Item

Abstract

This study attempts to constrain the lithospheric strength in Central Tibet by studying the rebound of the lithosphere and uppermost mantle subjected to loading or unloading due to lake Siling Co water level fluctuations.This lake is a large endhoreic lake at 4500~m elevation located North of the strike-slip right lateral Gyaring Co fault, and South of the Bangong Nujiang suture zone, on which numerous left-lateral strike slip faults are branching. The Siling Co lake water level has strongly changed in the past, as testified by numerous traces of palaeo-shorelines, clearly marked until 60 m above present-day level. Altimetric measures show that during the period 2003-2008 the Siling Co water level increased by about 0.7~m/yr, a remarkably fast rate given the large lake surface (1600~km2). To extent the lake level observation duration, we extract the lake contour from all cloud-free LANDSAT images available on the USGS GLOVIS server. The lake surface, used as a proxy for lake elevation, shows that the water level in the Siling Co lake in Tibet was more or less stagnant from 1973 to 1999 and increased between 2000 and 2010 at an average rate of 0.8 m/yr. A clear seasonal signal is superimposed on the interannual trend. The ground motion associated to the water level increase is studied by InSAR using all ERS and Envisat data on tracks 491 and 219 in the period 1992-2010, obtained through the Dragon ESA-MOST cooperation program. A redundant network of small baseline interferograms is computed with perpendicular baseline smaller than 500~m. The coherence is quickly lost with time (over one year), particularly to the North of the lake because of freeze-thaw cycles. The interferograms covering the period 1992-1999 show no detectable deformation, whereas the ones covering the period 2000-2010 present bowl shape pattern centered on the lake that extend from the shore to about ~100 km from the lake center. The amplitude is about 5 mm/yr close to the lake shores. The interferograms are also affected by decorrelation noise, residual orbital ramp and atmospheric delays. To increase the signal to noise ratio, they are first analysed in time assuming a constant deformation shape. We then obtain the temporal evolution of the deformation amplitude : the amplitude remains constant for the period 1992-1999, and increased from 2000 until 2010. This curve follows closely the lake level temporal evolution. An average velocity map for the period 2000-2010 is also produced to better assess the shape of the ground displacement. Both the amplitude temporal evolution and the average velocity map are used to explore possible rheological models : the elastic model could explain the amplitude evolution if static moduli are set about twice lower than the dynamic moduli derived from the lithosphere seismic velocity profiles. A visco-elastic model with a low viscosity lower crust can also adjust the data. We will discuss to what extent the models can be constrained considering the available data and its noise.