P-Wave Velocity Model at Utah FORGE Geothermal Field Using Travel-Time Tomography of DAS-VSP Data

At geothermal sites, estimating underground velocity structures from seismic data is essential for improving the accuracy of microseismic event localization, characterizing reservoir properties, and understanding subsurface dynamics. However, constructing detailed velocity models from surface seismic data is challenging due to the poor signal-to-noise ratio (SNR) and the complexity of near-surface structures with significant topographic variations. This study presents a novel application of P-wave travel-time tomography using distributed acoustic sensing (DAS) to develop a 2-D P-wave velocity model for the Utah Frontier Observatory for Research in Geothermal Energy (FORGE) geothermal site. Unlike conventional vertical seismic profile (VSP), the DAS technique enables dense spatial sampling along boreholes, providing higher resolution subsurface characterization. First, we extract the 2-D initial model from the legacy 3-D velocity model estimated from a 3-D active-source seismic survey. Then, we apply P-wave travel-time tomography with DAS-VSP data acquired in two boreholes close to each other. Consequently, the resulting velocity model shows that most first arrivals are close to real VSP data, even for far-offset data, and align well with geological and well-log data. The detailed model reveals lateral variations across three distinct depth zones: unconsolidated alluvium, consolidated alluvium, and granitoid. Thus, this study demonstrates the capability of DAS-VSP data and travel-time tomography in resolving localized subsurface structures and highlights its potential to address cycle-skipping issues in future high-frequency full-waveform inversion (FWI) studies requiring accurate initial velocity models.