G015-01
Drought Induced Groundwater Loss Around Great Salt Lake, Utah, Inferred from 3D GPS Displacements
Drought Induced Groundwater Loss Around Great Salt Lake, Utah, Inferred from 3D GPS Displacements
Monday, 14 December 2020: 16:00
Virtual
Abstract:
The American Southwest experienced a massive drought between 2012 and 2016 with significant impacts to regional water storage. During this period, the Great Salt Lake (GSL), Utah, lost 1.84 meters of water and Global Positioning System (GPS) data show significant changes in nearby station positions during the same time. Although the GPS data show expected uplift and extension localized on the GSL, preliminary analysis suggests that the observed GSL unloading alone cannot fit the GPS displacements and contributions from groundwater loss surrounding the GSL are likely. This study applies a damped least squares inversion to determine the quantity and distribution of groundwater removal consistent with the observed deformation. We test a large number of load distributions over a range of radial load rings and compare both the predicted vertical and horizontal displacements to the observations. We estimate loading coefficients for individual load belts using LoadDef, a stratified, gravitating, spherical load model produced by Martens et. al., 2019. Three dimensional inversion provides the most realistic distributions, compared to horizontal and vertical only solutions, and yield GSL unloading comparable with the observed water loss (i.e., a volume of 5.50 km3). The best model infers a radially decreasing mass loss up to 32 km from the edge of the lake at a total volume of 15.56 km3, over four times the volume observed by GRACE for the entire region. The maximum localized unloading is on the lake itself; however, the contribution of exterior groundwater loss is substantial and greatly improves the fit to the data. We observe a strong correlation between seismicity trends and drought severity; i.e., seismicity increases during drought periods, presumably due to unloading of the Wasatch fault system. This correlation is only seen for the area where the footprint of the inferred load is adjacent to the fault, not elsewhere.

