Separating Tectonic, Magmatic, Hydrological, and Landslide Signals in GPS Measurements near Lake Tahoe, Nevada-California

Blewitt, Geoffrey; Kreemer, Corné; Hammond, William C.; Plag, Hans-Peter
Nevada Geodetic Laboratory, Nevada Bureau of Mines and Geology, University of Nevada, Reno; gblewitt@unr.edu

We present two case studies of GPS measurements recorded in the region of Lake Tahoe, Nevada-California, where signals evidently represent some mix of tectonic, magmatic, and hydrological processes, and discuss clues that have been used to help unravel their relative contributions to the total signal. In both cases, the pattern of seismicity and its timing was critical, as well as consideration of spatial and temporal patterns of deformation that likely arise from each process. The first case study involves the Slide Mountain GPS Transient of 2003, previously reported by Smith et al., [2004, doi:10.1126/science.1101304], who inferred evidence for magma injection beneath the north shore of Lake Tahoe associated with deep (> 20 km) earthquake activity. During a 5 month period, a GPS station atop Slide Mountain, which lies 18 km northeast of the earthquake swarm, moved 6 mm horizontally and 8 mm upward. Separating the magmatic transient from tectonics and hydrology required extrapolation of the time series residuals derived from modeling of pre-transient data. The pre-transient data were modeled as secular tectonics (fit with station velocity parameters) plus seasonal hydrological loading (fit with annual and semi-annual sinusoids). We re-examine this event now that more than 6 years have passed since the end of the transient, using more recently advanced GPS processing techniques. Questions of interest include: (1) To what extent is the 5-month transient affected by seasonal hydrological signals that do not repeat exactly from year to year? (2) Does the post-transient station velocity differ from the pre-transient velocity, and if so, what does that imply for tectonic-magmatic processes at Lake Tahoe? We attempt to find clues to the latter question by interpolation of the regional strain field inferred by GPS farther from the swarm. We find that the magmatic transient occurred at the edge of a larger area, centered on the Steamboat Springs geothermal area, for which we find relatively high secular uplift and east-west extension. We postulate a regional magmatic doming process to be responsible for both the anomalous secular and transient deformation, and the presence of high temperatures in the area. In the second case study, we present results of transient GPS signals associated with the Mogul-Sommerset earthquake swarm of 2008 just west of Reno, Nevada. This case study illustrates how problematic hydrological signals were identified at one GPS station and were rejected based on site stability considerations. The tectonic (versus hydrological or magmatic) origin of the observed signal from the remaining stations was confirmed by (1) models and observations of the spatio-temporal pattern of surface displacements within a few kilometers of the swarm, (2) the timing of surface displacement in relation to that of seismic moment, and (3) the strain release pattern in relation to the regional inter-seismic strain field inferred by distant GPS stations. In addition, hypothetical landslide signals were rejected by careful experiment design informed by historical activity.