No-source tsunami forecasting for Alaska communities

Abstract:

The presented tsunami forecasting technique employs observations of the approaching tsunami at DART stations near the Aleutian trench to provide fast local forecasts for the Alaska communities. The suggested technique

• yields a prediction independent of the tsunami source estimate;

• increases forecast accuracy by using observations close to the target area;

• allows for checking the accuracy of the inversion-based forecast before the wave hits the coast.

We demonstrate this forecasting technology, introduced in (Power and Tolkova, 2013, Ocean Dynamics, 63(11), 1213-1232), with imitating real-time forecasts of the 2011 Tohoku tsunami at several coastal sites in Alaska (to be compared with the gage records). The coastal forecasts are generated as the wave is registered at regional DART stations (46402, 46043, 46409, 46410). Note that while the DART array spans the Pacific Rim, the inversion-based forecasting methodologies can incorporate data from only 1-3 stations in the vicinity of the tsunami origin. We present a forecasting method which complements existing forecasting tools by using tsunami observations in a region to generate regional predictions independent of the tsunami source estimate. This method allows to utilize observing capabilities of the DART array, as well as tsunami detectors in cabled underwater networks (e.g. NEPTUNE in Canada). Future instrumentation on submarine communication cables will supply larger selection of open-ocean measurements and many more opportunities for this method.

Figure: (Top) record of the 2012/10/28 Haida Gwaii tsunami at DART 46411; (Bottom) the tsunami record at Monterey tide gage (red) and its forecast (blue). The forecast is been made as the wave is been registered at the DART one hour before arriving at the gage (Power and Tolkova, 2013).