<p>Deep-seated, slow moving bedrock landslides are significant natural disasters with severe socio-economic repercussions. During the past decades, an acceleration of these hazards has been reported globally due to changes in seasonal freeze-thaw cycles, permafrost thawing, infrastructure development and other anthropogenic sources, changes of precipitation and groundwater levels, and variation in seismic activity. Interferometric Synthetic Aperture Radar(InSAR) is a powerful tool to map landslides movement from space and the Sentinel 1 C-band radar mission provides a high temporal resolution data source to investigate seasonal and intra-annual changes of landslide behaviour.</p><p>To construct a 2D/3D displacement field, we decompose a combination of different look angles and InSAR ascending and descending tracks of different sensors including Sentinel and ALOS 1 PALSAR data. The ionospheric delay for InSAR observations is estimated with a split range-spectrum technique because significant ionospheric total electron content variation is common in our study area in the Central Andes. Both statistical phase-based and weather model estimation approaches are implemented to minimize the effect of tropospheric signal on InSAR observations.</p><p>Our observations identify several areas with rapid translational slide movements exceeding 5-10 cm/y. Multi-annual and inter-annual behaviour of deformation is extracted through time series analysis and a hierarchical clustering approach is used to identify geographic areas with similar characteristics and rates. We show the wide-spread spatial distribution of unstable hill slopes in the Eastern Cordillera of the south-central Andes, especially at high elevations where field observations are difficult. We identify driving forces to be a combination of pre-existing geologic structures and climatic parameters.</p>
Crustal deformation in the south-central Andes backarc terranes as viewed from regional broad-band seismic waveform modelling