavy work load required to meet various local, state\nand national needs, messages to the FSS are periodic.\nThe Volume I tests have shown that if calibrated contoured data are\navailable routinely at the FSS, pilot briefers can interpolate between\nNational Weather Service advisories and maintain a \"user's watch\" of\nstorm locations and intensities. However, neither time nor expertise\nis available at the FSS to relate echo patterns to synoptic scale dis-\nturbances (wind, pressure, and moisture fields) and severe weather\nreports. Even with such data, it is difficult for meteorologists to\npredict changes in gross features of precipitation areas.\nFortunately, large severe storms tend to be steady-state and lend\nthemselves to tracking and extrapolation. The principal objective of\nthis study is to apply semi-automated computerized logic to identify,\ntrack, and extrapolate those storms of sufficient intensity and size\nto produce hazardous weather conditions and to map out a warning area\nfor the extrapolated storm positions.\n3. RATIONALE FOR SELECTING AND TRACKING ECHO CENTROID\n(OR WHY LEAVE A PERSON IN THE PICTURE)\nSeveral years experience in field operations at NSSL have stressed\nthe value of retaining the meteorologist for real time decision making.\nIt seems difficult, if not impossible, to anticipate all the compli-\ncated elements occasionally present in real weather situations, in a\ncomputer program.\nSeveral objective techniques have been suggested by Kessler and\nand Russo (1963), Wilson (1966), and Blackmer and Duda (1972), which\nrely on spatially correlating PPI information to derive echo motion and\nspeed. As shown in Figure 1, significant storms on the radar scope may\nhave quite different motions. Also under certain conditions severe\nstorms split with one portion often moving to the left of the average\nambient wind direction and faster than the wind speed, while the other\nportion moves to the right and slower than the ambient wind (Newton and\nFankhauser, 1964). Under these conditions a single speed and direction\nof motion for the whole scope would be misleading.\nEven if individual echoes are first isolated (Wilk, 1966), there are\nseveral drawbacks to using this approach. In a matrix analysis, a mini-\nmum of two PPI's must be stored at the same time requiring a large com-\nputer core. Generally, the two fields should have uniform grid density\nwhich an R, O system doesn't have. In an R, O system, data must be\nscan-converted to rectilinear coordinates before correlating the data.\nBoth scan conversion and correlation techniques are time consuming.\nCare also must be taken when scan converting to assure that spatial\naveraging has not changed the distribution of intensity integers.\n2","By contrast, the echo centroid extrapolation technique requires\nonly a small amount of core and is very fast. One can operate the\nprogram with radar data extracted manually from the PPI scope display.\nOne may also, in a relatively short period of time, use limited auto-\nmation to scan a PPI for centroid information, and display it regularly\nwithout full time monitoring. (Such a technique is presented in sec-\ntion 5.) Tests during the NSSL Spring Program (Wilk and Gray, 1970)\nindicate an operator can easily filter extraneous or unwanted data.\nSome storms may be moving beyond the radar scope's range while others\nmay be part of a broad band of non-severe stratiform rain whose overall\nmovement is slower and more persistent (fig. 2).\nAn operator may recognize splitting or merging storms which need\nto be treated as new echoes. (Computer programs to date have not\nproved reliable in echo matching and we make no attempt to do this\nhere.)\nOne final reason for leaving an operator in the picture is to\ninsure detection of system failures and to recognize spurious, non-\nmeteorological results when computerized objective analysis software\nsystems are in operation. The following sections are devoted to the\noperation of a man-machine mix using examples of real data sets.\n220\n210\n200\n160\n190\n180\nFigure 1. WSR-57 radar PPI, 100 n mi. Figure 2. WSR-57 radar PPI, 200 km\nrange, 20 n mi. range marks,\nrange, 40 km range marks, 1132 CST,\n1454 CST, April 3, 1964. Indi-\nSeptember 24, 1974. Light to\nvidual echoes are numbered one\nmoderate stratiform showers\nto five.\nindicated over most of the radar\nscope.\n3","4. MATHEMATICAL FORMULATION\n4.1 Introduction\nThree sources of information are used to construct a graphic pres-\nentation of a warning area. In order of calculation, they are (a) echo\ncentroid and shape, (b) echo motion, and (c) a measure of the variance\nof the echo motion. The method used to calculate echo centroid and\nshape basically requires fitting an arc length function to the echo's\nperimeter and was suggested by Blackmer and Duda (1972), , later developed\nby Ostlund (1974). Calculation of echo motion and variance using cen-\ntroid positions was developed by Barclay and Wilk (1970) and run opera-\ntionally during the 1970 Spring Data Collection Program.\n4.2 Echo Shape and Centroid Calculation\nUse of the arc-length function to describe echo shape requires that\none first determine echo perimeter. In the computer logic developed for\nthis report, data are entered into core and all bins with intensity less\nthan a specified level are first set to zero. Then, beginning with zero\ndegrees azimuth, the PPI is searched until an echo is found. Then the\nprogram isolates the echo, moving around the perimeter in a counterclock-\nwise direction until it comes upon the starting point. This logic dif-\nfers from Östlund's in at least two respects. First, the echo's perim-\neter is defined in an R,0 coordinate system; and second, the program\nminimizes echo area. The following two examples in B scan format illus-\ntrate these points.\nIn Figure 3 the arrows indicate\nRANGE (KM)\nthe path followed in the boundary\n21\n22\n23\n24\n25\n26\n27\n28\n29\n30\nsearch. S is the starting gate, X's\n241\nrepresent echo, dots--no echo.\nS\nECHO ONE\nEcho 1 is joined to echo 2 by a\n242\nX\nS\nsingle gate along a common radial.\n243\nThe program ignores that gate since\nECHO TWO\nX\nX\n244\nit would have to be used twice in\n245\n*\norder to close the boundary and com-\nS\n246\n*\nX\nbine echo 1 and 2. Likewise, between\nECHO THREE\necho 2 and echo 3 there is a common\n247\nX\nX\nX\ncorner. But because the corner gates\n248\nX\nwould have to be used twice, the\n249\nechoes are separated.\n250\n251\nNo gate is used twice except the\nstarting one and no gate is accepted\nunless there is an acceptable gate\nFigure 3. Representative echo\nbeyond it. The subroutine BNDRY\nsamples show the path and order\n(called from CONTUR) contains this\nof points to describe echo\nlogic.\nperimeter.\n4","The obvious result of the above, is that cores tend to be discrete\nwith a more regular shape.\nAfter a closed boundary is found, the area of echo is calculated by\nsumming up all bins within and including the perimeter. The area of a\nbin is given by (OTR12 - OTR22)/360°, which can be factored into\nOT(R1 - R2) (R1 + R2)/360°. Since the difference between R1 and R2 is the\ngatelength of the radar, G, and 0 is the angular difference between\nradials, the area of an individual bin is\nOTG (2R1-G)\n(1)\n360°\nIf an echo's area is less than a specified threshold, it is ignored\nand the program searches for a new echo. If an echo exceeds the speci-\nfied area, Fourier analysis of its shape is performed. If the area is\nmore than five times the specified area, the lowest intensity is purged\nand the remaining core treated as a new echo. This process is iterated\nuntil the echo is less than the specified area.\nOnce an echo meets the size criterion, the program enters subroutine\nOSTLND. Here the paired azimuth and range perimeter data are converted\nto Cartesian points. Fourier analysis of X (s) and (s) is performed\nwhere S is the arc length function. Mathematically these functions\nare:\n(2)\n(3)\n+\n.\nbn' 'n' and d may be expressed as\nThe coefficients\ncos (2nT's) ds\n(4)\nsin (2nT's) ds\n(5)\ncos (2nTTS) ds\n(6)\n5","sin ( 2nT S, 2nTs ds\n(7)\nwhere n, an integer, is the nth harmonic.\nEach coefficient may be rewritten:\n(s) cos ( 2nTs, 2nts ds\n(8)\nwhere M is the number of discrete points in the echo's boundary. Also,\nsince X (s) may be considered as consisting of a series of discrete line\nsegments, one may set X(si) = Pi + qidi Setting this expression into\nEq. (8) and integrating yields:\n(9)\nEq. (9) and the corresponding equation for each of the other coefficients\nare calculated in the computer. The 0th harmonic yields the mean of each\nseries and thus the centroid of the echo's shape. Eight harmonics in\naddition to the mean are calculated.\nBecause the values are derived initially from a polar scan, the.\ndensity of points about the perimeter is not constant, biasing the cen- -\ntroid location towards the radar. In a hypothetical case, using a circle\nof 10 km radius, the resulting centroid error varies as a function of\nrange (fig. 4) and is a maximum at 10 km. However, for an echo with\nstable motion there is little error because all centroids have the same\nbias.\n9\n8\nCENTROID RANGE BIAS\nFigure 4. Graphic presentation\n7\nof error in determining the\necho centroid due to non-\n6\nuniform perimeter density for\n5\na hypothetical circular echo\nwith a 10 km radius.\n4\n3\n2\n10km RADIUS\nI\no\no\n20\n40\n60\n80\n100\nRANGE (km)\n6","4.3 Echo Motion Calculation\nThe calculation of echo motion uses linear least squares (LSS) equa-\ntions fitted through an echo's past centroid positions expressed para-\nmetrically as a function of time as\nX(t) = A t + B.\n(10)\ny(t) = A t + B\n(11)\nwhere A and A are found by solving\ny\n(12)\n(t)2\n(13)\nThe ordinate axis intercepts, usually found by solving\n(14)\nand\n(15)\n,\nare here given as\n(16)\nand\n(17)\nwhere tl, Xl and Yl are the echo's last position in time and space. This\ncondition forces the LSS equations through the last point.\nFrom Eqs. (10) and (11) echo speed is simply\n(18)\nand the direction of motion is\nDIR = TAN-1(AXIA) + 180\n(19)\n.\n7","A measure of an echo's predictability in time and distance is\ndetermined by comparing the time of the latest centroid position to the\npredicted time of closest passage to that centroid position. The dif-\nference, At, is defined as the error in time, Et. The distance between\nthe predicted point of closest passage and the actual centroid location\nis defined as\n(20)\nwhere t is the time of closest passage and X and Y are the Cartesian\ncoordinates of the latest centroid position. We can solve for the\nunknown time, t, by first squaring terms in Eq. (20) and then differen-\ntiating them with respect to t yielding:\nX) + .\n(21)\nX -\nSetting the expression on the right equal to zero and solving for t yields\n(22)\nTherefore Et is t - tl where tl is the time of the latest centroid.\nGiven t, Ed can be calculated directly from Eq. (20).\nFinally, and are normalized to one hour and root mean square\nerrors (RMSE) computed from\n(23)\nand\n(24)\nfor n 3, where n is the number of points in a discrete track.\n4.4 Warning Area Calculations\nWe can now combine the results of sections 4.2 and 4.3 to determine\na warning area. Given a beginning time, tb, and ending time, te, we\nfirst solve Eqs. (10) and (11) for the starting and ending points Pb and\nPe, of the warning area specified in the time domain, which lie on the\n8","echo path (fig. . 5). A measure of a storm's predictability in time is\nincluded in determining Pb and Pe as follows:\nFor P.\nb\n(25)\n(26)\ny\nand for P\ne\n(27)\nX\n(28)\n.\nNext, two line segments are found which are parallel to and the\nsame length as the major axis of the echo and which pass through Pb and\nPe, respectively. The length and orientation of the line segments are\ndetermined by finding an ellipse which approximates the echo at hand.\nA parametric form of an ellipse is given by the zeroth and first\nharmonics of the echo. For convenience the ellipse is translated to the\norigin eliminating the zeroth harmonic from further calculations. From\nEqs. (2) and (3) Cartesian coordinates expressed as a function of the\narc length, S, for any point on the ellipse are given by\nCASE FOR w<10°\nAPPROXIMATING\nSAMPLE\nELLIPSE\nECHO\nECHO\nMOTION\nECHO\nMOTION\nR\nPB\nP.\nPO\n/\nTW\nE\nECHO\nof\nof\nMOTION\n%D\nFigure 5. Illustration of use of an approximating ellipse, linear\nleast squares predicted echo trajectory and the RMSE values\nto calculate graphic warning area for W > 10 degrees.\n9","(29)\n=\nY = +\n(30)\nAlso, an ellipse has the second degree form\n= K\n(31)\nwhere A, B, C and K are constants. By combining Eqs. (29) and (30) with\nEq. (31) we can find three equations with which to solve for A, B and C\nfrom which the orientation of the ellipse is found. A value for K is\nspecified below. Setting Eqs. (29) and (30) into (31) yields:\nK si( = +\n(32)\nB\n+\nExpanding and combining like terms gives us:\nK A = 2 + Bac + + Bbd +\n(33)\n+ (2Aab + B(ad + bc) +\nWhen 2TT/S = 0, sin(0°) = 0 and cos (0°) = 1; when 2TTS/L = 2\ncos (/2) = 0 and sin (2) = 1.\nUnder these two conditions Eq. (33) reduces to the following two\nidentities\nBac Cc2\n(34)\nK = Ab2 + bbd + cd2\n(35)\nSince they contain only constants they are valid for all S and K can be\nset into Eq. (33) as:\nK\n=\n(36)\n+ (2Aab + B(ad + bc) + 2cd)\n10","From trigonometry sin2w+cos2==1 = and similarly K sin2w+K c os 2 = K.\nHence, for Eq. (36) to be valid when sin (2TT/S) and cos (2ns/L) both # 0\n(37)\n2Aab + B (ad + bc + 2Ccd) = 0\nmust be true. Since a, b, c, and d are simply the coefficients of the first\nharmonic, only A, B, C and K are unknown.\nFrom Eq. (29) a maximized value of X(s) is (a2 + b21/2 and from\nEq. (30) a maximized value of Y(s) is (c2+d2)1/2. Therefore a maximum\nvalue for (X(s)2+y(s)2) 1/2 is a2 +c2+d2)1/2 which is set equal to\nK. This gives us a fairly accurate measure of the ellipse's semimajor\naxis. K, as computed above, will always be slightly greater than the\ntrue length of the semimajor axis. However, this is quite satisfactory\nsince the length of the echo's axis is the sum of several harmonics and\nnot just of the first harmonic alone.\nUsing determinants, A, B, and C may be found by solving Eqs. (34), ,\n(35) and (37) simultaneously. Specifically\n2\nK\nac\nC\nd2\nbd\nK\n1/(ad\nbc)\ncd\n0\n+\n(38)\nA =\nc2\na2\nac\nC\nd2\nb2\nbd\nbc)\ncd\n+\nExpanding the determinants and combining like terms yields:\nadc2)\nbc)\n+\n(39)\n(bc\nad)\n+\nB and C are computed likewise.\n11","Once the ellipse's coefficients have been found, its orientation can\nbe determined using the relationship (Morris and Brown, 1937)\ntan (2w) = ABC\n(40)\nor\n.\n(41)\nThere is an ambiguity as to which axis of the ellipse from which W is\nmeasured. This ambiguity may be resolved by considering the sign and\nrelative size of A, B and C. However, if the ellipse is first rotated\nthrough angle W eliminating B, its equation becomes\nA'x2 + c'y2 = K\n(42)\n.\nThen if A' is less than c', W is measured with respect to the major axis;\nif c' is greater than A', w is measured with respect to the minor axis.