USING PSINSAR TO DETECT THERMOKARST-INDUCED SURFACE SUBSIDENCE IN EBOLING MOUNTAIN ON THE QINGHAI-TIBET PLATEAU OF CHINA

Abstract:

Thermokarst, a process that characterizes landforms caused by thawing of ice-rich permafrost, is a key indicator of permafrost degradation. Surface dynamics of thermokarst processes on Qinghai-Tibet Plateau (QTP) of China, is still poorly quantified or understood. It is also challenging to detect and measure surface subsidence due to loss of subsurface ice over a large area.

In this study, the Persistent Scatterer Interferometric Synthetic Aperture Radar (PSInSAR) technique is used to retrieve surface subsidence located at Eboling Mountain, near the northern edge of the QTP. Persistent Scatterer (PS) refers to stable natural or man-made object with good coherence in temporal domain. Based on the assumption that the residual phase of nearby PS is smaller than , we can retrieve the unwrapped phase associated with the sparsely irregular data. Using 17 L-band ALOS-1 PALSAR images taken from 2006 to 2011, we find linear deformation trends of up to 50 millimeters per year and seasonal subsidence of up to 80 millimeters over the area where thermokarst gullies are present (location: 38.01° N ,100.90° E , Figure 1), associated with thermokarst development and seasonal thawing of active layer, respectively. Furthermore, to the northwest and northeast of the Eboling thermokarst area, there are several small areas with large subsidence trends, which may also be related to thermokarst processes.

This study demonstrates that the PSInSAR technique has the ability to map and quantify thermokarst-induced subsidence spanning multiple years using ALOS-1 PALSAR images. Yet as the thermokarst processes are much more complex in nature than a simple subsidence trend, more efforts are required (1) to validate the PSInSAR results with in situ measurements, (2) to better separate thermokarst-induced subsidence from seasonal ground elevation changes by including winter PSInSAR data and accounting for frost heave processes, and (3) to quantify potential aliasing problem due to the sparse temporal sampling rate of the PSInSAR data. This study promises a potential of using PSInSAR to identify thermokarst landforms, map and quantify surface subsidence due to permafrost degradation, and assess its impacts over large areas on the QTP.