Investigating Different Methodologies for a Sar Coherence Change Detection Product
Description:
The U.S. Geological Survey (USGS) provides rapid (within 30 minutes) estimates of earthquake-induced building damage and ground failure following significant events. These products are based solely on pre-event data and event-specific shaking estimates and do not include direct observations of damage or ground failure following an earthquake. To this end, the USGS is developing an intermediate-timeframe pipeline for post-earthquake products (within days to a week) that combines the current rapid estimation products with post-event observations with the goal of more accurately identifying the most affected areas. As a vital component of this pipeline, the USGS plans to generate an in-house Synthetic Aperture Radar (SAR) coherence-based Change Detection Map (CDM) using pre- and post-earthquake satellite observations to identify regions exhibiting surface property changes attributable to damage and ground failure. While NASA produces a similar product called the Damage Proxy Map (DPM), the USGS CDM is specifically being developed for integration into the USGS earthquake product pipeline. Here, we examine various methodologies to generate CDMs, with a focus on optimizing CDM robustness. We explore sample area dimension, the functional form of sampled coherence (e.g., boxcar versus Gaussian), and the combinations of SAR pairs (e.g., sequential-only versus multi-pairs). We generate CDMs using both Sentinel-1 (~30 m resolution) and Wide Ultrafine Radarsat-2 data (~10 m resolution), the latter of which has not yet been used to generate post-earthquake damage estimates. We compare our results against damage and landslide observations derived from high-resolution optical data (PlanetLab and WorldView, ~1-5 m resolution) from the 2023 Mw 6.8 Morocco earthquake. By exploring different CDM generation methodologies, we can constrain a CDM product’s accuracy and spatial limitations in identifying building-specific damage and resolving decimeter-scale ground failure.
Session: Creating Actionable Earthquake Information Products [Poster Session]
Type: Poster
Date: 5/1/2024
Presentation Time: 08:00 AM (local time)
Presenting Author: Paula
Student Presenter: No
Invited Presentation:
Authors
Paula Burgi Presenting Author Corresponding Author pburgi@usgs.gov U.S. Geological Survey |
|
|
|
|
|
|
|
|
Investigating Different Methodologies for a Sar Coherence Change Detection Product
Category
Creating Actionable Earthquake Information Products