\nNow if the slope of a line, the length between two points on that\nline, and the coordinate of one of the points are all known, we may\nsolve for the coordinates of the unknown point by combining the equation\nof a straight line with the equation for the distance between two points.\nFurther assume that the given line intersects the echo path at ',Y')\nwhere x' and Y' are known and the unknown coordinates are X and Y. Then\nY-Y'M(X- X') -\n(43)\nwhere M = tan W is the slope of the line and\n(44)\n-\nwhere R is the distance between (X,Y) and (X' Y'). We square Eq. (43) and\nset it into Eq. (44) yielding\n- x')2\n(45)\nFactoring and transferring terms yields\n(46)\nFinally, taking the square root of each side and solving for X gives us\n(47)\n12","and Y for each X is given by\nY' + M(X - X')\n(48)\nThus, we have found two points--one above and one below the echo path\nwhich define the initial warning boundary. There are also two points\nwhich define the final warning boundary. These four points define the\nwarning area. R in the above equations is equivalent to K in the general\nequation of the ellipse. However, in solving for the above points, R is\nmodified to include od. od is scaled linearly such that the total length\nfor R is given by\nat Pb\n(49)\nat Pe e\n(50)\n.\ntl is the time of the last echo observation.\nFrom Figure 6 it will be recognized that the warning area is a\nmodified parallelogram. However, as the echo's major axis becomes more\nclosely aligned with the echo motion, the parallelogram closes. There-\nfore, whenever the echo's motion and the echo's major axis are within\n10 degrees of each other, the minor axis of the best fit ellipse is used\nfor K and the slope of the lines passing through P and P respectively,\nare given as -1/XM where XM is the slope of the echo motion line passing\nthrough both Pb and Pe (fig. 6).\nCASE FOR w<10°\nSAMPLE ECHO\nMINOR\nAPPROXIMATING\nS\nAXIS\nD\nELLIPSE\nECHO\nW\nMOTION\nECHO\nMOTION\nTO\nPB\nPE\nECHO\n\"Of\",\nnot\nMOTION\nto\nFigure 6. Illustration of use of an approximating ellipse, linear least\nsquares predicted echo trajectory and the RMSE values to cal-\nculate graphic warning area for W 10 degrees.\n13","5. REAL TIME SYSTEM AND PROGRAM OPERATION\nSince the remote radar display system has the built-in capability\nto be interfaced to a computer, we adopted the programming philosophy\nof duplicating, as nearly as possible, a real-time operation. In this\nsection we shall first describe a model system and its components;\nsecond, describe the decision making and choices within the software\navailable to an operator; and third, offer some guidelines for using the\nprogram logic.\n5.1 Hypothetical Systems Configuration\nIn addition to the electronic components already described in Vol-\nume I, the system requires a central processor with a 50K decimal word\nmemory core. In order to operate in a pseudo-real-time manner, memory\ncycle time should be about one usec. (The Systems Engineering Laboratory's\nmodel 8600, on which the software was developed, has a memory cycle time\nof 600 nanosec.)\nSecondly, some sort of mass storage unit is needed. When not being\nused, the prediction and display logic resides there. Otherwise, the\nFortran program would have to be entered each time the system is used.\nAlso stored on disk are three data files: a) coordinates for graphically\ndisplaying the State of Oklahoma, b) coordinates for graphically dis-\nplaying the Victor Airways, and c) a list of Oklahoma airports. (The\nuse of these files is explained below.) Last, an I/O device such as a\nteletype or alphameric CRT with keyboard entry is needed. The operator\nmust manually insert commands and echo information into the software and,\nin turn, receives back numerical values of echo speed and direction of\nmotion and a measure of the predictability of echo motion. Information\nflow is shown in the systems flow chart (fig. 7).\n5.2 Hypothetical Software Logic\nFirst, let us assume that the program already resides on disk and\nhas been given the name ECHOPRED. The operator then merely enters\nECHOPRED to bring the program to an operational status. The operator's\nfirst decision is whether or not to initialize the program. (Figure 8\nillustrates the command structure which is presented in this section.)\nThis depends upon whether or not the operator is working a new storm\nday. For a new day or a long break in operation, the operator's response\nwill be 'YES', otherwise we presume he is still working the same storms\nand the response is 'NO', to the question, 'INITIALIZE'. When the answer\nis 'NO', echo information is retrieved from disk. The program should be\nleft operational during storm conditions; only if a power failure dis-\nrupts operation should the operator not initialize the program.\nNext, the program will ask for 'COMMAND'. Assuming this is the\nstart of operation or recovery after a failure, a systems check should\nbe made. After the operator responds with 'RQC', Radar Quality Control,\n14","REMOTE\nKEYBOARD\nRADAR CRT\nENTRY\nGRAPHIC\nCOMMANDS AND\nMASS\nWARNINGS\nCENTROID DATA\nSTORAGE\nSTATE\nOUTLINE\nCOMPUTER\n50 HZ\nSOFTWARE\nREFRESH\nVICTOR\nMODEM\nOF DISPLAY\nMEMORY\nAIRWAYS\nAND\nFOR CRT\nPREDICTION\nSTATE\nAIRPORTS\nNUMERICAL\nRESULTS FOR\nECHO PRED\nALPHAMERIC\nDISPLAY\nFigure 7. Schematic of systems flow chart.\nthe computer will type first 'PPI CHECK' with appropriate response being\n'YES' or 'NO', and then 'TEST PATTERN'. Again the operator responds\n'YES' or 'NO'.\nIn 'PPI CHECK' the program checks the housekeeping information (date,\ntime, azimuth) in detail and also counts the number of bins of each inten-\nsity in the PPI and presents this information to the operator. A system-\natic decrease in the frequency of higher intensities should occur when\nonly ground clutter returns are present. A low count at especially the\nfirst, second or fourth intensity levels may indicate hardware failure.\nA few random housekeeping errors will occur due to telephone line noise\nand are not serious.\nThe 'TEST PATTERN' is a computer generated field of seven concentric\nrings 20 degrees in width corresponding to each intensity switch sur-\nrounded by seven radial stripes of 10 degrees width starting with 0 deg-\nrees AZM, and repeated through 360 degrees. The purpose is to check the\nfidelity of the receiver memory. For least confusion, we recommend set-\nting the intensity switches to the following gray shade pattern:\n1 2 3 1 2 3 2.\nWhen the computer has finished with one or both of the above tests,\nit will again type 'COMMAND'. At this point the operator may wish to\nenter a value for the Time Weight Constant, TWC. After TWC is entered,\nthe computer will type 'HHMM' for which the operator enters a number such\n15","(Here and on adjoining page) Illustration showing various\nFigure 8. .\ncommands used in ECHOPRED.\nINITIALIZE\nYES\nCOMMAND\nRQC\nPPI CHECK\nYES\nTEST PATTERN\nYES\nIAZ TILT STC JUL TIME DLY GL TC\n0 0 0 164 110025\n0\n1\n1\nLAST AZIMUTH = 359 LAST RADIAL = 360\nINTENSITY BIT COUNT\n3\n4\n5\n6\n7\n1\n2\n686\n494\n0\n844\n1085\n803\n1128\nSET INTENSITY SWITCHES TO 1 2 3 1 2 3 2\nCOMMAND\nTWC\nHHMM\n30\nCOMMAND\nGCD\n20\nCOMMAND\nACC\nAREA/INTENSITY\n100\n3\nDAY/ TIME/ TILT\n164\n1110\n0\n2 STR ECHOES FOUND WITH AREA GREATER THAN 100 SQ KM\nAREA/AZIMUTH/RANGE\n170\n334\n142\n146\n350\n137\nDAY/ TIME/ TILT\n164 1120\n0\nSTOP\n2 STR ECHOES FOUND WITH AREA GREATER THAN 100 SQ KM\nAREA/AZIMUTH/RANGE\n137\n205\n337\n144\n354\n133\nCOMMAND\nENT\n16","N/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/ STDDS/ STDTM\n1 1110 334. 142.\n2 1110 350. 137.\n1 1120 337. 137.\n279.9\n53.1\n0.00\n0.000\n2 1130 337. 137.\n285.0\n61.4\n0.00\n0.000\n0\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERPLAY/ECHO NUMBERS\n1120\n1220\n100\nVIC\n1\n2\nWHE\nECHO NO/ AZM/ RNG\n1 19. 104.\n12.4 KM(+/- 0.0) AT 1302\n(1211 - 1354)\nCOMMAND\nAIR\nBTIME/ ETIME/ STD DEV/ ECHO NO.S\n1120 1220\n3.\n1\n2\nECHO\n1\nTIME AIRPORT\nDIST\nFTIM LTIM\nNO ENCOUNTERS PREDICTED\nECHO\n2\nTIME AIRPORT\nDIST\nFTIM LTIM\n1151 PERRY\n3.2 S\n1147 1154\n1 ENCOUNTERS LISTED\nCOMMAND\nPOS\nECHO NO/ HHMM\n2 1200\nAZM.RNG = 11.9 124.4 RAD 1SD= 0.0 RAD 3SD= 0.0\nCOMMAND\nDEL\nWHICH ECHOES\n1\n2\n* NO ACTIVE ECHOES\nCOMMAND\nIGN\nECHOES 1 AND 2 WERE DELETED BECAUSE THEY WERE NOT STRONG ENOUGH TO\nCONSTITUTE A HAZARD TO AVIATION.\nKEY\nCOMMAND\nBYE\n17","as 2400. If no entry is made for TWC the program uses a 30 minute default\nvalue. The TWC exponentially weights the influence that past centroids\nhave when predicting echo motion, giving greatest weight to the most recent\npoint (see section 5.3.1).\nAnother parameter the operator may wish to change is the Ground Clutter\nDistance, GCD. The default value is normally set at 20 km for the NSSL\nradar, to omit all of the ground targets from the analysis. When the\nground clutter is extended by abnormal propagation, spurious echoes may\nbe processed. By setting the size and intensity criteria high enough,\nthese echoes will be ignored. However, time will be lost determining\nthis fact. As echoes move into the ground clutter, spurious echoes will\ncomplicate the shape, but not seriously affect total echo area and cen-\ntroid position. Here a simple and expedient method is to tilt the radar\nantenna at two degrees which will effectively remove the ground targets\nfrom the scope while leaving the echo pattern mostly unchanged. When\nanomalous echoes are extensive, some program speed-up can be realized by\nsetting the GCD value artificially large, say 100 km. `However, the risk\nhere is that the operator will fail to reset the value as echoes approach\nthat range.\nThe next command by the operator instructs the computer to accept the\nremote radar data and locate echoes. Before doing this the computer will\nreply 'AREA/INTENSITY'. The operator must then respond with two values,\nfor example, 100 km2 and the 4th code switch. This means that only echoes\nwhose areas are greater than 100 km 2 and whose intensities are greater than\nor equal to the dBZ value corresponding to the fourth intensity switch are\ncontoured. For the data used in this report the dBZ value is about 40--a\nrainfall rate of 12 mm hr-1 (0.5 in hr-1). The program also checks for\nechoes whose area is five times that given. Whenever this occurs the lowest\nintensity in the echo is purged and the next intensity level checked to see\nif it meets 100 km 2 criterion. If it does, the information for the higher\nintensity core is saved as well as the lower intensity.\nAt the completion of the PPI, the computer writes out each echo's\narea and centroid location (azimuth and range); the echoes being sorted\nby intensity. The program will continue to accept new PPI information\nunless interrupted by a 'STOP' command typed in by the operator. The\nprogram does no matching of echoes between two PPI's and no information\nfrom a previous PPI is saved in the computer except that manually entered\nby the operator. Since two centroid positions are required to make a\nprediction, the operator should wait two scans before entering time, azi-\nmuth and range data.\nWhen the program has received a 'STOP' command, it will ask for a\nnew command. The operator will then want to enter the echo information.\nThis is done by first typing 'ENT'. The computer will then ask for an\necho number, time of observation and echo azimuth and range; it will type\nthe following:\n'N/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/ STDDS/ STDTM'\n18","The echo may have any number assigned between one and 99. The time is\nentered as a four digit number. The first two digits are for the hour\n(a 24-hour day is used); the second two digits are for the nearest whole\nminute. Azimuth is entered to the nearest whole degree; range to the\nnearest kilometer. After the first two entries and each subsequent entry\nfor the same storm, immediately following, the program will respond with\na speed, direction of motion and the standard deviation in distance and\ntime of the echo's motion (cf section 4). The program checks the manu-\nally entered data for simple entry errors. Also, if the last time\nentered is the same as the time of the last PPI scanned by the program,\nthe program will match the manually entered centroid position to the com-\nputer derived centroid position. The echo selected is the one whose cen-\ntroid distance is a minimum from that manually entered. If no match is\nfound, the observation is deleted and an error message generated. If\nthe computed echo speed is greater than 120 km/hr and error message is\nalso typed out but the observation is not purged. (Echoes moving at\nthat speed are rare.) When there are no more observations to enter, the\noperator enters a zero for the echo number; the program will respond by\nasking for a new command. One last operation which can be performed is\nto delete an observation by entering a minus sign in front of the echo\nnumber followed by time and centroid positions. Up to ten different\nechoes can be stored at one time.\nThere are four different commands the operator may wish to give the\nprogram now. They are 'DIS', 'WHE', 'AIR', and 'POS'.\nWhen 'DIS' is entered, the operator has asked for a graphic display\nof projected storm motion on the remote terminal. The program will ask\nfor a prediction time interval, a specified range, a graphic overlay,\nand the operator assigned numbers of echoes that the operator wishes to\nsee. The computer will type:\nBBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\nBTIME and ETIME are the beginning and ending times for the prediction\nperiod entered in the same four digit format as for entering echo time\ndata. A useful beginning time might be the time of the last PPI and the\nending time might be one hour later. RANGE lets the operator choose\nbetween a 200 or a 400 km range. If radar data were previously being\ntransmitted at one range, the operator would probably also want the\ngraphic display to be scaled the same. There are two choices for 'OVERLAY'.\nThey are the Oklahoma State outline entered as 'STA' and the low level\nVictor Airways entered as 'VIC' Esthetically, the former is more suited\nto 400 km range, while the latter to a 200 km range. If that entry is\nleft blank, no overlay is produced on the remote radar scope. The ECHO\nNUMBERS are those assigned by the operator.\nAnother option is to ask WHEn the centroid of a storm will be near-\nest a given point. After the operator has entered 'WHE' the program\nwill type:\n19","'ECHO NO/AZM/RNG'\nThe operator then enters the appropriate information, where AZM and RNG\ngive the position of the point in question, not the centroid of the echo.\nThe computer will return the distance from the point normal to the\nextrapolated echo path and the standard deviation of that distance. Time\nof arrival and two other times, one before and one after the predicted\ntime, are also computed. These times are t + Ot.\nAnother option available to the operator is to ask for the AIRports\nwhich lie in the echo path. The echo path is considered to be a cone\nwhose apex is the centroid position at the time of the last observation.