Comparing models of seasonal deformation to horizontal and vertical PBO GPS data

Abstract:

GPS monuments around the world exhibit seasonal displacements in both the horizontal and vertical direction with amplitudes on the order of centimeters. For analysis of tectonic signals, researchers typically fit and remove a sine function with an annual period, and sometimes an additional sine function with a semiannual period. As interest grows in analyzing small-amplitude, long-period deformation signals it becomes more important to accurately correct for seasonal variations. It is well established that the vertical component of seasonal GPS signals is largely due to continental water storage cycles (e.g. van Dam et al., GRL, 2001). Other recognized sources of seasonal loading include atmospheric pressure loading and oceanic loading due to non-steric changes in ocean height (e.g. van Dam et al., J. Geodesy, 2012). Here we attempt to build a complete physical model of seasonal loading by considering all of these sources (continental water storage, atmospheric pressure, and oceanic loading) and comparing our model to horizontal and vertical GPS data in the Western US. Atmospheric loading effects are computed from the National Center for Environmental Prediction 6-hourly global reanalysis surface pressure fields; the terrestrial water loading and ocean loading models are generated using SPOTL (Some Programs for Ocean Tide Loading; Agnew, SIO Technical Report, 2012) and parameters from NASA’s Land Data Assimilation Systems and the Estimating the Circulation and Climate of the Ocean model, version 4. We find that with a few exceptions, our seasonal loading model predicts the correct phases but underestimates the amplitudes of vertical seasonal loads, and is a generally poor fit to the observed horizontal seasonal signals. This implies that our understanding of the driving mechanisms behind seasonal variations in the GPS data is still incomplete and needs to be improved before physics-based models can be used as an effective correction tool for the GPS timeseries.