Merged Real Time GNSS Positioning and Seismic Moment Estimation in Support of NOAA Operational Tsunami Warning

Date: 4/25/2019

Time: 02:15 PM

Room: Puget Sound

Real-time measurements from increasingly dense Global Navigational Satellite Systems (GNSS) networks located throughout the circum-Pacific offer a growing observational platform for rapidly characterizing tsunami excitation. Currently over 17,000 GNSS instruments operate globally, many of which can provide on-the-fly characterization of transient ground displacements directly relatable to seismic slip and moment. This talk will summarize the status of a two-year effort to deliver low-latency GNSS solutions to NOAA’s two operational US tsunami warning centers in Palmer, AK and Honolulu, HI, along with GNSS-based seismic moment estimation routines compatible within their operational framework.

The usefulness to NOAA of real-time GNSS depend on their latency, resolution and completeness. Two additional operational requirements are redundant sources of GNSS solutions and that solutions be statistically identical irrespective of source. As a result, three independent real-time solutions produced by each of Scripps Institution of Oceanography, JPL/Caltech, and Central Washington University (PANGA) are shared internally prior to being formally merged within a simple 3-state Kalman filter for distribution to NOAA’s two tsunami centers. The merging filter runs simultaneously at each institution, allowing NOAA to access the merged solutions from a given station from any of the three providers. Solution sharing is accomplished using RMQ publication/subscription service. Merged solutions have been shown to offer lower scatter, higher resolution and completeness than what individual analysis provides, at a cost of ~10 seconds latency. We have implemented merging solutions from 110 stations, and see to expand these globally.

Both NOAA operational centers operate within an Earthworm environment. Static offset picking algorithms based on STA/LTA have been ported to Earthworm, as has the PGD moment estimation algorithm of Crowell (2011).

Presenting Author: Timothy I. Melbourne

Authors