\nThe cone expands downstream as a function of time (fig. 9)* After 'AIR'\nhas been entered as a command, the computer responds by typing\nBBTIME/ETIME/STD DEV/ECHO NO.S'\nBTIME and ETIME are enterable for, DIS. STD DEV is entered as a whole\nnumber (e.g , 1, 2, or 3). The expansion rate of the cone is determined\nby od (STD DEV). ECHO NO.S are the user assigned storm numbers.\nThe computer will return with the predicted encounters listed by\nstorm in the order of arrival time, ta. Also given is the distance to\nthe echo path at the arrival time and ta+ot and ta - 't .\nOne last question the operator may address is, \"What will the echo's\nPOSition be at a later time?\" After 'POS' has been entered, the computer\nwill type:\n'ECHO NO/ HHMM'\nThe operator enters the required data in the same manner as under\nthe command 'ENT'. After the above information is entered, the computer\nFigure 9. Illustration of\ncone used to locate air-\nports in an echo's path.\nOD\nPB\nPE\nInteger 3 was entered\nECHO\nfor 'STDDEV.\nPATH\n30\nD\n*Instead of a cone for determining what airports may be affected, the\nmodified parallelogram of the preceding section could be used.\n20","will return the azimuth and range of the centroid for the time given and\ntwo values which are od and 3rd.\nBecause computer storage restrictions permit only ten echoes to be\ntracked simultaneously, the operator will need to DELete echoes from time\nto time. The operator should consider first those echoes which are no\nlonger being actively tracked.\nAfter entry of 'DEL', the computer asks:\n'WHICH ECHOES'\nTo this the operator responds by entering the assigned echo numbers. Before\nasking for a new command, the computer lists the active echoes.\nOn occasion the operator may wish to enter information of a textual\nnature. This might include severe weather events associated with a partic-\nular storm, or storm tendencies for a new operator coming on duty. The\ncommand 'IGN' for IGNnore is entered. After this, a text of any number of\nlines may be entered. To restore the program to an operational mode, the\noperator types 'KEY' at the beginning of a new line.\nTo terminate the program, the operator types 'BYE' for a command.\n5.3 Some Practical Guidelines for Using Echo Prediction Software\n5.3.1 Time Weight Constant (TWC)\nAt the time the echo prediction logic was developed, it was recognized\nthat severe storms, especially tornado producers, \"turn right\" as they\nbecome severe (Newton and Fankhauser, 1964). In order to take the path\ncurvature into account in predicting future echo positions, an exponen-\ntially assigned weighting function was incorporated into the computer soft-\nware. Mathematically, the function is W = e-kAt/TWC where k is 1n 2, TWC,\nthe value entered by an operator and At the time interval between two\nobservations. The rate of decay is a function of time (fig. 10) such that\nwhen the time elapsed from the last point entered is equal to TWC, the\nvalue of the previous point is decreased by half. Although the prediction\nlogic was operationally tested in 1969 and 1970, TWC was always made large\nenough (24 hours) such that W=1. In other words, all centroid positions\ncarried the same weight when fitting LLS equations.\nAs a first step in testing the utility of changing the weighting\nfunction, points were arbitrarily entered at 10 degree intervals at a\nrange of 10 km through a 90 degree sector. The sampling interval was\nentered as five minutes. The results for k = 24 hrs, 30, 15 and 5 min-\nutes are shown in Figure 11a-d. An examination of the figures shows that\nthe greatest improvement in following the data trend is between the 15 and\n5 minute weighting function. Of the four parameters, direction, speed,\nstandard deviation of distance and standard deviation of time, the latter\n21","showed the greatest overall improve-\n100\nment--nearly a factor of three from\n.121 hours to .047 hours. Least\n90\naffected was the speed.\n80\ne -kAt/TWC\nW =\nIn another test, real data from\n70\ntwo storms were used to determine the\neffects of varying the weighting\n60\nfunction. One storm was tracked\nbetween 1225Z and 1310Z; the other\n50\nbetween 1247Z and 1310Z. Two time\nweight constants (TWC)--30 and 5 min-\n40\nutes-- were used. Also, because the\ndata density was greater than other\n30\ncases--2 to 3 minute intervals--two\npasses at each TWC were made, one at\n20\na 2 to 3 minute spacing and the other\nat a 4 to 6 minute spacing. The\n10\nresults are shown in Table 1. Con-\ntrary to the results in the first\no\no\n0.5\n1.0\n1.5\n2.0\ntest, there is little if any improve-\nAt in multiples of TWC\nment in od and ot between a 30 and a\nFigure 10. Illustration of response\n5 minute TWC for 2-3 minute spacing.\nUsing 5 minute data spacing and com-\ncurve for weighting function, W.\nparing the od and O t between a 30 and\n5 minute TWC shows some overall\nimprovement. By far the greatest improvement is to use 5 minute data\ninstead of 2-3 minute data. Several sources of error suggest why this\nis so. One is the natural variability of the echo; that is the random\ngain or loss of echo due to small scale echo changes. Second, radar\nsystem fluctuations of 1-2 dB would cause small scale changes. Third,\nand perhaps the most important, is the system resolution. Typically, an\necho might move 1-2 km in a 2 minute period. At a range of 100 km this\nmight be a change of centroid location of 1 degree azimuth or 1 km range\nor both. Such a motion results in a very noisy path. Barclay and Wilk\n(1970) also noted erratic echo movement from centroid positions when\nusing data of similar density. In the remoting system, one must also con-\nsider the time element. If one were to use radar data directly and had\nthe time of each radial, the centroid time would be precisely known. How-\never, the time of the data displayed on the remote scope is truncated to\nthe nearest minute so that the time the echo was sampled could conceivably\nbe one minute off. When the truncation error is a large fraction of the\nAT, the projected error from this cause tends to be large.\nLet us consider now the speeds and directions actually computed under\nthe different conditions. In general, the TWC had more effect than data\ndensity. Echo 1, with TWC equal to 5 minutes, accelerated sharply after\n1245 nearly doubling its speed by 1306. With a TWC of 30 minutes, the\nacceleration is smoothed considerably. The direction also shows consider-\nably more variance with a 5 minute TWC than with a 30 minute TWC. Echo 2\nshows basically the same features as Echo 1.\n22","21.0\nTRUE POINT TO POINT SPEED\n360\n20.8\n350\n20.6\nTWC\n340\n20.4\n0005\n330\n20.2\n0015\nTWC\n320\n20.0\n0030\n0005\n2400\n310\n19.8\n300\n19.6\n0015\n290\n19.4\n0030 &\n2400\n280\n19.2\n270\n19.0\no\nIO\n20\n30\n40\n50\no\nIO\n20\n30\n40\n50\nELAPSED TIME (min)\nELAPSED TIME (min)\n(a)\n(b)\n.10\n10\n2400\n0030\nTWC\n2400\n0015\n9\n0030\nTWC\n0015\n8\n.08\n7\n0005\n.06\n6\n5\n0005\n4\n.04\n3\n.02\n2\nI\no\no\n20\n30\n40\n50\no\nIO\n20\n30\n40\n50\no\n10\nELAPSED TIME (min)\nELAPSED TIME (min)\n(c)\n(d)\nFigure 11. Illustration of the effect of varying the TWC for an echo\nwhose track is curved in calculating (a) direction,\n(b) speed, (c) od, and (d) 't'\n23","Comparison of results of using a TWC of both 30 and\nTable 1.\n5\nminutes on data sampled at both 2-3 minutes and\n4-6 minutes for tracking two echoes.\nTWC = 30 MIN\nTWC = 5 MIN\nECHO 1\n2 MIN DATA\nTIME\nDIR\nSPD\nSTDDS\nSTD TM\nDIR\nSPD\nSTDDS\nSTD TM\n1230\n254.6\n23.8\n27.41\n.803\n247.3\n24.3\n27.41\n.803\n1236\n273.8\n37.6\n39.05\n.662\n272.1\n39.2\n40.64\n.614\n1240\n258.7\n34.6\n37.54\n.624\n252.2\n34.6\n38.67\n.583\n1245\n253.0\n36.1\n33.66\n.609\n248.4\n37.9\n34.59\n.579\n1251\n238.5\n39.7\n40.65\n.570\n227.9\n44.7\n39.87\n.546\n1255\n228.7\n43.2\n41.17\n.620\n216.0\n52.0\n38.63\n.594\n1300\n224.8\n50.5\n42.34\n.670\n219.2\n62.3\n41.19\n.612\n1306\n218.8\n55.4\n47.77\n.661\n210.3\n67.6\n45.60\n.602\n1310\n219.3\n55.1\n49.99\n.695\n217.0\n58.4\n47.67\n.655\n5 MIN DATA\n1230\n251.0\n24.0\n0.00\n.000\n251.0\n24.0\n0.00\n.000\n1236\n273.9\n36.4\n14.10\n.394\n276.2\n38.8\n14.10\n.394\n1240\n263.4\n34.1\n18.90\n.410\n259.4\n34.0\n19.20\n.429\n1245\n257.1\n36.7\n17.63\n.403\n252.2\n38.8\n17.52\n.422\n1251\n238.4\n39.3\n28.34\n.368\n223.6\n45.7\n27.20\n.386\n1255\n228.2\n43.6\n31.01\n.405\n214.7\n53.6\n27.09\n.415\n1300\n227.0\n48.7\n29.75\n.448\n222.3\n60.1\n28.25\n.397\n1306\n220.9\n53.0\n32.47\n.436\n211.2\n65.6\n30.34\n.377\n1310\n220.7\n53.6\n32.62\n.437\n216.6\n58.7\n31.25\n.412\nECHO 2\n2 MIN DATA\n1230\n-\n-\n-\n-\n-\n-\n-\n-\n1236\n-\n-\n-\n-\n-\n-\n-\n-\n1240\n-\n-\n-\n-\n-\n-\n-\n-\n1247\n346.1\n74.1\n0.00\n0.000\n346.1\n74.2\n0.00\n0.000\n1251\n270.7\n38.9\n62.34\n1.099\n268.6\n36.3\n62.06\n0.944\n1255\n247.0\n41.7\n57.18\n1.020\n241.1\n42.1\n56.50\n1.049\n1300\n233.5\n66.4\n56.79\n1.362\n230.9\n73.9\n53.82\n1.372\n1306\n228.3\n74.7\n52.68\n1.236\n226.3\n79.8\n48.81\n1.201\n1310\n229.6\n70.0\n48.35\n1.146\n230.4\n66.3\n45.93\n1.155\n5 MIN DATA\n1230\n-\n-\n-\n-\n-\n-\n-\n-\n1236\n-\n-\n-\n-\n-\n-\n-\n-\n1240\n-\n-\n-\n-\n-\n-\n-\n-\n1247\n-\n-\n-\n-\n-\n-\n-\n-\n1251\n286.1\n33.2\n0.00\n0.000\n286.1\n33.2\n0.00\n0.000\n1255\n251.8\n38.2\n32.21\n0.075\n245.0\n41.3\n32.21\n0.075\n1300\n235.8\n65.0\n36.27\n0.912\n231.8\n78.6\n32.46\n0.863\n1306\n231.6\n72.0\n32.85\n0.817\n228.9\n79.1\n29.19\n0.774\n1310\n232.6\n66.9\n30.79\n0.803\n232.5\n64.5\n27.63\n0.772\n24","Although two storms are admittedly a small sample, our experience in\na large number of cases makes us believe that these results apply to other\nstorms as well. Essentially, what is indicated is that for a five minute\nforecast of echo motion, one should use a five minute TWC, with five minute\ndata resolution. However, for a 30 to 60 minute forecast, a larger TWC,\nsuch as 30 minutes, should be used. Over this length of time, trends in\nthe overall echo motion are more important than short term fluctuations\nwhich should be smoothed out. Also, since warning areas are mapped out\nbased on echo speed and motion, large variance from one time to the next\nwould only confuse the user and cause the warning areas to shift consid-\nerably from one prediction PPI to another.\n5.3.2 Selection of Area and Intensity Criteria\nIn choosing the threshold area and intensity the user is hampered by\nbeing limited to ten echoes. However, the ability to assimilate and follow\neven ten is questionable. On a scope containing many echoes, therefore,\nthe user's attention should be directed to the largest and/or strongest\nechoes.\nBarclay and Wilk (1970) determined that for echo extrapolation using\ncentroid data, threshold values of 103 to 104 mm 6. m 3 for isolated\nstorms\nand 102 to 10 3mm6m-3 for squall lines worked best. Based on his own expe-\nrience, the author believes these are good criteria. With higher inten-\nsities, tracking is difficult because the lifetimes of the intense cores\nare short and in squall lines the probability of mergers and splits also\nincreases. If one uses too low an intensity, information concerning those\nareas most hazardous to aircraft is lost. A general rule-of-thumb is to\nuse the lowest intensity for which a discrete echo can still be defined.\nIn a master's thesis in 1969, R. A. Houze, Jr. defined three areal\nsizes associated with New England precipitation: \"synoptic scale areas\"\non the order of 104 - 105 mi² and a duration of about 10 hours to pass a\npoint; \"mesoscale areas\" on the order of 102 - 103 mi 2 and a duration of\nabout one hour; and \"cells\" with a 1-10 mi 2 area and lasting about one minute.\nThe scale size which is associated with severe weather and with which we\nhave concerned ourselves in this report, is the mesoscale. The synoptic\nscale pattern is dependent on larger atmospheric disturbances and its\nmovement is better forecast by existing NWS software. Also, the average\nradar intensity does not constitute a hazard to aviation from strong shear\nor turbulence. Small, intense cells, on the other hand, are of too short\na duration to be tracked and where they do occur--imbedded in mesoscale\nsystems--the entire area should be avoided.\nThe average area of eight extremely severe isolated storms which have\noccurred in Oklahoma over the past six years was 208 n mi 2 (713 km2) with\na\nrange of 122 to 427 n mi 2 Squall lines generally are about three times the\nsize of isolated echoes, although cells within squa11 lines are usually\nslightly smaller than severe, isolated storms. Since the areas given\nabove are for total storms, the actual area threshold used is reduced by a\nfactor of 2 or 3 depending on the intensity threshold. In summary, the\narea threshold ranges between 100 and 1000 km while the intensity varies\n25","102mm6\n3\nto 10 mm 6 m - 3\n-\nfrom\nExperience guides determination of the best\nm\n.\ncombination for any given storm system.\nOnce the criteria are established, the user is well advised to resist\nfrequent change. The purpose in establishing criteria is to give the user\na history for summarizing events on the display. Frequent changes will\ncause loss of continuity of pattern.\nIf it becomes difficult to match echoes due to frequent splitting or\nmerging, however, then a lower intensity and a larger area should be\nselected. Conversely, if persistent significant core elements are being\nomitted, then higher intensity and smaller area threshold are indicated.\n5.3.3 Splits and Mergers\nOne of the biggest problems in using echo centroids to track and\nextrapolate future positions is how to handle splitting or merging storms.\nFor example, if a storm splits into two distinct cores, should you regard\none of the cores as a continuation of the old echo and tag the other core\nas a new echo or should both be treated as new echoes? A similar problem\nexists with mergers.\n180\nThe best procedure the author\nhas found is to follow trends in\narea and centroid position for each\n160\ncore in question. Typically, in a\nfive minute period, an echo will\n140\n2 TAN'(X/2R)\nmove from 3 to 6 km. If motion is\nalong a radial, then the centroid\n120\nposition is simply a function of\nrange. If, however, the azimuth\nchanges, then the incremental change\n100\nis range dependent. Figure 12\nshows the change in azimuth as a\n80\nfunction of range for motion tan-\ngent to a circle at that range.\n60\nThe important consideration remains\nto look for discontinuities in\neither range or azimuth. A change\n40\nin range 10 km and/or 10 degrees\ngreater than expected coupled with\n20\na 20 percent or greater change in\narea, should be considered a new\no\necho.\no\n10\n20\n30\n40\n50\n60\n70\nRANGE (km)\n6. TEST CASES\nThree days were selected for\nFigure 12. Illustration of the angular\nprogram testing. On two days,\nchange in centroid position as a\nJune 6 and June 16, 1975, squall\nfunction of range. X is the dis-\nlines producing damaging surface\nplacement distance of the centroid.\n26","winds moved across the State; the third case on November 2, 1974, produced\nlocalized flooding in the northwest Oklahoma City with cumulative rainfall\namounts in excess of five inches.