Exciting news! We're transitioning to the Statewide California Earthquake Center. Our new website is under construction, but we'll continue using this website for SCEC business in the meantime. We're also archiving the Southern Center site to preserve its rich history. A new and improved platform is coming soon!

Group B, Poster #074, Tectonic Geodesy

Automated block closure and resolution tests in dense block models: preliminary results

Jayson P. Sellars, & Eileen L. Evans
Poster Image: 

Poster Presentation

2022 SCEC Annual Meeting, Poster #074, SCEC Contribution #12264 VIEW PDF
Block models, in which the seismogenic crust is divided into microplates bounded by faults, are a powerful tool for estimating fault slip rates from geodetic observations such as from the Global Navigation Satellite System (GNSS) and/or geologic constraints. However, block models require that all faults in the model connect and form closed blocks. Typically, the process of forming closed blocks is done by inspection based on mapped fault traces. Manually closing blocks is a tedious task which leads to creating only a handful of block models in a low-density system of known faults in the area of interest. Furthermore, given ambiguity in block model closure, assessing how well block models tru...ly preform remains a challenge. In this work, we compare block models with varying geometries in both toy models and larger-scale models. Larger-scale models are based on Southern California and generated with a block closure algorithm. The block closure algorithm automatically closes mapped fault segments into closed blocks using a modified cone search with adjustable parameters and allows us to produce suites of block geometries to explore the influence of block closure choices on estimated slip rates. Here, we perform resolution tests by employing the block closure algorithm to the dense Fault Sections database of the 2023 update to the USGS National Seismic Hazard Model (NSHM2023) in southern California. We estimate initial fault slip rates regularized with total variation regularization (TVR) within a representative block geometry; TVR allows robust slip rate estimation within a dense block model with many poorly constrained blocks. We use these estimated rates to generate a forward model. We then perform TVR inverse models on our algorithmically generated suite of block geometries, constrained by forward data, to assess block model performance. Toy model comparisons, as well as preliminary Southern California-based models, show that TVR slip rate estimates fit best on models that are similar in geometries to the representative model.