Landslide Monitoring with ALOS/PALSAR data in Mountain Area

Abstract:

InSAR is a relatively new technique with a high potential in earth observation, which has made great success in monitoring urban areas deformation. At present, although there are a considerable number of applications in the complicated mountain areas, it is hard to obtain sufficiently high-density stable point targets in these regions. So scientists have been trying to solve this bottleneck problem and improve accuracy in mountain areas.

In this work, we present the landslide measurement result in complicated topographic region using ALOS/PALSAR data. The test area is selected around highlands of the boundary between China and India. We choose 13 scenes of ALOS/PALSAR images from May 2007 to February 2011. The main landforms in this experimental region are bare rock and soil, ice and snow, the vegetation in the alpine area. Due to the lithology of the strata and the undulations extent of the terrain, it is prone to cause landslides in the event of rainfall, earthquakes, snow melt or human activities.

The traditional PS algorithm has a higher requirement for a long time series data collection, especially in low-coherence area of vegetation cover. As the collected data and stable points are relatively less in this experimental area, we plan to study the time series InSAR analysis coherence model and error model, and extend its application to the extra-urban regions. The approach has been carried out to increase the density of stable points, which are mainly distributed on the top of mountain and ridge areas. And using the 13 images we find several subsidence areas by this technique. The result shows that the top of mountain is relatively stable and the suspected landslide areas are mainly along the ridge, which is in accordance with the actual situation. Then the mechanism and stability analysis of landslide is discussed. Meanwhile, some other measurement data in experimental area is available for cross validation, such as optical data and TerraSAR-X data. And a comparative study of the different data is analyzed as well. This study will extend time series InSAR analysis to landslide monitoring, fault deformation and other geological science research applications. Moreover, this research is also conducive to monitor the periodical changes of surface and climatic and environmental changes over time.