Elastic Characterization at FORGE ~P-wave Tomography and VSP Subsurface Imaging~


Nori NAKATA, Don W. VASCO, Hongrui QIU, Peidong SHI, Federica LANZA, Ben DYER, Tong BAI, and Coral CHEN

Key Words:

subsurface imaging, FORGE, tomography, VSP


Stanford Geothermal Workshop







Paper Number:


File Size:

1514 KB

View File:


For understanding subsurface geology, faults locations and fracture distribution, the detailed subsurface structure information such as velocity models and reflectivity images is essential. The velocity models can be used for finding accurate earthquake locations and characterization. We use 3D seismic survey in 2018 and a walkaway VSP survey in 2022 to refine the elastic model at FORGE, Utah. We apply travel-time tomography for the 3D seismic survey, which contains nearly 1100 vibroseis shots and 1700 receivers. To increase the picking accuracy, we develop and test two approaches: a machine-learning-based (ML) picker and a frequency-dependent traveltime picker. For the ML picker, we adapt the EQTransformer with crosscorrelation to pick P-wave travel times. Although the EQTransformer is designed for earthquake waveforms, it can still pick traveltimes of active-seismic data compared to other ML pickers. The frequency-dependent picker is a new approach for the waveforms before correlating vibroseis sweep. In the waveforms, the sweep signals are recorded as linear upsweep signals. We use a time-frequency analysis to find the linear trend, which corresponds to the arrival time of the wave. Both methods match reasonably well with some differences, which will be discussed. Then we apply a tomograophic inversion with the Eikonal solver to estimate the subsurface velocities with higher resolution than the model developed for migration. The VSP survey was recorded after the 2022 April stimulation. The survey contains 106 vibroseis shot points, two DAS systems in nearby boreholes (78-32A and 78-32B), and two geophone systems (58-32 and 78-32B). Direct P and S waves are clearly observed with some reflections. Geophones generally have higher signal-to-noise ratio than DAS. We apply reflection imaging using reverse-time migration and velocity modeling with travel time tomography. The velocity model in top 100 m and deeper than 500 m will be updated with the VSP data.

ec2-3-215-186-30.compute-1.amazonaws.com, you have accessed 0 records today.

Press the Back button in your browser, or search again.

Copyright 2023, Stanford Geothermal Program: Readers who download papers from this site should honor the copyright of the original authors and may not copy or distribute the work further without the permission of the original publisher.

Attend the nwxt Stanford Geothermal Workshop, click here for details.

Accessed by: ec2-3-215-186-30.compute-1.amazonaws.com (
Accessed: Friday 24th of May 2024 06:24:20 AM