Welcome to the NOAA Institutional Repository | A tsunami forecast model for Elfin Cove, Alaska - :11191 | Office of Oceanic and Atmospheric Research (OAR)
Stacks Logo
Advanced Search
Select up to three search categories and corresponding keywords using the fields to the right. Refer to the Help section for more detailed instructions.
 
 
Help
Clear All Simple Search
Advanced Search
A tsunami forecast model for Elfin Cove, Alaska
  • Published Date:
    2015
Filetype[PDF - 31.23 MB]


This document cannot be previewed automatically as it exceeds 5 MB
Please click the thumbnail image to view the document.
A  tsunami forecast model for Elfin Cove, Alaska
Details:
  • DOI:
    doi:10.7289/V5VH5KTQ
  • Personal Authors:
  • Corporate Authors:
    NOAA Center for Tsunami Research
  • Series:
    NOAA OAR special report
    PMEL tsunami forecast series ; vol. 13
    Contribution (Pacific Marine Environmental Laboratory (U.S.)) ; no. 3405
    Contribution (University of Washington. Joint Institute for the Study of the Atmosphere and Ocean) ; no. 2089
  • Document Type:
  • Description:
    Operational tsunami forecasting by NOAA's Tsunami Warning Centers relies on the detection of tsunami wave trains in the open ocean, inversion of these data (transmitted via satellite) to quantify their source characteristics, and real-time modeling of the impact on threatened coastal communities. The latter phase of the process involves, for each such community, a pre-tested forecast model capable of predicting the impact, in terms of inundation and dangerous inshore currents, with sufficient resolution and within the time constraints appropriate to an emergency response. To achieve this goal, considerable advance effort is required to tune each forecast model to the specific bathymetry and topography, both natural and manmade, of the impact area, and to validate its performance with a broad set of tsunami sources. Where possible, the validation runs should replicate observed responses to historical events, but the sparse instrumental record of these rare but occasionally devastating occurrences dictates that comprehensive testing also include a suite of synthetic scenarios that represent potential extreme events. During the forecast model design phase, and in research mode outside the pressures of an emergency situation, more detailed and slower-running models can be investigated. These models, referred to as reference models, represent the most credible numerical representation of tsunami response for a study region, using the most detailed bathymetry available and without the run-time constraint of operational use. Once a reference model has been developed, the process of forecast model design is to determine where efficiencies can be gained, by reducing the grid resolution and increasing the model time step, while still adequately representing the salient features of the full solution. This report addresses the tsunami aspects of the natural hazard spectrum, documenting the reference and forecast model development for Elfin Cove, Alaska, and its vicinity. Though sparsely populated, the region is traversed by several segments of the Alaska Marine Highway and features important marine resources, including commercial and recreational fishing and the potential for tidal power extraction. Additionally, Glacier Bay is a popular venue for cruise ships and other tourist activity. The forecast model performs satisfactorily in hindcasts of major historical tsunamis and its stability in tests of large synthetic events around the Pacific basin has been demonstrated. However, it should be noted that forecast model amplitudes consistently underestimate those produced in reference model runs. The disparity is generally only a few percent in the early phases of the tsunami wave train but may be significantly larger for later waves. It is suggested that a safety factor of the order of 10% be applied in operational use of forecast model projections. During testing, the forecast model simulated 4 hr of real time in 12.92 min. While this exceeds the 10 min target for this metric, the modest increase is justified by the regional coverage provided by the intentionally enlarged model domain. [doi:10.7289/V5VH5KTQ (http://dx.doi.org/10.7289/V5VH5KTQ)]

  • Supporting Files:
    No Additional Files