An Open-Source Web Platform for Real-Time Seismic Impact Assessment Integrating USGS Live Feeds, Physics-Based Intensity Models, and Multi-Source Remote Sensing Geospatial Datasets

Rapid seismic impact assessment remains a critical challenge for disaster risk reduction, particularly in regions with limited ground instrumentation. This study presents an open-source, browser-based platform that integrates real-time earthquake data from the United States Geological Survey (USGS) Federated Digital Seismic Network with geospatial and remote sensing datasets to deliver physics-based, on-demand earthquake impact assessments worldwide.

Shaking intensity is computed using multiple published Intensity Prediction Equations (IPEs) expressed on the Modified Mercalli Intensity (MMI) scale. Seismic wave travel times are derived from the ak135 Earth model, while site amplification is represented using NEHRP Vs30 classifications. The platform also integrates WorldPop population data, global land-cover information, and OpenStreetMap infrastructure data to estimate population exposure and infrastructure distribution. Additional analytical tools include seismicity heatmaps, tectonic plate visualization, Gutenberg–Richter analysis, aftershock characterization, and interactive spatial dashboards.

Validation against USGS ShakeMap reference zone radii was conducted for ten historical earthquakes ranging from Mw 6.0 to Mw 9.1 across four tectonic settings. The Atkinson and Wald (2007) model reproduced MMI ≥ 6.0 moderate-shaking zone radii with a mean absolute error of 13.3% and a bias of −1.3%, with six of ten events within ±15% error. Performance for MMI ≥ 8.0 severe-shaking zones was lower, with a mean absolute error of 40.1% and a bias of −14.6%, primarily due to the point-source approximation for large finite-fault ruptures. These results demonstrate the potential of the platform for rapid, reproducible, first-order seismic impact assessment while highlighting the need for finite-fault and spatially distributed site-response models for improved near-source predictions.