Spatial and Temporal Variations in Seasonal and Anthropogenic Ground Movements Recorded by Continuous GPS in Southern California

Abstract:

Continuous Global Positioning System (GPS) stations in many locations around the world record mm-scale quasiperiodic motions that may arise due to seasonal variations in precipitation, and/or anthropogenic groundwater removal. The causes and spatial patterns of these motions are difficult to determine because diffuse GPS station spacing. In southern California, a dense GPS network, large seasonal motions, and seasonally variable precipitation make the region ideal for characterizing both seasonal and anthropogenic motions. In this study, we utilize data from 57 permanent GPS sites in the Plate Boundary Observatory network. In order to quantify periodic seasonal motions, we perform a Discrete Fourier Transform (DFT) on detrended GPS time series to calculate the phases and amplitudes of the annual harmonics. To compare ground motions to precipitation patterns, we also perform a DFT on monthly NOAA precipitation data, which allows us to calculate phase differences between the rainfall and GPS recorded motions. Phase calculations show that sites overlying the Santa Ana aquifer and sites not overlying the aquifer have phase differences of 45 and 180 days suggesting that many sites respond quickly to seasonal rainfall patterns. We find that ground movement above the Santa Ana aquifer has significantly different characteristics than the surrounding ground movement suggesting an anthropogenic source. Furthermore, existing InSAR data shows significant localized motions, which have been proposed to be anthropogenic. This suggests that anomalous annual phases and/or amplitudes in GPS time series may help to identify stations that are recording significant anthropogenic motions without the need for other data. To remove these hydrologic motions from the secular velocity estimates, we use inverse dislocation models that simulate aquifer compaction. These models are then used to effectively “correct” the GPS velocities to yield more accurate tectonic estimates of motion.