Date of Award


Degree Name

Doctor of Philosophy


Geological Sciences


Marianne S. Karplus

Second Advisor

Diane I. Doser


With increasing earthquakes in the Delaware Basin since 2009, earthquake studies, including accurate hypocenters, are critically needed in the Delaware Basin to identify the structures producing earthquakes, and to determine if they are related to unconventional petroleum development and production. In 2018, with funding from the Texas Seismological Network (TexNet), the University of Texas at El Paso deployed and maintained a nodal network of 25 Magseis Fairfield Z-Land Generation 2 5-Hz seismic nodes in the Pecos, Texas region of the Delaware Basin, known as The Pecos Array. The network was deployed from November 2018 until the beginning of January 2020, with an additional two months of data recorded in September and October 2020. The network collected continuous 3-component data with a 1000-Hz sampling rate. The spacing of the nodes varied from ~2 km in town to ~10 km farther away from the city center. The primary goal of this network was to improve estimation of event hypocenters, which will help to determine why there has been an increase in earthquakes over the past several years. We summarize the scientific motivation, deployment details, and data quality of this network. Data quality statistics show that we successfully collected continuous data with signal-to-noise ratios that allow us to detect and locate events, hundreds of them being estimated at ML < 0.50. This unique dataset is contributing to new seismotectonic studies in the Delaware Basin.Using the Pecos Array, we analyzed five months (January â?? May 2019) of data using the machine learning algorithm, called PhaseNet, to assess how well it picked and located events compared to manual picks and earthquake locations from TexNet. We found that P-phase picks differed by an average of 0.0780 seconds, and S-phase picks differed by an average of 0.1829 seconds compared to the analysts. We used these phases to associate and relocate earthquake events using HypoInverse and compared the resulting two catalogs. We found that events from both catalogs align with shallow normal faults and maximum horizontal compressive stresses observed in the area. Ninety-eight percent of all events occur above 5 km in depth, with an average error of < 1.5 km. Recent research suggests wastewater disposal and fracking to be the cause of the earthquakes. While we did not specially study the source of the earthquakes, our events fall at the depths where these procedures are occurring. We conclude that PhaseNet is a good alternative compared to standard automatic pickers, and that our network was able to successfully constrain hypocenter locations in the Pecos, TX region. The Pecos Array was used to conduct a local 3-D tomography study in the Pecos, Texas region using the SIMULPS algorithm. Previous tomography studies have been conducted to update the velocity model of the entire Permian basin, but these studies covered a broader area, and their best velocity results were in the upper 35 km of the crust. With our denser array, localized earthquakes, and dense ray paths in the Pecos, Texas area, we have created a detailed 3-D velocity model for the top 8 km of this region. We observe velocity values that are consistent with reported geologies in the basin, and slower velocities on faults compared to areas without faults. We found that our new model is 28% slower compared to the DB1D velocity model that TexNet uses.




Recieved from ProQuest

File Size

106 p.

File Format


Rights Holder

Jenna Lynn Faith