\nThe three cases above were chosen because of their danger to the avia-\ntion community. In the case of fast moving squalls, the inherent danger\nis from the turbulence and high winds in and around them. In the case of\nflooding, aside from the intensity of storms, the sheer persistence of heavy\nrain in one location implies time delays for air traffic leaving and/or\narriving, or for circumnavigating those features. Certainly, it is of the\nutmost importance for a pilot, en route, to know of adverse conditions which\nwill not clear his destination by his ETA.\nThe case studies were made to develop the logic to generate graphic\nforecasts and to determine the feasibility of using the polar coordinate\ndisplay as a graphics terminal. The analyses excluded quantitative veri-\nfication of the predicted echo motion. The cases were analyzed using\nvariations of the echo tracking logic, involving different cutoff limits\nfor magnitudes of od and ot in expanding the warning areas.\nIn simulating real time operations, radar data archived on magnetic\ntape were reformatted to \"look\" like data received by the remote radar\ndisplay program. A second tape was generated by the echo prediction logic\nwhich contained the graphics information, also in the format for the\nremoting system. The graphics tape was then played through a tape drive*\nand transmitted to the receiver where the graphics were photographed.\nThere are three components to the graphics: individual warning areas\nfor one or more storms; a contour at a specified intensity threshold for\neach echo; and a background reference field, showing the Victor Airways or\nthe state outline. Each component is coded at one of three 'intensity'\nvalues to produce differential gray shading. Warning areas were coded in\nthe seventh intensity level (not normally used in the real time echo dis-\nplay) and displayed at the brightest gray-shade level; echo contours were\ncoded in the second intensity level and displayed at the medium gray-shade\nlevel; and the Victor Airways were coded in the first intensity level and\ndisplayed at the lightest gray shade level.\nAfter reviewing the radar data on the remote scope, threshold criteria\nwere selected and an initial run was made using the echo prediction logic.\nThe area and centroid data from this run were matched and a second run was\nmade using these results to generate the warning areas. The following\ncases illustrate the logic and generation of the graphics.\nCase 1, June 6, 1975\nA stationary front was indicated on the 12Z surface map (fig. 13) as\nextending from the Oklahoma Texas panhandles northeastward across the\n*As part of the specifications, a half-inch magnetic tape interface at\nthe transmitter was provided.\n27","southern third of Kansas; a cold front was entering Nebraska to the north.\nThe upper level flow at 500 mb (fig. 14) was from the west-northwest.\nStorms broke out along the stationary front in Kansas in the early after-\nnoon and generally moved southeast in the same direction as the flow\naloft. Abundent moisture was available in central Oklahoma (18-20 g kg -1)\nand the air was potentially unstable. Surface winds were from the south-\nsoutheast at 15-20 knots. By 2000 CST, a well-developed dry line oriented\nnorth-south existed in the Texas and Oklahoma panhandles as shown by the\nsurface analysis with streamlines in Figure 15. Cold air produced by the\nstorms in western Kansas had produced a pseudo cold front as well. There-\nfore, an area of strong convergence formed in northwest Oklahoma (fig. 16).\nAs the Kansas storms were following the same track, growth was favored on\nthe southern flank of these storms and a line formed propagating southward.\nAs shown in Figure 17, the closest echoes at 1600 were about 200 km\naway along the Kansas-0klahoma border. Because the motion at that time\nwas east-southeast, the echoes were not expected to move within Doppler\nradar range. Hence, operations ceased and, between 1625 and 1830 CST,\nno radar data were collected. The radar, monitored at 1830 CST, indicated\na large echo was still beyond Doppler radar range. Within two hours,\nsignificant changes occurred. A line of storms developed in the northwest\nquadrant and proceeded to move southward into central Oklahoma.\nFor experimental development of the echo prediction and display logic,\nanalysis was begun at 2040 CST when the echoes were within 120 km. A con-\ndensation of echo positions and motions is presented in Table 2. The com-\nplete program command structure with annotation is included as Appendix 1.\nFor this storm period, the size and intensity criteria selected were\n250 km2 2 and code switch 4 (corresponding to 40 dBZ), respectively. Data\nat 0 degree elevation were available at five minute intervals. The time.\nweight constant was set at 2400 (24 hr) and the ground clutter distance\nat 20 km.\nL\nH\nFigure 13. June 6, 1975, 12Z, sur- Figure 14. June 6, 1975, 12Z, 500 mb\nface analysis adapted from DOC,\nanalysis adapted from DOC, NOAA,\nNOAA, EDS daily weather maps,\nEDS daily weather maps, Weekly\nWeekly Series.\nSeries.\n28","SURFACE MAP TIME 2000 CST DATE 060675\nCBIPO\nyour's\non\nor\n570\nOWN ALGOS\nBE-B-N\nMOVE\n076\n9081 2\n500\n(6082\n15.\n10\n575\nVG E\nHUT\n[WOOL\n12th\n008\nMOVE\nL\nrane\n500\nOVHD\nTGCS DENT\nTOTEL\n04\n00\nCYO\nLIGIC\nERO\n03\nDETRES\nEND\n061\nTOMO BONS\nEBE\nLTGIC\nQ\n1406\n035\n30\nQ\nages\n\"\nFCC\n260-0\nFTC\n260-5\n1208\n30\nPoro\nCVS\nn\nX335\n10\nQ\nis\n408\nTXX\nCB GE\nQ56\n1206\n306\n10\n15\nLTGIC\nFigure 15. Subsynoptic surface data provided by NWS, FAA and NSSL sta-\ntions, 2000 CST, June 6, 1975, with streamlines superimposed.\n06-6 -75 2000CST\nAMO 50\n50\n50 200\n250\n150 100\n050\n50 0\nFigure 16. Divergence field derived\n-100\n-50\nfrom surface data, 2000 CST,\n00\n50\nJune 6, 1975. All values are\nX 10-6.\n250\n150\n0\n0\n-100\n-150\n-100\n-50\n-50\n100\n0\n-50\n-50\n50\n0 50 50\n0\n0\n0\n-50\nDIVO\n29","Figure 17. WSR-57 radar PPI,\nNSSL\n400 km range, 100 km range\nJUNE\nmarks, 1600 CST, June 6,\n6\n1975.\n1975\nWSR-5\n4947\nGray shades for each intensity level were set at 1220333 and resulted\nin echoes being contoured at cancellation level on the remote scope.\nAt 2040 CST, as shown in Figure 18, two large cells were located by\nthe computer search at azimuth 29°, range 136 km and at azimuth 335°, range\n157 km. After the second PPI at 2045 CST, the same two cells were matched\nmanually to the previous cells. They were labeled echo 1 and echo 2 and\ntheir respective times and positions were entered for tracking. Signifi-\ncantly different motions were obtained: echo 1 moved from west-northwest\n(293°) at 93 km hr-1 while echo 2 moved from north-northwest (335°) at\n52 km hr-1. Warning graphics were then displayed for these echoes (fig. 19).\nThe starting position for each warning area is 15 minutes after the last\nobservation, the ending position is one hour and 15 minutes later. A11\nsubsequent warning areas were also for one hour duration starting 15 min-\nutes downstream. A circle was drawn on the graphics display after the PPI\nat 2055 (fig. 20) to represent the echo shape. The program does not save\necho shape information for more than one PPI. Whenever a graphic display\nof a warning area is requested for an echo for which no entry was made for\nthe last PPI, a circle with a 10 km radius automatically replaces the\ninitial shape function.\nA third echo located at azimuth at 60° range 195 km, at 2050 showed\nno movement by 2055 CST. The resulting warning area was a line drawn along\nthe major axis of the echo (fig. 20). With the addition of a third point\nat 2100 CST, an exorbitantly large ot was calculated. When the prediction\nfield was first displayed, the large value of Ot distorted the warning area.\nTherefore, a \"cutoff\" limit was introduced such that whenever ot exceeded\nthe prediction interval, te - tb, ot was set to (te - tb) /2. Later, a\n\"cutoff\" limit was also established for od restricting it to the length of\nan echo's axis used in defining the warning area. The choice of these\nlimits was purely arbitrary and may need further adjustment based on a\nlarge sample of data. Obviously, many tracks, based on the first few obser-\nvations, will show considerable scatter. As additional locations are added\nover a larger period of time, a better estimate of the mean velocity will\n30","Table 2. Echo centroid positions, direction and speed predicted,\nand the RMSE values for June 6, 1975.\nECHO\nTIME\nAZIMUTH\nRANGE\nDIRECTION\nSPEED\nod\not\n(KM HR-1)\n(CST)\n(DEGREES)\nNO.\n(KM)\n(DEGREES)\n(KM)\n(HOUR)\n1\n2040\n29.\n136.\n-\n-\n-\n-\n2\n2040\n335.\n157.\n-\n-\n-\n-\n1\n2045\n32.\n137.\n292.5\n92.9\n-\n-\n2\n2045\n335.\n153.\n335.0\n51.5\n-\n-\n1\n2050\n34.\n140.\n283.1\n78.7\n14.36\n0.182\n2\n2050\n336.\n148.\n319.0\n58.2\n17.90\n0.098\n3\n2050\n61.\n188.\n-\n-\n-\n-\n1\n2055\n36.\n144.\n276.9\n75.9\n16.73\n0.162\n3\n2055\n61.\n188.\n360.0\n0.0\n-\n-\n1\n2100\n38.\n146.\n277.1\n73.1\n15.58\n0.159\n2\n2100\n335.\n136.\n334.2\n64.3\n23.26\n0.102\n3\n2100\n61.\n189.\n241.0\n6.1\n6.06\n*\n4\n2100\n298.\n194.\n-\n-\n-\n-\n6\n2100\n310.\n155.\n-\n-\n-\n-\n2\n2105\n336.\n132.\n331.8\n62.8\n24.05\n0.142\n3\n2105\n60.\n187.\n138.9\n12.2\n20.27\n*\n4\n2105\n297.\n192.\n356.8\n49.0\n-\n-\n6\n2105\n308.\n152.\n9.8\n76.7\n-\n-\n2\n2110\n333.\n124.\n339.1\n66.1\n40.99\n0.230\n3\n2110\n59.\n187.\n139.7\n20.2\n18.42\n*\n4\n2110\n296.\n188.\n345.0\n55.0\n12.69\n0.110\n1\n2115\n44.\n162.\n272.1\n79.2\n15.85\n0.171\n2\n2115\n333.\n118.\n341.2\n68.1\n38.06\n0.215\n3\n2115\n61.\n188.\n137.7\n8.1\n16.84\n*\n4\n2115\n295.\n184.\n340.6\n57.4\n12.34\n0.109\n1\n2120\n48.\n160.\n277.9\n80.7\n44.36\n0.159\n2\n2120\n334.\n115.\n340.5\n66.6\n37.05\n0.274\n3\n2120\n60.\n189.\n153.9\n7.1\n17.39\n*\n4\n2120\n295.\n178.\n329.4\n57.6\n25.53\n0.112\n5\n2120\n305.\n131.\n-\n-\n-\n-\n1\n2125\n50.\n165.\n279.9\n82.1\n41.50\n0.155\n2\n2125\n332.\n105.\n341.5\n69.1\n36.33\n0.389\n3\n2125\n60.\n191.\n178.8\n7.0\n18.34\n-\n4\n2125\n296.\n166.\n312.9\n66.7\n50.69\n0.313\n5\n2125\n305.\n126.\n305.0\n61.0\n-\n-\n7\n2125\n13.\n189.\n-\n-\n-\n-\n1\n2130\n50.\n173.\n278.3\n81.8\n46.15\n0.167\n2\n2130\n332.\n101.\n341.7\n69.5\n34.56\n0.383\n4\n2130\n295.\n163.\n310.0\n68.5\n47.67\n0.316\n5\n2130\n304.\n122.\n318.3\n55.8\n14.75\n0.121\n7\n2130\n12.\n187.\n71.1\n46.1\n-\n-\n*In the run used for generating the graphics the values for ot exceeded\n99999.999. By deleting the observation for cell 3 at 2055 CST, Ot was\nreduced to about three hours.\n31","be made. At 2105 CST three cells were associated with the squall line\n(fig. 21). . Generally, the core in the middle cell in the line was too\nsmall to be tracked.\nNo major changes occurred in the pattern but several more PPI's and\ngraphics (figs. 22-27) on this data are presented as examples. The last\nPPI at 2130 CST and graphic warning areas which were displayed at a 400 km\nrange with the State outline. The last radar PPI at 2249 CST (fig. 28)\n(about the ending time of the last warning area) shows that the line had\nmoved as expected.\nas\nFigure 18. (left, above) Remote\nradar display PPI, 200 km range,\n2040 CST, June 6, 1975.\nFigure 19. (above) Remote radar\ndisplay with computer generated\nwarning areas, echo contours and\nVictor airways, June 6, 1975.\nPrediction period is 2100 to\n2200 CST.\nFigure 20. (left) Remote radar\ndisplay with computer generated\nwarning areas, echo contours and\nVictor airways, June 6, 1975.\nPrediction period is 2110 to\n2210 CST.\n32","Figure 21. Remote radar display\nFigure 22. Remote radar display\nPPI, 200 km range, 2105 CST,\nPPI, 200 km range, 2115 CST,\nJune 6, 1975.\nJune 6, 1975.\nFigure 23. Remote radar display\nFigure 24. Remote radar display\nwith computer generated warning\nPPI, 200 km range, 2120 CST.\nareas, echo contours and Victor\nJune 6, 1975.\nairways, June 6, 1975. Predic-\ntion period is 2130 to 2230 CST.\n33","Figure 25. Remote radar display\nFigure 26. Remote radar display\ncomputer generated warning\nPPI, 200 km range, 2130 CST,\nareas, echo contours and Victor\nJune 6, 1975.\nairways, June 6, 1975. Predic-\ntion period is 2135 to 2235 CST.\n4\n3\nNSS\nJUNE\n6\n1975\nWSR.\nFigure 27.\nRemote radar display\nFigure 28. WSR-57 radar PPI,\nwith computer generated warning\n200 km range, 40 km range marks,\nareas, echo contours and Okla-\n2249 CST, June 6, 1975.\nhoma State outline, 400 km range\nJune 6, 1975. Prediction time\nis 2145 to 2245.\n34","Case 2, June 16, 1975\nOn this day, moisture from the Gulf of Mexico and drier continental\nair were separated by a stationary surface front extending from the Okla-\nhoma panhandle across southern Oklahoma and northern Arkansas (fig. 29).\nDuring the next 24 hours this boundary moved northeastward as a warm front\nahead of a cold front approaching from the northwest.\nThe flow at 500 mb was essentially zonal at 20-30 kt (fig. 20). Mix-\ning ratio values during the afternoon of the 16th were 16-18 g kg-1, and\nthe air mass was potentially unstable. When convective temperature was\nreached after 1500 CST, numerous showers and thunderstorms developed over\nKansas, Oklahoma, north Texas and New Mexico. The sequence of satellite\nphotographs and WSR-57 PPI displays (figs. 32a, b, c, and 33a, b, c) trace\nthe growth and movement of these storms during the afternoon.\nUntil 1830 CST, activity was beyond 200 km, and the radar integrator\nwas range delayed (fig. 32) to provide 1 km resolution data between 200\nand 400 km (Sirmans and Doviak, 1973). After 1830 CST, data collection\nwas normal with data recorded at five minute intervals.\nThe area of thunderstorms visible over northern Texas on the GOES\nsatellite photos appeared as a line of moderate to intense echoes on the\nWSR-57 display. These storms intensified and moved rapidly across central\nOklahoma as an organized squall line, preceded by a strong damaging gust\nfront. The NSSFC issued a tornado watch at 1815 CST valid from 2000 CST\nto 0200 CST to cover western Oklahoma, but it lagged spatially behind the\nstorms because of their unexpected acceleration. The steering level wind\nvelocity was 210 degrees at 25 kts (46 km hr-1). However, line motion was\nmuch faster from 270 degrees at 45 kts (80 km hr-1). Figures 34 and 35,\n30 minutes before tracking began, shows the storm influenced surface stream-\nline and convergence patterns, respectively.\nH\nH\nFigure 29. June 16, 1975, 12Z sur- Figure 30. June 16, 1975, 12Z 500 mb\nface analysis adapted from DOC,\nanalysis adapted from DOC, NOAA,\nNOAA, EDS daily weather maps,\nEDS daily weather maps, Weekly\nWeekly Series.\nSeries.\n35","(a)\n(b)\nFigure 31. SMS-3 satellite pic-\ntures, June 16, 1975:\n(a) 1545 CST (2145Z)\n(b) 1615 CST (2215Z)\n(c) 1745 CST (2345Z)\n(c)\n36","90\n0000\nNSSL\nNSSL\nJUNE\nJUNE\n16\n16\n1975\n1975\nWSR-57\n37\n379q\n(a)\n(b)\nFigure 32. WSR-57 radar PPI,\n400 km range with 200 km\nrange delay to first gate,\n40 km range marks, June 16,\nNSSI\n1975:\nJUNE\n(a)\n1545 CST\n(b)\n1615 CST\n(c)\n1745 CST\n(c)\n37","SURFACE MAP TIME 1800 CST DATE 061675\n0\nFigure 33. Subsynoptic surface data provided by NWS, FAA, and\nNSSL stations, 1800 CST, June 16, 1975, with stream-\nlines superimposed.