Seasonal perturbations in interseismic deformation of North-East India

GPS time series is manifestation of longterm plate motion along with contribution from some non-tectonic forcing. The non-tectonic forces causing deformation of the crust are, to say a few, ocean and body tides of the Earth, interhemispheirc exchange of fluids, snow melt, groundwater runoff, Glacial Isostatic adjustment, atmospheric loading, etc. The Earth as a whole responses to external forces as an elastic body. As it was shown by Darwin in 1882[Darwin, G.H. 1882], changes of weight of the atmospheric column due to variations of pressure result in crustal deformations called atmospheric pressure loading [http://gemini.gsfc.nasa.gov/aplo/] (The word 'crust' was not in use during Darwin's time. Instead he wrote as 'distortion of the upper strata of the Earth'). The magnitude of the effect can be upto 15-20mm and it can vary as much as 10mm within a 24 hour period. As a result, this physical phenomenon affects the coordinates of geodetic sites (VLBI, GPS, SLR, DORIS etc) and should be accounted for in all high precision geodetic analyses. On the other hand, the solid Earth crust also deforms due to the largest mass movement on the solid earth through global hydrological cycle, the exchange of water between ocean, land, and ice sheets[Braitenberg, C., and Zadro, M. 2001]. In this paper emphasis is laid on the study of seasonal perturbations in GPS time series due to atmospheric loading. The results are based on GPS measurements from 2004-2011 in North-eastern India. To asses the influence of atmospheric pressure loading on GPS time series we implement atmospheric pressure loading model in our GAMIT processing. It has been observed that model has most impact in vertical component. The difference in displacement reaches upto ~1-2cm in vertical whereas ~2-3mm in horizontal components. In addition, power spectral analysis is opted to find out the frequency of perturbations and seasonal signals in the time series. Power spectral densities demonstrate that colored noise is mostly abundant in the lower frequency domains whereas at higher frequencies the spectrum mostly corresponds to white noise. Power spectrum of time series are estimated using periodorgram method. The spectral indice of most components are within -1<α<1 and suggest a dominance of fractal white noise in the time series. Autocorrelation studies also favoring the presence of white noise by showing highest peak at iΔt=0. In order to investigate the seasonality, time series is decomposed to find its trend and the seasonal signal has been extracted from the power spectrum. The power spectrum of every component (North, East and Up) shows its biggest peak at 0.73cycles/year and next prominent peak is found at 2.15cycles/year at one of the stations. This assessment of deformation associated with non-tectonic forcing, like atmospheric loading, along with tectonics is vital since it not only affects the earthquake nucleation process and stress build up in the crust but also effectively alters the drainage pattern which is the lifeline of the North-Eastern India.