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 A, Poster #199, SCEC Community Models (CXM)

3D full waveform inversion of the Salton Seismic Imaging Project active source dataset

Rasheed Ajala, Patricia Persaud, & Folarin Kolawole
Poster Image: 

Poster Presentation

2023 SCEC Annual Meeting, Poster #199, SCEC Contribution #12978 VIEW PDF
High-frequency (>1 Hz) structural imaging is a long-term challenge the community must address to advance earthquake science and better understand crustal deformation processes. Developing automated tomography tools and deploying dense seismic arrays help overcome significant obstacles, but building large-scale detailed models remains computationally intractable. Model hybridization offers a paradigm shift to the problem by allowing the development and incorporation of small-scale high-frequency models in perceived regions of interest — sedimentary basins, fault zones, expensive infrastructural areas — into background longer-wavelength models. Here, we work toward developing a high-frequen...cy full waveform tomographic model of Imperial Valley using the dense Salton Seismic Imaging Project (SSIP) dataset and embedding it into existing SCEC community velocity models (CVMs). Previous results with the SCEC CVMs and detailed basin-scale SSIP travel time tomographic models verify the feasibility of the approach in the global enhancement of low-frequency wavefield predictions of the SCEC models. For the merging parameters used in the validation exercise, the hybrid version outperformed CVM-H 15.1 in almost all cases as opposed to CVM-S 4.26, indicating that CVM-S 4.26 may be more susceptible to inversion update conflicts from the SSIP dataset. The evaluation of representations of CVM-H 15.1 also showed better performance without surface elevations. These results inform our choice of the longer-wavelength model as a topography-squashed hybrid CVM-H 15.1 model. We present ongoing developments in the inversion process.