\nTesting of the echo prediction\nlogic began with the 1830 CST observa-\n06-16-75 1800CST\n0 100\n150 -500\ntion. Using intensity and area thresh-\n-250\nolds of 40 dBZ and 250 km 2, , respectively,\n400\n0\n100\nthe storm centroids were determined at\n300\n-50\n5 minute intervals. The four-level -\n-100\n-150\nintensity code for the remote display\nwas set to light, medium, cancel and\n100\n250\nbright (1220333). . The time weight con-\nstant (TWC) was 2400, and the minimum\n200\n-150\nrange (GCD) was 20 km. A11 warning areas\n150\n100\n100\n50\nwere for one hour interval starting with\n-50\nthe time of the last observation.\n-50\n50\nAfter two PPI's (figs. 35 and 36)\n0\ncentroid positions for cell 1 were\n-50\n-50\n0\n0\n0\n150\nentered and an estimate of the echo's\nDIVO\nmotion (from 337 degrees at 44 km hr-1)\nFigure 34. Divergence field derived\nwas obtained (Table 3) . Figure 37 shows\nfrom surface data, 1800 CST,\nthe subsequent warning area based on\nJune 16, 1975.\ntwo centroid positions.\n38","Figure 35. Remote radar display\nFigure 36. Remote radar display\nPPI, 200 km range, 1830 CST,\nPPI, 200 km range, 1835 CST,\nJune 16, 1975.\nJune 16, 1975.\nIt is important to note here that\nas an echo enters the scope, its cen-\ntroid position will be influenced by\nthe area change. As a result, using\ncentroid positions to calculate echo\nmotion will underestimate true echo\nspeed. Likewise, direction of motion\nwill be affected. If the increase in\neach area occurs only at one end of a\nline, the calculated direction of\nmotion will be pulled towards that end.\nUntil such time as an echo or squall\nline has fully entered the scope, it\nis probably better to track a point\nalong its leading edge. (That was not\ndone for this case.)\nFrom Table 3, an increase in line\nspeed is evident through the first half\nhour of tracking. Because of the bound-\nFigure 37. Remote radar display\nary problem just described, large values\nwith computer generated warning\nof and od resulted. An example of\narea, echo contour and Victor air-\na PPI and its associated warning area\nways, June 16, 1975. Prediction\nat 1900 CST are shown in Figures 38 and\nperiod is 1835 to 1935 CST.\n39.\n39","Table 3. Echo centroid positions, direction and speed predicted,\nand the RMSE values for June 16, 1975.\nSPEED\nECHO\nTIME\nAZIMUTH\nRANGE\nDIRECTION\nod\not\n(KM HR-1)\n(CST)\n(DEGREES)\n(KM)\n(DEGREES)\n(KM)\n(HOUR)\nNO.\n1\n1830\n280.\n188.\n-\n-\n-\n-\n1\n1835\n279.\n186.\n338.0\n45.9\n-\n-\n1\n1840\n278.\n183.\n331.3\n49.3\n6.59\n0.070\n1\n1845\n279.\n179.\n302.5\n39.2\n31.11\n0.516\n1\n1850\n276.\n174.\n314.0\n52.1\n43.65\n0.843\n1\n1855\n273.\n171.\n324.0\n63.3\n46.94\n0.845\n1\n1900\n271.\n165.\n325.4\n71.5\n43.74\n0.812\n2\n1905\n262.\n170.\n-\n-\n-\n-\n3\n1905\n287.\n151.\n-\n-\n-\n-\n167.\n305.9\n55.2\n2\n1910\n261.\n-\n-\n3\n1910\n285.\n142.\n315.6\n135.9\n-\n-\n0.134\n2\n1915\n261.\n159.\n276.1\n69.7\n39.15\n0.299\n3\n1915\n287.\n133.\n287.0\n110.4\n59.72\n264.\n146.\n321.3\n77.7\n41.41\n0.761\n1\n1920\n41.81\n0.720\n263.\n140.\n318.7\n78.4\n1\n1925\n77.8\n46.93\n0.690\n1\n1930\n263.\n132.\n315.3\n78.7\n44.85\n0.668\n1\n1935\n260.\n126.\n313.3\n1\n1940\n258.\n120.\n311.7\n79.3\n43.72\n0.640\n309.0\n79.1\n50.48\n0.626\n1\n1945\n259.\n111.\n4\n1950\n263.\n110.\n-\n-\n-\n-\n4\n1955\n264.\n108.\n219.9\n36.2\n-\n-\n4\n2000\n267.\n97.\n235.9\n91.2\n28.38\n1.627\n5\n2005\n263.\n82.\n-\n-\n-\n-\n5\n2010\n263.\n74.\n263.0\n105.0\n-\n-\n110.4\n0.01\n5\n2015\n263.\n64.\n263.0\n0.082\n6\n2020\n285.\n51.\n-\n-\n-\n-\n66.6\n6\n2025\n284.\n46.\n294.1\n-\n-\n62.1\n20.79\n0.099\n6\n2030\n287.\n41.\n276.9\n40","Figure 38. Remote radar display\nFigure 39. Remote radar display\nPPI, 200 km range, 1900 CST,\nwith computer generated warning\nJune 16, 1975.\narea, echo contour and Victor air-\nways, June 16, 1975. Prediction\nperiod is 1900 to 2000 CST.\nAt 1905 CST, the program logic isolated two discrete cells within the\nline at intensity 40 dBZ where there had been only one five minutes before.\nThe split-off of the smaller of the two cells caused the centroid of the\nlarger cell to be shifted nine degrees in azimuth and moved back 5 km. After\nthe split, the two cells were identified as cell 2 and cell 3 and tracked\nfor the next ten minutes. The PPI at 1915 CST (fig. 40) and the graphics\nfor that time (fig. 41) indicate quite different speeds for each cell (see\nalso Table 3)\nAt 1920 CST, cells 2 and 3 merged. Since the resulting centroid position\nwas consistent with the extrapolated position for cell 1, the merge was\nreassigned as cell 1 (fig. 42). It was traced until 1945 CST. Two radar\nPPIs and their respective warning areas (figs. 43-46, at 1930 and 1945 CST)\nshow the line's movement during this period.\nAfter 1945 CST, in the initial pass through the data, the large core\nwhich had been cell 1, fragmented and attempts to match the new cells proved\nfutile. Therefore, a second pass was made with new threshold criteria of\n1000 km2 and intensity level 3. The gray shading was recoded at 1203333 to\nreflect the decrease in the intensity threshold.\nCell 4, tracked from 1950 to 2000 CST, indicated motion in a north-\neasterly direction due, primarily, to growth on the north end of the line.\nMotion had been southeasterly previously. Shown in Figures 47 and 48 is a\nradar PPI and graphic warning for 2000 CST.\n41","Figure 40. Remote radar display\nFigure 41. .\nRemote radar display\nPPI, 200 km range, 1915 CST,\nwith computer generated warning\nJune 16, 1975.\nareas, echo contours and Victor\nairways, June 16, 1975. Predic-\ntion period is 1915 to 2015 CST.\nFigure 42.\nRemote radar display\nFigure 43.\nRemote radar display\nPPI, 200 km range, 1920 CST,\nPPI, 200 km range, 1930 CST,\nJune 16, 1975.\nJune 16, 1975.\n42","Figure 44.\nRemote radar display\nFigure 45. Remote radar display\nwith computer generated warning\nPPI, 200 km range, 1945 CST,\narea, echo contour and Victor\nJune 16, 1975.\nairways, June 16, 1975. Predic-\ntion period is 1930 to 2030 CST.\nFigure 46. Remote radar display\nFigure 47. Remote radar display\nwith computer generated warning\nPPI, 200 km range, 2000 CST,\narea, echo contour and Victor\nJune 16, 1975.\nairways, June 16, 1975. Predic-\ntion period is 1945 to 2045 CST.\n43","At 2005 CST, due to a substantial\nchange in area and range, the largest\necho was relabeled as cell 5. Like\ncell 4, cell 5 was also tracked for only\n15 minutes. As shown in Figures 49 and\n50, and Table 3, the warning area and\necho motion after 10 minutes reflect\nthe line's overall speed and direction\nof motion better than any earlier time.\nA substantial change in area and\ncentroid position again occurred about\n2020 CST with the additional growth on\nthe north end of the line. Cell 5 was\ndropped. The larger echo, renumbered\nas cell 6, moved slower, although the\ndirection was the same. The PPI and\nwarning area (fig. 51 and 52) are shown\nfor 2030 CST when tracking ceased.\nFigure 48. Remote radar display\nProbably a longer sampling inter-\nwith computer generated warning\nval should have been used for this case\narea, echo contour and Victor\n(e.g. 15 minutes). However, several\nairways, June 16, 1975. Predic-\nfactors have to be considered. The\ntion period is 2000 to 2100 CST.\noptimum sampling interval determined\nby Wilk and Gray (1970) was 45 minutes.\nObviously, one can't wait that long before making a prediction. Also, the\nlifetime of the storm may be less than an hour. Conversely, if one samples\ntoo frequently, lack of spatial and temporal resolution will produce ficti-\ntiously large errors.\nAnother dilemma arises when matching echo manually. The more frequently\none samples, the easier it is to follow echo motion and to account for splits\nand merges. (On some occasions it was difficult to match five minute data.)\nHowever, an unwarranted amount of time may be spent tracking small cells of\nshort duration which neither produce severe weather nor have sufficient pre-\ndictability to provide meaningful extrapolations.\nCase 3, November 2, 1974\nThe storms developed early in the morning of November 2. On the previous\nday a stationary front extended across southern Texas and northern Louisiana\n(fig. 53) During the day, it evolved into a warm front which moved into\nsouthern Oklahoma (fig. 54). The position of the front changed little during\nthe afternoon and evening of November 2 (fig. 55) as an upper level low\ndeveloped over California, and maintained a stationary pattern of southwest-\nerly flow aloft over Oklahoma.\nAt 500 mb (fig. 56) a broad trough covered most of the United States\nwith two low pressure centers, one centered over northeast Wyoming, the other\nover northern California and Nevada. Between November 1 and 3 the northern\n44","Figure 49. Remote radar display\nFigure 50. Remote radar display\nPPI, 200 km range, 2015 CST,\nwith computer generated warning\nJune 16, 1975.\narea, echo contour and Victor\nairways, June 16, 1975. Predic-\ntion period is 2015 to 2115 CST.\nFigure 51. Remote radar display\nFigure 52. Remote radar display\nPPI, 200 km range, 2030 CST,\nwith computer generated warning\nJune 16, 1975.\narea, echo contour and Victor\nairways, June 16, 1975. Predic-\ntion period is 2030 to 2130 CST.\n45","H\nH\nH\nH\nH\nFigure 54. November 2, 1974, 12Z\nFigure 53. November 1, , 1974, 12Z\nsurface analysis adapted from\nsurface analysis adapted from\nDOC, NOAA, EDS daily weather\nDOC, NOAA, EDS daily weather\nmaps, Weekly Series.\nmaps, Weekly Series.\nlow migrated northeastward into Canada;\nthe other moved southward as a separate\nclosed low. This resulted in a shift\nof the primary trough to a northeast-\nsoutheast orientation and caused the jet\nstream to retrograde westward (figs. 57\nand 58) Since there was insufficient\nfrontal lift and no surface heating,\nL\nthe triggering mechanism for the early\nH\nmorning storms on November 2 was prob-\nably a short wave, produced in the lee\nof the Rocky Mountains. Over central\nOklahoma, mixing ratio values were\n12-14 g kg-1, and low clouds and high\nrelative humidity were widespread\n(fig. 59). At 0600 CST, Oklahoma City\nFigure 55. November 3, 1974, 12Z\nreported a ceiling of 200 ft, Hobart\nsurface analysis adapted from\nand Clinton-Sherman, 150 km to the west,\nDOC, NOAA, EDS daily weather\n200-300 ft ceilings, and Ardmore, 600 ft.\nmaps, Weekly Series.\nAt stations north of Oklahoma City,\nrain and fog were reported.\nSurface winds were generally light 5-10 kt (fig. 59) which was unrepre-\nsentative of the mean flow. Surface mixing was not occurring, as indicated\nby the winds recorded at 444 m on the WKY tower (Goff and Zittel, 1974),\nwhich were 20-30 kts form the south-southeast. This flow more accurately\nreflects the true inflow into the storms.\nTesting of the echo prediction logic began with the 1225Z observation\nand continued until 1210Z. In this case, digital radar data were available\nevery two to three minutes. However, graphic displays were produced as\n46","H\nH\nFigure 56. November 1, 1974, 12Z\nFigure 57. November 2, 1974, 12Z\n500 mb analysis adapted from\n500 mb analysis adapted from\nDOC, NOAA, EDS daily weather\nDOC, NOAA, EDS daily weather\nmaps, Weekly Series.\nmaps, Weekly Series.\nbefore at approximately five minute\nintervals. The thresholds for inten-\nL\nsity and area were 40 dBZ (intensity\nswitch 4) and 150 km2 , respectively.\nH\nThe coding of gray shades for photog-\nraphy was 1220333. The time weight\nconstant was 2400 and the minimum\nrange (to omit ground clutter) was\nL\n20 km.\nBecause the radar pattern changed\nlittle during the 45 minutes of track-\ning, only selected radar PPIs and\ngraphic PPIs are shown for this case\n(figs. 60 through 69).\nFigure 58. November 3, 1974, 12Z\nWarning areas were drawn for a\n500 mb analysis adapted from\none-hour prediction interval starting\nDOC, NOAA, EDS daily weather\nfrom the time of the last observation.\nmaps, Weekly Series.\nIn the first PPI at 1225Z, two\ncells were isolated and labeled cell 1\nand cell 2, respectively. Cell 1, at 250 degrees azimuth and 76 km range,\nwas tracked for the entire 45 minute period. As shown in Table 4, excluding\nthe first prediction, cell motion was southeasterly gradually shifting to the\nnortheast and accelerating. By 1310Z, cell motion was from 220 degrees azi-\nmuth at 53 km hr-1.\nCell 2 split after the first PPI becoming two cells, These merged later\nat 1234Z and tracking of cell 2 was resumed (table 4). This cell, imbedded\nin the line northwest of Oklahoma City, moved from 240 degrees azimuth about\n47","SURFACE MAP TIME 1200 Z DATE 110274\n156110\n200/00\n013\nissue\nFW\nMIX\nIN\nx\nFigure 59. Subsynoptic surface data provided by NWS, FAA and NSSL\nstations, 1200Z. November 2, 1974, with streamlines\nsuperimposed.\n- 1 . At 1245Z when a sudden increase in the line's area occurred,\n60 km hr\ncell 2 was dropped.\nAfter cell 2 split at 1228Z, one of the new cells was labelled cell 3\nand tracked until 1234z. The other cell was never tracked because little\nmotion was shown.\nCell 5 was isolated at the north end of the line at 1230Z and tracked\nfor 15 minutes. During that time it moved from a more southerly direction\n(214 degrees) than the other cells.\nAt 1245Z, a much larger cell was isolated due to the increase in inten-\nsity and size. This cell, cell 6, was tracked until testing ceased at 1310Z.\nAs has been pointed out in previous sections, the large values for 't\nand od (table 4) are in part the result of too frequent sampling. In some\ncases, the graphics examples do not show the high values of ot and od because\nthey exceeded the predetermined limits, as mentioned in case 1.\n48","Table 4. Echo centroid positions, direction and speed predicted,\nand the RMSE values for November 2, 1974.\nECHO\nTIME\nAZIMUTH\nRANGE\nDIRECTION\nSPEED\nod\not\n(KM HR-1)\n(KM)\n(DEGREES)\n(KM)\n(HOUR)\nNO.\n(Z)\n(DEGREES)\n1\n1225\n251.\n76.\n-\n-\n-\n-\n2\n1225\n311.\n95.\n-\n-\n-\n-\n33.1\n1\n1228\n250.\n75.\n303.3\n-\n-\n95.\n3\n1228\n304.\n-\n-\n-\n-\n23.8\n27.41\n0.803\n1\n1230\n251.\n74.\n256.0\n95.\n215.0\n98.8\n3\n1230\n306.\n-\n-\n5\n1230\n349.\n126.\n-\n-\n-\n-\n283.1\n30.8\n43.72\n0.755\n1\n1232\n249.\n73.\n5\n1232\n350.\n127.\n235.1\n72.5\n-\n-\n0.729\n1\n1234\n249.\n71.\n278.3\n36.6\n41.90\n2\n1234\n315.\n92.\n247.7\n47.9\n-\n-\n168.9\n15.0\n15.84\n1.060\n5\n1234\n349.\n127.\n1\n1236\n249\n70.\n274.1\n37.3\n38.72\n0.675\n2\n1236\n317.\n92.\n242.8\n53.3\n20.54\n0.502\n39.77\n0.675\n1\n1238\n250.\n69.\n265.2\n35.6\n59.6\n17.93\n0.619\n2\n1238\n319.\n91.\n242.1\n36.72\n1.329\n5\n1238\n351.\n130.\n214.7\n40.9\n34.6\n37.31\n0.635\n1\n1240\n250.\n68.\n259.9\n0.556\n2\n1240\n320.\n90.\n243.2\n61.7\n18.14\n34.0\n35.21\n0.601\n1\n1242\n250.\n67.\n256.7\n242.5\n64.9\n18.45\n0.546\n2\n1242\n322.\n90.\n35.8\n33.48\n0.618\n1\n1245\n250\n64.\n254.1\n83.\n6\n1245\n311.\n-\n-\n-\n-\n0.593\n63.\n250.5\n36.4\n33.63\n1\n1247\n251.\n74.1\n5\n1247\n310.\n81.\n346.1\n-\n-\n0.599\n1\n1249\n252.\n61.\n246.8\n37.8\n33.45\n62.2\n69.93\n0.944\n6\n1249\n313.\n80.\n268.5\n7\n1249\n353.\n132.\n-\n-\n-\n-\n240.6\n40.84\n0.576\n1\n1251\n255.\n60.\n39.0\n1.099\n271.2\n39.4\n62.41\n6\n1251\n312.\n80.\n39.81\n0.555\n59.\n236.1\n39.9\n1\n1253\n256.\n260.7\n43.3\n58.90\n1.076\n6\n1253\n314.\n79.\n41.75\n0.624\n1255\n259.\n57.\n231.4\n41.9\n1\n41.6\n57.34\n1.015\n6\n1255\n315.\n80.\n248.2\n40.65\n0.677\n54.\n228.9\n44.4\n1\n1257\n260.\n54.6\n57.82\n1.402\n6\n1259\n319.