T11D-2920
Crustal Velocity Structure of the Rio Grande Rift and Rocky Mountains from Local Earthquakes and Blasts Recorded by USArray and CREST

Monday, 14 December 2015
Poster Hall (Moscone South)
Jenny Nakai, University of Colorado, Boulder, CO, United States, Anne F Sheehan, University of Colorado at Boulder, Boulder, CO, United States and Susan L Bilek, New Mexico Tech, Socorro, NM, United States
Abstract:
Arrival times from over 3,100 earthquakes and 2,800 mine blasts recorded at USArray Transportable Array (TA) and other regional broadband seismic stations are inverted to find the regional P and S velocity structure in Colorado and New Mexico. Knowledge of the crustal structure in Colorado will help inform to what extent this structure and composition influences the isostatic compensation of high topography in the region. The relationship between the Rio Grande rift and the surrounding physiographic provinces remains enigmatic, and neither the geology nor geophysical surveys have clearly resolved the rift in Colorado. Therefore, tomography may supply more information about velocity variations along the rift within the context of the Colorado Plateau, the Great Plains, and the Rocky Mountains. Thus far, applications with the TA data to resolve P wave crustal structure are rare due to distant station spacing and small magnitude and shallow (mid to upper crustal) local earthquakes. The depths of the earthquakes range from 4 km to the mid-crust, so we expect dense ray coverage in the upper crust. In order to increase the number of crossing rays, we use mine blast P wave arrivals to constrain shallow surface structure. A total of 70,000 P wave arrivals and 18,000 S wave arrivals constitute the dataset. We develop a reliable 1D regional velocity model using 500 of the largest earthquakes with a fixed Vp/Vs ratio from the arrival data, then use this model as an input to investigate the feasibility of utilization of a 3D inversion algorithm. While use of a 3D inversion algorithm will be explored, construction of a series of 1D velocity and Vp/Vs models may prove to be more robust.