INVESTIGATING GEODETIC AND GEOLOGIC VERTICAL VELOCITY DISCREPANCIES ALONG THE SOUTHERN SAN ANDREAS FAULT SYSTEM USING MULTIPLE COMPARISON TECHNIQUES AND DEVELOPING A GROUNDWATER CORRECTION

Geodetic and geologic measurements of vertical deformation in southern California record localized zones of uplift and subsidence along the San Andreas Fault System (SAFS) derived from tectonic motions, and anthropogenic sources such as groundwater pumping and hydrocarbon extraction. Investigating the relationship between geodetic and geologic vertical data is nontrivial, as these datasets differ in geographic coverage area, spatial resolution, signal source, and associated uncertainties. Here we compare 888 geodetic velocities (from Earthscope PBO) that range from -11.1 to 7.6 mm/yr (average uncertainty of 1.7 mm/yr), and 1627 geologic velocities (from the SCEC Vertical Motion Database) that range from -3.5 to 3.6 mm/yr (average uncertainty of 0.13 mm/yr). These datasets are not spatially co-located, so several different interpolation techniques (surfacing, masking, Delaunay triangulation) were utilized for optimal analysis of the data. While the number of data points for each comparison largely depends on the chosen interpolation technique, all methods suggest that the relationship between the geodetic and geologic vertical data is not 1:1. In regions of subsidence, for example, the geodetic rates are often twice as fast as the geologic rates, suggesting that the respective motion sources of the data play an important role. A vertical velocity crustal deformation model was used as a purely tectonic reference for the two data sets, and initially provided a poor fit to the GPS data. In order to improve this fit, a groundwater correction was developed to extract groundwater pumping effects recorded in the GPS data. A ground subsidence/groundwater level change ratio of .006 was derived from published measurements, and applied to changes in groundwater levels from regional well log data. Applying the groundwater correction improved the GPS-model fit by 33% and also doubled the correlation between the geodetic and geologic data. As we continue to explore the vertical motion discrepancies of the SAFS using regional tide gauge records along the California coast, we emphasize the fundamental importance of correcting vertical deformation data for groundwater pumping and hydrocarbon extraction, and encourage future deployment of geodetic arrays in locations complimentary to existing geologic observations.