\n80.\n237.8\n46.9\n39.60\n0.685\n227.1\n1\n1259\n262.\n52.\n54.73\n1.313\n79.\n236.7\n59.5\n6\n1259\n320.\n42.70\n0.680\n51.\n226.7\n48.3\n1\n1300\n261.\n57.53\n1.359\n80.\n234.2\n65.0\n6\n1300\n322.\n224.3\n50.2\n50.80\n0.701\n267.\n51.\n1\n1302\n1.292\n69.2\n55.49\n81.\n231.5\n6\n1302\n324.\n51.8\n49.67\n0.685\n269.\n50.\n222.3\n1\n1304\n1.235\n229.2\n72.6\n53.43\n6\n1304\n326.\n82.\n48.48\n0.670\n270.\n49.\n221.0\n52.8\n1\n1306\n1.190\n229.0\n73.3\n52.15\n6\n1306\n327.\n81.\n52.9\n51.53\n0.710\n1\n1308\n267\n47.\n221.1\n1.149\n81.\n229.0\n72.6\n50.14\n6\n1308\n328.\n50.72\n0.699\n46.\n220.9\n53.2\n1\n1310\n270.\n1.143\n69.8\n48.96\n6\n1310\n328.\n80.\n229.8\n49","Figure 60.\nRemote radar display\nFigure 61.\nRemote radar display\nPPI, 200 km range, 1225Z,\nPPI, 200 km range, 1230Z,\nNovember 2, 1974.\nNovember 2, 1974.\nFigure 62. Remote radar display\nFigure 63.\nRemote radar display\nPPI, 200 km range, 1240Z,\nwith computer generated warning\nNovember 2, 1974.\nareas, echo contours and Victor\nairways, November 2, 1974. Pre-\ndiction period is 1240 to 1340Z.\n50","Figure 64.\nRemote radar display\nFigure 65. Remote radar display\nwith computer generated warning\nPPI, 200 km range, 1251Z,\nNovember 2, 1974.\nareas, echo contours and Victor\nairways, November 2, 1974. Pre-\ndiction period is 1251 to 1351Z.\nFigure 67. Remote radar display\nFigure 66. Remote radar display\nwith computer generated warning\nPPI, 200 km range, 1300Z,\nareas, echo contours and Victor\nNovember 2, 1974.\nairways, November 2, 1974. Pre-\ndiction period is 1300 to 1400Z.\n51","Figure 68. Remote radar display\nFigure 69. Remote radar display\nPPI, 200 km range, 1310Z,\nwith computer generated warning\nNovember 2, 1974.\nareas, echo contours and Victor\nairways, November 2, 1974. Pre-\ndiction period is 1310 to 1410Z.\n52","7. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS\nThe conclusions and recommendations included here are divided into three\nareas: A) Phase II objectives, B) software refinements, and C) future tests\nand suggestions for implementation.\nPhase II Objectives\nA.\nAfter consideration of research on storm motion and predictability in\nprogress at NSSL, Lincoln Laboratories, Massachusetts Institute of Technology,\nand the National Weather Service Techniques Development Laboratory, we con-\nclude that statistical techniques for automation of tracking and warning pro-\ncedures are incomplete and probably will require substantial improvements in\nhardware (e.g., more sophisticated radars and signal processing equipment)\nand significant changes in the operational configurations of manpower. We\nbelieve that until such time as our ability to measure and understand severe\nstorm dynamics, a simplified man-machine mix using echo centroids for track-\ning and extrapolating echo motion represents the best technique for issuing\nadvisories at the Flight Service Station.\nWe also believe that the remote terminal described in Phase I is an\nadequate method for remotely displaying radar imagery; and the Phase II study\nhas demonstrated the feasibility of using the R, 0, coordinate system as a\ngraphics terminal. To this end we have developed logic which:\n1. Shows current echo positions and coverage by displaying contoured\nechoes at user defined area and intensity threshold and shows that\nechoes from severe storms can be isolated routinely using an inten-\nsity threshold of 30-40 dBZ and an area threshold of 150-1000 km2.\nThis threshold is sensible because ninety-nine percent of the storms\nwith hail have a maximum intensity above 30 dBZ and moderate turbu-\nlence can be expected in storms of that intensity.\nDisplays a graphic warning area which is derived from echo size and\n2.\nmotion and expanded downstream for a user specified prediction inter-\nval to show a measure of the storm's predictability.\nProvides the user a choice of two computer-generated background\n3.\nreference maps - the State of Oklahoma outline, suitable for a\n400 km range display, and the low level Victor airways, suitable\nfor a 200 km range display. Other maps can be added with only a\nslight impact on the existing logic.\n4. Achieves high visibility and easy interpretation by using different\ngrey shades for the graphics elements. The best results were\nachieved when the warning areas were coded \"bright\", the echo con-\ntours coded \"medium\", and the reference maps coded \"dim\".\nWhat we have not done is to:\n1. Generate alphameric messages on the remote radar display system.\nAlthough possible, this was rejected for three reasons:\n53","a) degradation of the display's resolution as a function of range\nwould require excessively large characters to maintain readability;\nb) the logic would be complex, time consuming and require additional\ncore, and c) low cost, high speed alphameric displays are available\nfor use as a satellite display.\nTransmit simulated real time test cases of the graphics products to\n2.\nthe Wiley Post FSS. However, their personnel were kept informed of\nour work and on occasion did have an opportunity to see a simulated\ntest case at NSSL. Reactions were quite favorable to the graphics\ndescribed above.\nIncorporate growth and decay trends within the graphics. They were\n3.\nfound to be, on a storm-by-storm basis, too short-lived to extrapo-\nlate. General statistics, such as echo coverage, average intensity,\netc., were not acceptable to FSS briefers as guidance information.\nB. Software Refinements\nThe computer techniques used for this report have proved reliable and\nstable and no major modifications are suggested. Small changes will be nec-\nessary to facilitate real time operation and improve aesthetic appearances.\nAlthough developed for an R, O coordinate system, the logic can be adapted\nwith a considerable modification effort to an X,Y type display. Additional\nmodifications of existing logic might include:\n1. Creating a permanent file for the echo shape function so that a\ncontoured echo, regardless of the time of the last entry, will\nretain its shape.\n2. Converting the logic to determine the airports in the paths of\nechoes from a cone to the rectangular area used for generating\nwarning areas. (It may be desirable to perform the airport search\nwithin the program area which generates graphics.)\n3. Routing radar PPI information into the computer for blending with\nbackground reference fields before displaying on the remote terminal.\nThis would require a hardware change and should be switch control-\nlable to avoid computer dependency.\n4. Including the echo's depth when drawing a warning box. The contour\nof large, slow moving echoes may actually circumscribe a box under\nthe existing logic.\nIntroducing the weighting function into the calculations of ot and\n5.\nod. Currently, only the centroid positions are weighted to have\ndecreasing influence with time. However, the errors Et and Ed\nshould also have decreasing influence with time. We know that with\nfew points to estimate echo motion, the errors will be large. How-\never, as more points are added we should not be penalized by the\nlarge earlier errors. Conversely, if having tracked a storm for\n54","awhile, its predictability decreases, we should be made aware of\nthis fact.\n6. Dividing the scope into sectors. Some program speed-up could\nprobably be realized if certain quadrants did not have to be\nscanned for echo.\nFuture Tests and Suggestions for Implementation\nC.\nBased on experience with the three case studies, the author believes\nsome skill is needed to use the echo centroid tracking logic presented in\nthis report. Basic to a successful operation is the selection of area and\nintensity thresholds and sampling intervals for the different types of storm\nsituations (e.g., isolated severe storms, squall lines, and slow moving\nflood producers). We believe that some training of FSS personnel will be\nneeded.\nA second consideration is the configuration between man and machine.\nOne possibility is to have one person responsible for identifying the haz-\nardous storms and operating the echo tracking logic. Then, all user (pilot\nbriefer) requests are channeled to the \"hazards\" briefer who is cognizant\nof the complete echo pattern. A second possibility is to allow each pilot\nbriefer access to a computer terminal which lists input and output of echo\nlocations and velocities (e.g., echoes which have been assigned user num-\nbers for tracking could be so labeled).\nOne very important question which needs to be answered is how much time\nis needed for decision making. Not only initially when establishing thresh-\nolds, but during operation when matching echoes manually. This, of course,\nwill vary with radar patterns of differing complexity. (The author had the\nadvantage of being able to study the numerical results at leisure, but was\nhandicapped by not having a simultaneous radar display for comparison.)\nFor the above reasons, a limited operational test is needed. This test\nshould be divided into two sections of one month's duration each during a\nstorm season. The first half of the test would be to prepare software logic\nfor operational use and to test (a) the feasibility of operationally using\nthe existing logic, and (b) determining the optimum man-machine configuration.\nIt should include sufficient hardware (e.g., remote alphameric terminal con-\nnected to a time share larger computer) to allow efficient operation of the\nECHOPRED logic without major modifications.\nIf the outcome of the first test is favorable, the system should then be\ninstalled at a Flight Service Station, such as Wiley Post, for operational\ntest and evaluation by FSS staff.\n55","ACKNOWLEDGMENTS\nThe author wishes to especially thank Mr. Kenneth E. Wilk for his\nsuggestions and helpful discussions during all portions of this work\nand for his careful review of this report; also Mr. James Muncy, FAA\nTechnical Officer, who supported this work.\nThe author appreciates the dedicated work of NSSL's supporting staff\nin the preparation of this report.\nREFERENCES\nBarclay, P. A., and K. E. Wilk, 1970: Severe thunderstorm radar echo\nmotion and related weather events hazardous to aviation operations.\nESSA Tech. Memo. ERLTM-NSSL 46, 63 pp.\nBigler, S. G., 1969: Dial A Radar. Bulletin, AMS, 50, 6, 428-430.\nBlackmer, R. H., Jr., and R. 0. Duda, 1972: Application of pattern recog-\nnition techniques to digitized radar data. Proceedings, 15th Radar\nMet. Conf., AMC, Boston, Mass.; 138-143.\nGoff, R. C., and W. D. Zittel, 1974: The NSSL/WKY-TV tower data collec-\ntion program: April-July 1972. NOAA Tech. Memo. ERL NSSL-68, 45 pp.\nKessler, E., and J. T. Russo, 1963: Statistical properties of weather\nradar echoes, Proceedings, Wea. Radar Conf., AMS, Boston, Mass., ,\n25-33.\nMorris, M., and 0. E. Brown, 1937: Analytic Geometry. McGraw-Hill Book\nCo., New York, 144 pp.\nOstlund, S. S., 1974: Computer software for rainfall analyses and echo\ntracking of digitized radar data, NOAA Tech. Memo. ERL-WMPO-15, 82 pp.\nSirmans, D., and R. J. Doviak, 1973: Meteorological radar signal inten-\nsity estimation. NOAA Tech. Memo. ERL NSSL-64, 80 pp.\nWilk, K. E., 1966: Motion and intensity characteristics of the severe\nthunderstorms of April 3, 1964. ESSA IERTM-NSSL 29, 9-21.\nWilk, E. E., and K. C. Gray, 1970: Processing and analysis techniques\nused with the NSSL weather radar system. Proceedings, 14th Wea.\nRadar Conf., AMS, Boston, Mass., 369-374.\nWilson, J. W., 1966: Movement and predictability of radar echoes, ESSA\nTech. Memo. ERLTM-NSSL 28, 30 pp.\n56","APPENDIX A\nAnnotated Command Structure, June 6, 1975\nCOMMAND\nTWC\nHHMM\n2400\nCOMMAND\nGCD\n20\nCOMMAND\nACC\nAREA/IVTEVSITY\n250\n4\nDAY / TIME/ TILT\n157\n2040\n0\n2 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n1 will try tracking only very strong\n136 cell 1 - new\n29\n988\n157 cell 2 - new 1 (VST) cells this run\n335\n1907\n2 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\nignore severe cells imbedded in the\n327\n161\n501\nVST cells above\n155\n359\n340\nDAY / TIME/ TILT\n157 2045\ne\nSTOP\n2 VST ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n137 cell 1 - some growth evident\n32\n1127\n153 cell 2\n1974\n335\n2 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n154\n407\n328\n149\n496\n338\n57","COMMAND\nENT\nV/ HHMY/ AZM/ RNG/ DIRECTION/ SPEED/\nSTDDS/\nSTDTM\n1 2040\n29. 136.\n1 2045\n32. 137.\n292.5\n92.9\n0.00\n0.000\n2\n2040\n335.\n157.\n2\n2045\n335.\n153.\n335.0\n51.5\n0.00\n0.000\n0\n0\ne.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n2100\n2200\n2ld\nVIC\n1\n2\n0\n0\n0\n0\n0\n0\n0\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME/ TILT\n157 2050\n0\nSTOP\n4 VST ECHOES FOUND WITH AREA GREATER THAN 250 SC\nKM\nAREA/AZIMUTH/RANGE\n1182\n34\n140 cell 1\n353\n61\n188 cell 3 - new\nSno assignment, wait to see if cell\n304\n310\n163\n(will be retained in next PPI.\n1884\n336\n147 cell 2\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n943\n335 144\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/\nSPEED/\nSTDCS/\nSTDTM\n1 2050 34. 140.\n283.1\n78.7\n14.36\n0.182\n2 2050 336. 148.\n319.0\n58.2\n17.90\n0.098\n0\n0\ne.\n0.\n58","COMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n0\n0\n1\n2\n0\n0\nD\n0\n0\n200\nVIC\n2105\n2205\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME/ TILT\n157 2055\n0\nSTOP\n3 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n145 cell 1\n1225\n36\n188 cell 3\n388\n61\n(no assignment as several small cells in line\n2509\n324\n147\n(merged, may be result of radar power fluctuation\nSEV ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\n1\nAREA/AZIMUTH/RANGE\n336 140\n895\nCOMMAND\nENT\nN/ HHMM/ AZMI RNG/ DIRECTION/ SPEED/ STDCS/\nSTDTM\n1 2055\n36.\n144.\n0.162\n276,9\n75.9\n16.73\n3 2050\n61.\n188.\n3 2055\n61.\n188.\n360.0\n0.0\n0.00\n0.000\n0\n0\n0.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO\nNUMBERS\n0\nG\n0\n0\n0\nVIC\n1\n3\n2\n0\n211l\n2210\n220\n59","COMMAND\nACC\nAREA/INTENSITY\n25l\n4\nDAY / TIME/ TILT\n157 2100\ne\nSTOP\n5 VST ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n146cell 1\n1077\n38\n189cell 3\n34l\n61\n194cell 4\n357\n298\n155cell 6\n561\n310\n136cell 2 - centroid position consistent with last cell 2\n1656\n335\nposition\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n954\n335\n133\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/\nSPEED/\nSTDCS/\nSTDTM\n1 2120 38. 146.\n277.1\n73.1\n15.58\n0.159\n2 2100 335. 136.\n334,2\n64.3\n23.26\n2.102\n-3 2055 61. 188.\n3 2100 61. 189.\n241.0\n6.1\n0.00\n0.000\n0\n0\ne.\nl.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO\nNUMBERS\n1\n2\n3\n0\n0\n0\n0\ne\n0\n2115\n2215\n2ld\nVIC\n60","COMMAND\nACC\nAREA/INTENSITY\n25l\n4\nDAY / TIME/ TILT\n157 2105\n0\nSTOP\n6 VST ECHOES FOUND WITH AREA GREATER THAN 250 SG KM\nAREA/AZIMUTH/RANGE\n156 icell 1 split, wait to see if split remains before making\n762\n39\n(new assignment\n252\n45\n122\n187cell 3\n432\n60\n192cell 4 - rapid growth but centroid position consistent\n600\n297\n152cell 6\n586\n308\n132cell 2 - growth evident, but centroid position is\n1946\n336\nconsistent\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n333 127\n1087\nCOMMAND\nENT\nN/ HHMM/ AZMI RNG/ DIRECTION/ SPEED/ STDCS/ STDTM\n2 2105 336. 