Accuracy of Real-Time Tsunami Forecast Based on Real-time Data

Abstract:

The goal of tsunami warning, as a system for prediction of potential tsunami impact at coastlines, requires rapid detection, assessment and model assimilation of available tsunami real-time data.

Historically, the earthquake magnitude has been used as a proxy of tsunami impact estimates. Real-time seismic data has been available for real-time processing for decades and ample amount of seismic data is available for an elaborate post event analysis. This measure of tsunami impact carries significant uncertainties for quantitative tsunami impact estimates, since the ratio between the energy of an earthquake and resultant tsunami (averaging at around 0.1%) varies significantly from one event to another. The other historical stream of data for tsunami warnings have been tide-gage real-time data that are difficult to interpret and have limited forecast value, since tide gages record tsunami impacting the coastlines already.

New real-time data streams that have become available to tsunami warning system include real-time deep-ocean tsunami detection (DART system) and GNSS real-time data streams. The data provide new opportunity for real-time tsunami forecast systems. Assimilation of these new data into tsunami forecast capability has potential of increasing the accuracy of the tsunami forecast, thus reducing false alarm rates, providing quick practical tsunami forecast products and, ultimately saving more lives during the tsunami warning event.

Implication of including new real-time data on tsunami forecast accuracy will be discussed.