National Oceanic and
Atmospheric Administration
United States Department of Commerce


FY 2009

Extracting tsunami source parameters via inversion of DART® buoy data

Percival, D.B., D. Arcas, D.W. Denbo, M.C. Eble, E. Gica, H.O. Mofjeld, M.C. Spillane, L. Tang, and V.V. Titov

NOAA Tech. Memo. OAR PMEL-144, 22 pp (2009)

The ability to accurately forecast the hazards to coastal communities from earthquake-generated tsunamis requires the knowledge of certain tsunami source parameters. The Short-term Inundation Forecast for Tsunamis (SIFT) tool developed by the National Oceanic and Atmospheric Administration estimates these source parameters based upon a combination of observed data from DART® buoys and a set of precomputed models. We give a detailed description of the procedures within SIFT Version 2.0 for estimating the source parameters and for assessing the sampling variability in the estimated parameters. We illustrate these procedures using detided 1-min data collected by two DART® buoys (21414 and 46413) in the northwestern Pacific during the 15 November 2006 Kuril Islands tsunami event, generated by an offshore moment magnitude 8.3 earthquake. We discuss the advantages of using a least squares (L2) method for fitting model tsunami time series to the data. The model series originate from two model source segments (50 km by 100 km) near the earthquake epicenter. The fitting procedure estimates the source parameters of each segment, both individually and together. Realistic confidence limits on the source parameters are obtained by fitting the residual series (from the data-fitted model) to a first-order autoregressive model to account for the obvious correlation between adjacent residual values. Using data from both buoys improves the source parameter estimates, at an operational cost of some delay in time for the tsunami to reach the farther station. An appropriate fit uses the first full tsunami wave at each station with no improvement found by extending the fitted data segment further. Formulae are derived for estimating the source parameters and computing the confidence limits for multiple DART® buoys and source segments.

Feature Publications | Outstanding Scientific Publications

Contact Sandra Bigley |