132.\n331.8\n62.8\n24.05\n0.142\n3\n2105\n60,\n187.\n137.5\n11.7\n22.61\n2.883\n4\n2100\n298.\n194.\n6\n2100\n310.\n155.\n4\n2105\n297,\n192.\n0.000\n356.8\n49.0\n0.00\n6\n2105\n308.\n152.\n0.000\n9.8\n76.7\n0.00\n0\n0\ne,\nD.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\ne\n0\n6\n0\n0\n2120 2220 220\n2\n3\n4\n61","COMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME/ TILT\n157 2110\n0\nSTOP\n4 VST ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n162 no assignment, still not convinced echo 1 won't reappear\n879\n40\n187 cell 3\n531\n59\n188 cell 4\n772\n296\n124 cell 2\n1854\n332\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n726\n337 123\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/\nSTDCS/\nSTOTM\n3 2110 59. 187.\n138.9\n19.7\n20.00\n2.746\n4 2110 296. 188.\n345.0\n55.0\n12.69\n0.110\n2 2110 333. 124.\n339.1\n66.1\n40.99\n0.230\n0\n0\n2.\nl.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n2125 2225 200\nVIC\n1\n2\n3\n4\n0\n10\n0\n0\n0\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME/ TILT\n157 2115\n0\nSTOP\n62","4 VST ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\narea is consistent with last cell 1 area and is\n162cell 1\n1060\n44\nonly one in the area, rapid movement is indicated\n188cell 3\n497\n61\n184cell 4\n815\n295\n(cell 6 and smaller part of cell 1 split\n118cell 2\n1868\n333\ndropped)\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n797\n337 117\nCOMMAND\nENT\nN/ HHMM/ AZMI RNG/ DIRECTION/ SPEED/\nSTDCS/\nSTDTM\n1 2115\n44,\n162.\n272.1\n79.2\n15.85\n0.171\n3\n2115\n61.\n188.\n135,5\n7.1\n17.95\n3.333\n4\n2115\n295,\n184.\n340.6\n57.4\n12.34\n0.109\n2 2115 333. 118.\n341,2\n68.1\n38.06\n0.215\n0\n0\nl.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n2130\n2230\n220\nVIC\n1\n3\n4\n2\ne\n0\nD\n0\n0\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME/ TILT\n157 2120\n0\nSTOP\n63","5 VST ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n160 cell 1.\n1236\n48\n189 cell 3\n452\n60\n178 cell 4\n674\n295\n131 cell 5 - possibly old cell 6, but just as easy to reassign\n323\n305\n115 cell 2 I. some area decrease, but centroid position\n1699\n334\nconsistent\n1 SEV ECHOES FOUND W ITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n608\n334\n107\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/\nSTDCS/\nSTDTM\n1\n2120 48. 16l.\n277.9\n80.7\n44.36\n0.159\n3\n2120\n60,\n189.\n155.2\n6.2\n18.65\n3.435\n4 2120 295. 178.\n329.4\n57.6\n25.53\n0.112\n2\n2120\n334,\n115.\n340.5\n66.6\n37.05\n0.274\n0\n0\nl.\nl.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n2135\n2235\n2lg\nVIC\n1\n3\n4\n2\n£\n0\nD\n0\n0\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY/ TIME/ TILT\n157 2125\n0\nSTOP\n64","6 VST ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n377\n13\n189 cell 7 - new\n1421\n50\n165 cell 1 - growth trend indicated\nconsistent\n191 cell 3 - large area decrease, but centroid position/\n267\n60\n166 cell 4) frapid growth observed in both cells, but\n908\n296\n126 cell 5f (centroid positions consistent\n668\n305\n105 cell 2\n1639\n332\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n637\n336 104\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/\nSTDDS/\nSTDTM\n5 2120\n305.\n131.\n1 2125\n50.\n165.\n279.9\n82.1\n41.50\n0.155\n3 2125\n60.\n191.\n184,5\n6.5\n19.48\n3,190\n4 2125 296. 166.\n312.9\n66.7\n50.69\n8.313\n5\n2125\n305.\n126.\n325,0\n61.0\n0.00\n0.000\n2\n2125\n332.\n105.\n341.5\n69.1\n36.33\n0.389\n0\n0\n0.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n2140\n2240\n220\nVIC\n1\n3\n4\n5\n2\n0\nis\n0\n0\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME/ TILT\n157 2130\ne\nSTOP\n65","5 VST ECHOES FOUND WITH AREA GREATER THAN 250 SG KM\nAREA/AZIMUTH/RANGE\n187 cell 7 - rapid growth from previous PPI\n12\n515\n173 cell 1 - growth trend reversing\n50\n1376\n163 cell 4\n295\n935\n122 cell 5\n304\n644\n(cell 3 dropped by program)\n101 cell 2\n332\n1732\n2 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SG KM\nAREA/AZIMUTH/RANGE\n57 156\n260\n333\n93\n624\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/\nSTDCS/\nSTDTM\n13, 189.\n7 2125\n12, 187.\n7 2130\n0.00\n0.000\n71.1\n46.1\n173.\n1 2130\n5l.\n46.15\n2.167\n278.3\n81.8\n4 2130 295, 163.\n47.67\n0.316\n310,0\n68.5\n5 2130 304. 122.\n14.75\n0.121\n318.3\n55.8\n2 2130 332, 101.\n341.7\n69.5\n34.56\n0.383\n0\ne.\n0.\n0\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\nR\n0\n5\n2\n0\n0\n4\n2145 2245 4lg STA\n7\n1\nCOMMAND\nBYE\n66","APPENDIX B\nAnnotated Command Structure, June 16, 1975\nCOMMAND\nTWC\nHHMM\n2400\nCOMMAND\nGCD\n20\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME / TILT\n167\n1830\n0\n1 VST ECHOES FOUND WITH AREA CREATER THAN 250 SG KM\nAREA/AZIMUTH/RANGE\n188 cell 1 - new {will track VST cells this run\n282\n1795\n1 SEV ECHUES FOUND WITH AREA GREATER THAN 250 SG KM\nAREA/AZIMUTH/RANGE\n1330\n281\n189 ignore SEV cells imbedded in the VST cells above\nDAY / TIME / TILT\n167 1835\n2\nSTOP\n1 VST ECHOES FOUND WITH AREA GREATER THAN 25% SG KM\nAREA/AZIMUTH/RANGE\n186 cell 1\n2221\n279\n1 SEV ECHOES FOUND WITH AKLA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n279 186\n1744\n67","CUMMAND\nENT\nSTDIM\nV/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/\nSTCOS/\n1 1830\n288.\n188.\n1 1835 279, 186.\n0.900\n3.00\n338.0\n45.9\n0\nis\nE.\nD.\nCOMMAND\nDIS\nBTIME/ETIME/RAVGE/UVERLAY/ECHO NUMBERS\n2\n0\nis\nD\na\n0\n1835 1935 200\nVIC\n1\n(I)\nCOMMAND\nACC\nAREAVINTENSITY\n259\n4\nDAY / TIME/ TILT\n167 1840\n0\nSTOP\n1 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n278 183 cell 1 - growth due to squall line entering PPI\n2648\n2 SEV ECHOES FOUND WITH AREA CREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n1429\n269\n188\n482\n293\n185\nCOMMAND\nENT\nSTDTM\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/\nSTODS/\n1 184 2 28. 183.\n6.59\n0.070\n331.3\n49.3\n0.\n0\n0\n2.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\ne\n0\na\nis\n184l 1940 200\nVIC\n1\n&\n0\nis\n68","COMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME/ TILI\n167 1845\ne\nSTOP\n1 VST ECHOES FOUND WITH AREA GREATER THAN 250 Sq KM\nAREA/AZIMUTH/RANGE\n179 cell 1 - line still growing\n3067\n279\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n27l 181\n1661\nCOMMAND\nEVT\nSTDTM\nHHMM/ AZM/ RNG/ DIRECTION/ SPEED/ STCDS/\nN/\n1 1845 279, 179,\n31.11\n0.516\n302.5\n39.2\n2\n0\n2.\nD.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\na\nis\n2\n1845 1945 200\nVIC\n1\n0\nis\nis\nCOMMAND\nACC\nAREA/IVTENSITY\n250\n4\nDAY / TIME/ 11LT\n167\n1856\n2\nSTOP\n69","2 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n174 cell 1\n(cell split on north end of line, will\n2863\n276\n186 no assignment wait to see if small cell is retained)\n265\n381\n1 SEV ECHOES FOUND WITH AREA CREATER THAN 254 SC KM\nAREA/AZIMUTH/RANGE\n1636\n268 179\nCOMMAND\nENT\nN/ HHMM/ AZM RNG/ DIRECTION/ SPEED/ STCDS/ STOTM\n1 1852 276, 174.\n314.0\n52.1\n43.65\n0.843\n&\n2.\n1)\nis.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHU NUMBERS\n1852\n1950\n248\nVIC\n1\nI\n0\nD\n0\na\n0\ne\nCOMMAND\nACC\nAREA/ITENSITY\n25l\n4\nDAY / TIME / TILT\n167 1855\nI\nSTOP\n1 VST ECHUES FOUND WITH AREA GREATER THAN 25% SG KM\nAREA/AZIUTH/RANGE\n2675\n171cell 1 - area decreasing, small cell dropped\n273\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIUTH/RANGE\n1362\n266 172\nCOMMAND\nENT\nN/ HHMM/ 1211 RNG/ DIRECTION/ SPEED/ STCDS/ STOIM\n1\n1855\n273.\n171.\n324.0\n63.3\n46.94\n0.845\n0\nI\nCo\n0.\n70","COMMAND\nDIS\nBTIML/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n1855 1 955 202 VIC\n1\n&\n0\n0\n0\n0\nis\nX\nCOMMAND\nACC\nAREA/INTENSITY\n258\n4\nDAY/ TIME/ TILT\n167\n1980\na\nSTOP\n1 VST ECHOES FOUND WITH AREA GREATER THAN 250 SG KM\nAREA/AZIMUTH/RANGE\n2505\n165 cell 1 - area still decreasing, centroid position\n271\nconsistent\n1 SEV ECHUES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n1274\n266\n167\nCOMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/\nSTLOS/\nSTDTA\n1 1999 271. 165,\n325.4\n71.5\n43.74\n0.812\n0\n0 E. 0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n1928 2000 202\nVIC\n1\n0\n0\n0\n0\n(A\nR\nI\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY /\nTIME/ TILT\n167\n1985\nV\n71","2 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZINUTH/RANGE\n2238\n262\n170 cell 2 - new shift in centroid position significant\n360\n287\n151 cell 3 - new (with this split, redefine echoes\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n1449\n263\n166\nDAY / TIME/ TILT\n167 1910\na\nSTOP\n2 VST ECHOES FOUND WITH AREA GREATER THAN 252 SG KM\nAREA/AZIMUTH/RANGE\n2380\n167cell 2 - area and centroid position consistent\n261\n262\n285\n142cell 3 - (obvious from PPI same cell as above\nrapid motion indicated)\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n1626\n261 163\nCOMMAND\nENT\nv/\nHHMM/ AZMI RNG/ DIRECTION/ SPEED/\nSTCOS/\nSTDTM\n2\n1905\n262.\n170.\n2\n1910\n261.\n167.\n305,9\n55,2\n9.00\n0.000\n3\n1905\n287.\n151.\n3\n1910 285. 142.\n315.6\n135.9\n0.00\n0.000\nECHO SPEED GREATER THAN 120 KM/HK, CHECK THIS CB.\n&\nis\n2.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n1912 2010 200\nVIC\n2\n3\nis\nis\n0\n0\n0\n72","COMMAND\nACC\nAREA/INTENSITY\n25l\n4\nDAY / TIME / TILT\n167 1915\n&\nSTOP\n2 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ\nKM\nAREA/AZIMLTH/RANGE\n2431\n159 cell 2\n261\n133 cell 3 - area lost in previous PPI regained\n364\n287\n1 SEV ECHOFS FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n1482\n261 156\nCOMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/\nSPEED/\nSTCDS/\nSTOTM\n2\n1915\n261.\n159.\n276.1\n69.7\n39,15\n0.134\n3\n1915\n287,\n133.\n287.0\n110.4\n59.72\n0.247\n0\n0\nl.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n1915 2015\n200\nVIC\n2\n3\n(I)\nis\n0\n0\nis\n&\nCOMMAND\nACC\nAREA/INTENSITY\n25l\n4\nDAY / TIME/ TILT\n167 1920\nV\nSTOP\n73","1 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n2792\n146 cell 1 - (cells 2 and 3 merged, extrapolated position\n264\nfor old cell 1 appears reasonable)\n2 SEV ECHUES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n574\n254\n169\n591\n265\n136\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/\nSTCDS/\nSTDIM\n1 1920 264, 146,\n321,3\n77.7\n41.41\n0.761\nK\nI\n8.\nD.\nCOMMAND\nDIS\nBTIME/ETIM/RANGE/OVERLAY/ECHO NUMBERS\n1922 2022\n202\nVIC\n1\nD\n0\n0\nD\n0\n0\ne\nCOMMAND\nACC\nAREA/INTENSITY\n25l\n4\nDAY / TIME / TILT\n167 1925\nE\nSTOP\n1 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n2757\n263\n140cell 1\n2 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SQ KN\nAREA/AZIMUTH/RANGE\n518\n253\n166\n435\n272\n126\n74","COMMAND\nENT\nN/ HHMM/ 4 2M / RNG/ DIRECTION/ SPEED/\nSTCDS/\nSTOTM\n1 1925 263. 140.\n313.7\n78.4\n41.81\n0.720\n10\ng\nD.\nD.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ELHO NUMBERS\n1925\n2025\n200\nVIC\n1\n0\n0\n0\nD\n0\n6\n&\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME / TILT\n167\n1930\n0\nSTOP\n1 VST ECHCES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n132cell 1 - area starting to decrease\n2562\n263\n2 SEV ECHOES FOUND WITH AREA GREATER THAN 253 SQ KM\nAREA/AZIMUTH/RANGE\n555\n252\n155\n442\n269\n119\nCOMMAND\nENT\nN/\nHHMM/ AZM/ RNG/ DIRECTION/ SPEED/\nSTCDS/\nSTCTM\n1 1930 263. 132.\n315.3\n77.8\n46,93\n3.690\n0\nis\nD.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n1930 2030\n200\nVIC\n1\nD\nx\n0\ns\nis\n2\n75","COMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY / TIME / TILT\n167 1935\n0\nSTOP\n1 VST ECHOES FOUND WITH AREA GREATER THAN 252 SC KM\nAREA/AZIMUTH/RANGE\n2254\n26l\n126 cell 1 - area decreasing rapidly\n2 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SC KM\nAREA/AZIMUTH/RANGE\n522\n251\n150\n525\n268 111\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/\nSTCDS/\nSTDTM\n1 1935 260. 126.\n313.3\n78.7\n44,85\n0.668\n0\n&.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n1935\n2035\n202\nVIC\n1\nis\n0\nD\nD\n0\nW\n8\nCOMMAND\nACC\nAREA/INTENSITY\n250\n4\nDAY/ TIME / TILT\n167 1940\nP.\nSTOP\n1 VST ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUTH/RANGE\n120 cell 1\n2390\n258\n76","2 SEV ECHOES FOUND WITH AREA GREATER THAN 250 SQ KM\nAREA/AZIMUIH/RANGE\n146\n476\n251\n614\n267\n103\nCOMMAND\nENT\nSTDEN\nN/ HHM AZM / RNG/ DIRECTION/ SPEED/\nSTOOS/\n1\n1940 258. 120.\n43,72\n9.640\n311.7\n79.3\nP.\na.\n0\nI\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\nis\nis\n194l 2040 200\nVIC\nl\nit\nis\nis\nv)\nCOMMAND\nACC\nAREA/IVTENSITY\n250\n4\nDAY / TIME/ TILT\n167 1945\n0\nSTOP\nI VST ECHOES FOUND wiTH AREA GREATER THAN 250 SG KM\nAREA/AZIMUTH/RANGE\n259 111 cell 1\n2246\n2 SEV ECHCES FOUND WITH ARE4 GREATER THAN 25% SQ KM\nAREA/AZIMUTH/RANGE\n417\n245\n138\n367\n267\n95\nCOMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/ STCDS/\nSTDTM\n1 1945 259. 111.\n309.0\n79.1\n50.48\n0.626\n0\n2\nK.\n0.\n77","COMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n1945\n2045\n200\nVIC\n1\nG\n0\nis\nV\n0\n&\nx\nCOMMAND\nACC\nAREA/INTENSITY\n1800\n3\nDAY / TIME / UIT\n2 (cell 1 broke into several cores after this time;\n167\n1946\na larger area at a lower intensity was used after this\ntime)\n2 STR ECHOES FOUND WITH AREA GREATER THAN1000 SG KM\nAREA/AZIMUTH/RANGE\n9510\n2491\n354\n115 this data taken at long range, no assignment made\n4848\n263\n1 VST ECHOES FOUND WITH ARLA GREATER THAN1000 SC KM\nAREA/AZIUTH/RANGE\n3962\n345 246\nDAY / TIME / TILT\n167\n1952\na\n1 STR ECHOES FOUND WITH AREA GREATER THAN1000 SQ KM\nAREA/AZIMUTH/RANGE\n112cell 4 - new\n4745\n263\nDAY/ TIME / TILT\n167\n1955\n&\nSTOP\n1 STR ECHOES FOUND WITH AREA GREATER THAN1000 SQ KM\nAREA/AZIMUTH/RANGE\n108 cell 4 - indicated motion unusually slow\n4743\n264\n78","COMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/ STCDS/ STDTM\n4\n1950 263. 110.\n4\n1955\n264.\n128.\n219.9\n36.2\n0.00\n0.001\n&\nA\nV.\n0.\nCOMMAND\nDIS\nBTIME/FTIME/RANGE/OVERLAY/ECHO NUMBERS\n1955 2055 200\nVIC\n4\n(A\nD\n19\n0\n0\nD\n&\nCOMMAND\nACC\nAREA/IVTENSITY\n1000\n3\nDAY / TIME / TILT\n167 2000\ne\nSTOP\n1 STR ECHOES FOUND WITH AREA GREATER THAN1000 SQ KM\nAREA/AZIMUTH/RANGE\n97cell 4 - - (centroid position change very large,\n4849\n267\necho seems unstable)\nCOMMAND\nENT\nV/\nHHMM/ AZM / RNG/ DIRECTION/ SPEED/\nSTCDS/\nSTETM\n4 2000 267,\n97,\n235.9\n91.2\n28.38\n1.627\n&\nis\nE.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n2222 2100 200 VIC 4\nD\n0\ng\n0\na\nD\n&\nCOMMAND\nACC\nAREA/INTENSITY\n1200\n3\nDAY / TIME / UILT\n167\n20x5\ne\n79","2 STR ECHOES FOUND WITH AREA GREATER THAN1000 SQ KM\nAREA/AZIMUTH/RANGE\nchange\n4345\n263\n82 cell 5 - new {significant area and centroid position/\n1368\n294\n155 no assignment - large patch of stratiform rain\nDAY / TIME/ TILI\n167\n2010\n2\nSTOP\n2 STR ECHOLS FOUND WITH AREA GREATER THAN1000 SQ KM\nAREA/AZIMUTH/RANGE\n4319\n263\n74 cell 5 - area and centroid position consistent\n1322\n295\n155 no assignment - little motion from previous PPI\nCOMMAND\nENT\nN/\nHHMM/ AZM / RNG/ DIRECTION/ SPEFD/\nSTCDS/\nSTCTM\n5 2005\n263.\n82.\n5\n2016\n263.\n74.\n263.10\n1x5.0\n0.00\n0.000\nx\n2.\nis\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVEKLAY/EC HC NUMBERS\n2010 2110\n200\nVIC\n5\n0\n0\n0\n&\nD\n0\n&\nCOMMAND\nACC\nAREA/INTENSITY\n1022\n3\nDAY / TIME/ TILT\n167 2015\n0\nSTOP\n2 STR ECHOES FOUND WITH AREA GREATER THAN1000 SQ KM\nAREA/AZIMUTH/RANGE\n4186\n64 cell 5 - rapid motion indicated\n263\n1434\n296\n153 no assignment - see above\n80","COMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/\nSTCDS/\nSTDTM\n5 2015 263, 64.\n263.0\n110.4\nD.ll\n0.082\n10\nD l. 0. .\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n2015 2115 208 VIC\n5\n0\n0\n'D\n&\nis\nis\nR\nCOMMAND\nACC\nAREA/INTENSITY\n1022\n3\nDAY / TIME / TILT\n167\n2020\n0\n2 STR ECHOES FOUND WITH AREA GREATER THAN1000 SG KM\nAREA/AZIMUTH/RANGE\ngrowth on north end of line caused\n5162\n285\n51 cell 6 - newf\nsignificant shift in centroid and area\n1543\n296\n145 no assignment\n1 VST ECHOES FOUND WITH AREA GREATER THAN1000 SQ KM\nAREA/AZIMUTH/RANGE\n1250\n248\n82\nDAY / TIME/ TILT\n167 2025\ne\nSTOP\n2 STR ECHOES FOUND WITH ARLA GREATER THAN1000 SQ KM\nAREA/AZIMUTH/RANGE\n5203\n284\n46 cell 6 - area and centroid position consistent\n1442\n297\n148 no assignment\n1 VST ECHOES FOUND WITH AREA GREATER THAN 1000 SQ KM\nAREA/AZIMUTH/RANGE\n1247\n238 77\n81","COMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/ STCDS/ STDTM\n6 2020\n285.\n51.\n2025\n284.\n46.\n6\n0.00\n0.000\n294.1\n66.6\n0\nis\nl.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n0\nis\n2025 2125\n200\nVIC\n6\n0\nD\n0\na\nCOMMAND\nACC\nAREA/INTENSITY\n1000\n3\nDAY / TIME / TILT\n167 2038\n0\nSTOP\n2 STR ECHOES FOUND WITH AREA GREATER THAN1000 SG KM\nAREA/AZIMUTH/RANGE\n41 cell 6 - some decrease in area, centroid position\n287\n487€\n1778\n388\n145 no assignment\nconsistent\nCOMMAND\nENT\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/ STCDS/ STCTM\n6 2030 287.\n41.\n276.4\n62.1\n20.74\n0.099\n0\n0\nl.\nD.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n0\nis\n2030 2130\n200\nVIC\n6\nx\n0\nL\nV\nCOMMAND\nBYE\n82","APPENDIX C\nAnnotated Command Structure, November 2, 1974\nCOMMAND\nTWC\nHHMM\n2400\nCOMMAND\nGCD\n20\nCOMMAND\nACC\nAREA/IVTEVSITY\n150\n4\nDAY/ TIME / TILT\n306 1225\n0\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n154\n251\n76 cell 1 - new\n540\n311\n95 cell 2 - new\nDAY/ TIME / TILT\n306 1228\n0\n3 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n219\n25l\n75 cell 1\n}\n189\n304\n95 cell 3 - new\nSplit of cell 2, skip 4 in sequence,\n322\n321\n94 no assignment\nmay use later\nDAY / TIME/ TILT\n306\n1230\n0\nSTOP\n4 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n225\n251\n74 cell 1\n179\n95 cell 3\n306\n94 no assignment (no motion from previous time)\n337\n321\n222\n126 cell 5 - new\n349\n83","COMMAND\nENT\nV/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/ STCDS/\nSTCIM\n1 1225\n251.\n76.\n1\n1228\n25l.\n75.\n303.3\n33.1\n0.20\n0.000\n1\n1230\n25L.\n74.\n256.0\n23.8\n27.41\n0,803\n3\n1228\n384.\n95.\n3\n1230\n386.\n95.\n215.0\n98,8\nD.DD\n2,000\n&\n0\nK.\nlo\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n123l 1330 202 VIC\n1\n3\n0\ns\n0\n8\nki\n&\nl\nCOMMAND\nACC\nAREA/INTENSITY\n15l\n4\nDAY / TIME / TILT\n386\n1232\n0\n4 VST ECHOES FOUND WITH AREA GREATER THAN 150 56 KM\nAREA/AZIMLTH/RANGE\n21l\n249\n73cell 1\n194\n3et\n94cell 3\n364\n321\n93no assignment\n193\n352\n127cell 5\nDAY / TIME/ UIT\n3ee\n1234\n&\n3 VST ECHOES FOUND WITH AREA GREATER THAN 150 SG\nKM\na\nAREA/AZIMUTH/RANGE\n71cell 1\n192\n249\n(merge of cell 3 with unassigned cell,\n55l\n92cell 2\n315\ncentroid position consistent with old cell 2)\n127cell 5\n216\n349\nDAY / TIME/ TILT\n306\n1236\ne\nSTOP\n84","2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\nrecell 1\n249\n198\n92cell 2\n558\n317\nCOMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/ STCDS/ STCTM\n1 1232 249. 73.\n283.1\n30.8\n43.72\n0.755\n1 1234 249.\n71.\n278.3\n36.6\n41.90\n0.729\n1 1236 249. 70.\n0.675\n274.1\n37.3\n38.72\n5 1230 349. 126.\n5 1232 35l. 127.\n0.000\n235.1\n72.5\n0.80\n5 1234 349. 127.\n168.9\n15.0\n15,84\n1.060\n2 1225 311.\n95.\n2 1234 315.\n92.\n247.7\n47.9\n0.20\n0.000\n2 1236 317,\n92.\n242.8\n53.3\n20.54\n0.502\n0.\n&\nis\nk.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO\nNUMBERS\n0\n0\n0\nis\nR\nl\n1\n2\n5\n1236\n1336\n202\nVIC\nCOMMAND\nACC\nAREA/INTENSITY\n152\n4\nDAY / TIME/ TILT\n306\n1238\n&\n3 VST ECHOES FOUND WITH AREA GREATER THAN 150 SG KM\nAREA/AZIMUTH/RANGE\n69cell 1\n196\n258\n551\n319\n91cell 2\n130cell 5\n152\n351\n85","DAY / TIME / TILT\n306 1240\n&\nSTOP\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 S6 KN\nAKEA/AZIMUTH/RANGE\n216\n25l\n68\n546\n328\n9l\nCOMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/\nSTCDS/\nSTCTM\n1 1238 25l.\n69.\n265.2\n35.6\n39.77\n0.675\n1 1240 25l.\n68.\n259.9\n34.6\n37,31\n0.635\n2 1238 319.\n91.\n242.1\n59.6\n17.93\n0.617\n2\n1240\n322.\n90.\n243.2\n61.7\n18,14\n0,550\n5 1238 351. 130.\n214.7\n40.9\n36.72\n1.329\n0\n0\ne.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n124l 1340 200\nVIC\n1\n2\n5\n0\n0\nl\nD\nis\ne\nCOMMAND\nACC\nAREA/INTENSITY\n15l\n4\nDAY / TIME / TILT\n326\n1242\n&\n3 VST ECHOES FOUND WITH AREA GREATER THAN 150 S6 KM\nAREA/AZIMLTH/RANGE\n67cell 1\n194\n250\n86 no assignment\n159\n284\n92 cell 2\n581\n322\n86","DAY / TIME / TILT\n386 1245\ne\nSTOP\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 S6 KM\nAREA/AZIMUTH/RANGE\n64cell 1\nseveral cells merged, may be result of\n17l\n258\n83cell 6 - new\nincrease in intensity or radar power\n944\n311\nfluctuation\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SG KM\nAREA/AZIMUTH/RANGE\n157\n327\n92\nCOMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEEDA\nSTCDS/\nSTEIM\n1 1242 25l.\n67.\n256.7\n34.0\n35.21\n0.601\n1\n1245\n25l.\n64,\n254.1\n35.8\n33.48\n0.618\n2\n1242\n322.\n90.\n242.5\n64.9\n18.45\n8.540\n6\n1245\n311,\n83.\nk\nD\nl.\nl.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\n1245 1345 202 VIC\n1\n2\n0\n0\n0\n0\nis\n&\ne\nCOMMAND\nACC\nAREA/INTENSITY\n150\n4\nDAY / TIME / TILT\n306\n1247\ne\n87","2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n63 cell 1\n157\n251\narea loss in both cells, but centroid positions\n81 cell 6,\n863\n310\nconsistent\nDAY / TIME / TILT\n306\n1249\ne\n3 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n61 cell 1\n153\n252\n80 cell 6\n85l\n313\n130cell 7 - new\n197\n353\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SG KM\nAREA/AZIMUTH/RANGE\n225\n325\n84\nDAY / TIME/ TILT\n306\n1251\ne\nSTOP\n3 VST ECHOES FOUND WITH AREA GREATER THAN 150 SC KM\nAREA/AZIMUTH/RANGE\n60cell 1\n194\n255\ngrowth evident in both cells\n80cell 61\n946\n312\n151\n353\n132cell 7\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n257\n326\n85\n88","COMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/\nSTCDS/\nSTDTM\n1 1247 251.\n63.\n0.593\n33.63\n250.5\n36.4\n1 1249 252.\n61.\n33.45\n0.599\n246.8\n37.8\n1 1251 255. 60.\n40.84\n0.576\n240.6\n39.0\n6 1247 31l.\n81.\n0.00\n0.000\n346.1\n74.1\n6 1249 313. 80.\n69.93\n0.944\n268.5\n62.2\n6 1251 312. 80.\n1.099\n62.41\n271.2\n39.4\n132.\n7\n1249\n353.\n7 1251 353. 132.\n180.0\n0.0\n0.00\n0.000\n0.\ne\n0\nl.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\ne\n0\nD\nR\n7\n0\n0\nVIC\n1\n6\n1251 1351\n20e\nCOMMAND\nACC\nAREA/INTENSITY\n150\n4\nDAY / TIME / TILT\n1253\n0\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n59 cell 1\n256\n199\n79 cell 6 growth continuing\n971\n314\n2 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n77\n177\n302\n85\n267\n327\nDAY / TIME/ TILT\n306 1255\ne\nSTOP\n89","2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n}}\n57cell\n241\n259\nboth cells growing, cell 1 accelerating\n80 cell\n1023\n315\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n316 78\n471\nCOMMAND\nENT\nSTCTM\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/\nSTCDS/\n1 1253 256. 59.\n0.555\n39.81\n236.1\n39.9\n1 1255 259.\n57.\n41.75\n0.624\n231.4\n41.9\n6 1253 314.\n79.\n1.076\n260.7\n43.3\n58.90\n6 1255 315.\n80.\n1.015\n57.34\n248.2\n41.6\ne\n0\n0.\n0.\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\ne\ne\n0\n0\n0\nVIC\n1\n6\n0\n0\n1255\n1355\n202\nCOMMAND\nACC\nAREA/INTENSITY\n15l\n4\nDAY / TIME / TILT\n1257\n0\n306\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n54cell 1\n295\n26l\n80cell 6\n319\n1219\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n293\n326\n80\n90","DAY / TIME/ TILT\n0\n1259\n306\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n262 52cell 1\n26l\n79 cell 6\n320\n1222\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SC KM\nAREA/AZIMUTH/RANGE\n327\n78\n231\nDAY / TIME/ TILT\n0\n306 1380\nSTOP\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SC KN\nAREA/AZIMUTH/RANGE\n51 cell 1\n262\n278\n80 cell 6\n322\n1223\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SG KM\nAREA/AZIMUTH/RANGE\n327 77\n239\nCOMMAND\nENT\nSTCTM\nSTCDS/\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/\n1 1257 26l. 54.\n0.677\n40.65\n44.4\n228.9\n52.\n1 1259 262.\n0.685\n39.60\n46.9\n227.1\n1 1300 261. 51.\n0.680\n42.70\n226.7\n48.3\n6 1257 319.\n80.\n1.402\n57.82\n54.6\n237.8\n6 1259 32l. 79.\n1.313\n54.73\n59.5\n236.7\n6 1300 322. 80.\n1.359\n57.53\n65.0\n234.2\n0\nl.\n0.\ne\n91","COMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\ne\ne\n0\n0\n0\n0\n0\n6\n1\n13ll 14ll 202 VIC\nCOMMAND\nACC\nAREA/INTENSITY\n4\n150\nDAY / TIME/ TILT\n0\n1302\n306\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 S6 KM\nAREA/AZIMUTH/RANGE\n51 cell 1\n267\n278\n81 cell 6\n324\n993\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SC KN\nAREA/AZIMUTH/RANGE\n329 77\n207\nDAY/ TIME / TILT\ne\n1304\n306\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SC KM\nAREA/AZIMUTH/RANGE\n50 cell 1\n269\n266\n82 cell 6\n326\n1248\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SC KM\nAREA/AZIMUTH/RANGE\n78\n331\n202\nDAY / TIME / TILT\n0\n306 1306\nSTOP\n92","2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n49 cell 1\n278\n27l\n81 cell 6\n327\n1072\n1 SEV ECHOES FOUND WITH AREA GREATER THAN 150 SG KM\nAREA/AZIMUTH/RANGE\n79\n332\n205\nCOMMAND\nENT\nN/ HHMM/ AZM / RNG/ DIRECTION/ SPEED/ STCDS/ STCTM\n1 1302 267. 51.\n0.701\n50.80\n50.2\n224.3\n1 1304 269. 50.\n0.685\n49.67\n222.3\n51.8\n1 1306 27l. 49.\n0.670\n48.48\n52.8\n221.0\n6 1302 324. 81.\n1.292\n55.49\n231.5\n69.2\n6 1304 326. 82.\n1.235\n53.43\n229.2\n72.6\n6 1306 327. 81.\n1.190\n52.15\n73.3\n229.0\n0\nl.\n0.\n0\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\nl\n0\ne\n0\n0\n0\n6\n0\n1326 1406 20 € VIC\n1\nCOMMAND\nACC\nAREA/INTENSITY\n150\n4\nDAY/ TIME/ TILT\ne\n306\n1308\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SQ KM\nAREA/AZIMUTH/RANGE\n47 cell 1\n267\n290\n81 cell 6\n328\n1052\n93","DAY / TIME/ TILT\ne\n306 1310\nSTOP\n2 VST ECHOES FOUND WITH AREA GREATER THAN 150 SC KM\nAREA/AZIMUTH/RANGE\n46cell 1\n27l\n285\n82cell 6\n328\n1004\nCOMMAND\nENT\nSTCTM\nSTCDS/\nN/ HHMM/ AZM/ RNG/ DIRECTION/ SPEED/\n1 1308 267. 47.\n0.710\n51.53\n221.1\n52.9\n1 1310 270.\n46.\n0.699\n50.72\n220.9\n53.2\n6 1308 328.\n81.\n1.149\n50.14\n229.0\n72.6\n6 1310 328. 80.\n1.143\n48.96\n229.8\n69.8\n0.\n0\nl.\ne\nCOMMAND\nDIS\nBTIME/ETIME/RANGE/OVERLAY/ECHO NUMBERS\ne\n0\ne\n0\n0\n0\n0\n1\n6\n20e\nVIC\n1310\n1410\nCOMMAND\nBYE\n94","APPENDIX D\nComputer Program Listing for ECHOPRED\nCOMMCN XB(900), XE(20) YB(903), E(227),SL1222) PHIB(220) PHIE(20P)\nIGTLENG,\n*IGD, BCXTM,IHSK(19)\nDIMENSION IPCINT 12 '),LNSAV(42)\nDIMENSION NCB(12), AXI10),\nDIMENSION STE2(13),SCE2(12\nDIMENSION SWT(18),\nDIMENSION Sw(12),FTIME(1I)\nDIMENSION JCIRI521\nDIMENSION JEC(12), AREA(12),STORM(E)\nDIMENSION\nDIMENSION IKNT(8)\nDIMENSION ELPS13)\nREAL #8 SW,\nINTEGER\nINTEGER*2 IECAZ\nCHAR IECRNG\nCHAK CATE,IHSK\nDATA CP/w,SW'.'s ','SE', IF ', NE',''N 'NW' \"W\nDATA CAIR,QBYE,GENT,GCIS,QDEL.QTWC,QPOS,QWtE,QIGN\"\n#/ /\nDATA CKEY,QACCQGCC, GBQC/3FKEY 3HACC, 3HGCD, 3HBQC/\nDATA STORM/24H LGT MCC STR VST SEV EXTA\nFADIRIX,Y)=RTD*ATAN2IY,X1+189.\nHOUR(K)=FLOAT(K/19)+FLOAT(MODIK, 12911/68.\nJHHNN(T)=T60.+40IFIX(T)\n1127 FORMATI215,2F5.3\n1108 FORMAT(1X,12, $15,2F5.21\n1706\nFORMAT(1115\n6796\nFORMAT(1X,11151\nMXNCE=12\nMEN=180\nMXEP1=NXNCE+1\nRT0=57.2957795\nALN2=3.6931\nOTR=0.0174532925\nD4BP=4.213.1415927\nREAD STATICN NAMES ONT C DISK\nC\nREWIND 2\n1731 READ(5,1732)TN1,TN2 TN3, XA, YA\n1732 FURMAT(304,3X,2F8.2\nIFITN.E6.4HTAF )GC TO 1733\nCONVERT AM TC KM FOR AIRPORTS AND STORE ON DISK\nXA=XA*1.852\nYA=YA*1.852\nWRITE(2,1732)TN1,TA2,TN3,XA,YA\nGO TC 1731\n1733 END FILE 2\nENTER OVERLAYS CNTC CISK\nREWIND 3\nDO 1798 I=1,2\n1786\n1787 FORMAT(7F10.4)\nTO 1788\nL=1\nCALL SETLIN(L)\n*2)\nGO TC 1786\n1788 END FILE 3\n1790 CONTINUE\nDEFAULT TIME WEIGHT CONSTANT IS 30 MINUTES.\nC\nTWC=-ALN2*0.5\nDEFAULT GROUND CLUTTER DISTANCE FOR ECHO CONTOURING IS 20 KM\nC.\nIGC=29\nIF NC TIME IS AVAILABLE FROM BFNDIX DISPLAY BDXTM = a.\nC.\nBDXTN=0.\nC\nREAD COMMAND\nDO 1620 1=1,10\nJEC(I)=2\n95","DO 1020 J=1,4\n00 1x26 K=1,12\n1022 COF(I,J, K)=\nDO 1819 I=1,MEN\n1310\nMAP(I)=g\nI=1\nDO 1040 I=I,MXNCE\n1035\nIENC(I)=?\n1040\nNOB(I)=J\nWRITE(6,2101)\n2199\n2101\nC***360 DEPENDENT\nREAD(50,2105)CMD\nC*******\nWRITE16,2106)CMC\n2100 FORMATILX, 24A31\n2105 FORMAT(24A3)\nIFICMC E6. CENT )GL TO 1136\nEG. QGCD)GC TC 123.\nIFICMC\nIFICMC EG. CTWC)GC TO 1306\nIFICNC EC. CBYE)GC TO 1420\nIFICMC E6. CIGNIGO TO 150.\nIFICMC EC. QCEL)GC TC 1632\nIFICMC EC. CAIR)GC TO 173.\nIFICMC EC. CDIS) GO TO 1822\nIFICMC E6. (FOS)GO TO 198%\nIFICME EG. QACC)GC TO 2230\nIFICMC EG. CWHE)GC TO 3003\nIFICME EC. CBQC)GC TO 5003\nWRITE16,2112)CMC\n2110 FORMATITX,A3, INV/LIC.')\nGO TC 2193\nC\nTIME WEIGHT CONSTANT\nC\nENTER ECHC CBSERVATION\nC\n1131 DO FORMATIC 1102 N/ FHMNI AZM/ RNG/ DIRECTION/ SPEED/ STDDS/ STDTM )\n1100 WRITE(6, 1101)\nI=1, NXNCE\n1102 JEC(I)=3\nREAD ECHO CBSERVATION\nC\n1105 JE=\nC***360 DEPENDENT\nREAD(50, 1107)JE, JTM, EAZM, ERNG\nWRITE16, 1108)JE, JTM, EAZM, ERNG\nJ=1ABS(JE)\nIFIJ . EG. F)GO TC 2130\nPERFORM INPUT VALICITY CHECK.\nC\nIF(J .CT. MEN)GC TC 1160\nIF(MCC(JTM,198) .GE. 60)GO TO 1165\nIFIEAZN.GT.3601GC TO 1170\nETIM=HCUR(JTM) COMPUTE TIME IN FOURS, X AND Y COORDINATES.\nC\nAZM=EAZM*CTR\nXT=DCCSIAZM)*ERNG\nYT=DSINAZM)ERNG\nCHECK FCR NEW ECHO.\nC\nL=NAP(J)\nIFIL . GT . 01) GC TO 1132\nIFIJE .LT. C)GO TO 1135\nNEW ECHC. ALLCCATE.\nC\nDO 1115 L=1,MXNCE\nIF(IENC(L) . EQ. as GO TO 1125\n1115 CONTINUE\nALL WCRKING ECHCS USED.\nC\nWRITE16,1120)IENO\n*360 DEPENDENT\n1120 FORMAT(*UNABLE TC ACCOMODATE NEW ECHO.'\nCELETE CNE OF THE FOLLOWING*/1214\nGO TC 2190\nSTART NEW ECHO WITH INDEX L\nC\n1125 NOB(L)=1\nIENCIL)=J\nMAP(J)=L\nE2(L)=0.\nSTE2(L)=0.\nSWIL)=1.\nSWT(L)=ETIM\n96","SWT2IL)=ETIN*2\nSWXIL)=XT\nSWTXIL)=XT*ETIM\nSWYIL)=YT\nSWTY(L)=ETIMYT\nFTIME(L)ETIN\n1129 GO TC 1125\n1130 IFIFTINEIL).GT.ETIMTIMETIM+24.\nC\nCOMPUTE WEIGHT FACTOR\nW=EXPITWC#(ETIM-FTIME(L)))\nC\nTEST FCR CBSERVATION DELETION.\nIF (JE.GT.0) GC TO 1140\nC\nDELETE PREVIOUS DESSERVATION\nIF(ETIM .LE. FTIME(L))GO TU 1135\n1133 WRITE(6,1134)\n1134 FORMATI YOU CANT CELETE SOMETHING THATS NOT THERE')\nGO TC 210\n1135 NUB(L)= NCB(L)-1\nIF(NCe(L) LE. 2)GC TO 1137\nSW(L)=SWIL)-W\nSWT(L)=SWT(L)-W*ETI\nSWT2(L)=SWT2IL)-W*ETIM**2\nSWXILI=SWXIL)-W#XT\nSWYIL)=SWYIL)-W#YT\nSWTX(L)=SWTXIL)-W#TIM*XT\nTTY(L)=SWTYIL)-WTIMYT\nGO TC 1135\n1137 IENC(L)=0\nMAPIJI=D\nGO TC 1125\nC\nADD NEW CASERVATION\n1140 DT=ABS(ETIM-FTIME(LI)\nIF (CT EG. ?.)GC TO 1175\nIF(NCBIL) .LT. 2)GC TO 1145\nC\nCOMPUTE ERRCRS\nPTIME=(AXIL)#(XT-BX(L))+\n(=AXIL)*PTIMEeX(L)\nPY=AYIL)*PTIME+PY(L)\nPRDST=SQRTIIPX-XT)#*2+(PY-YT)**2\nERTIN