Time series SAR interferometry approach for landslide identification in mountainous areas of Western Ghats, India

Monitoring of Earth’s surface processes is efficiently performed by Differential SAR interferometry (DInSAR) techniques with the availability of open access Copernicus Sentinel-1A and 1B SAR data products having 6 days temporal resolution. With over 750 landslides triggered in Kodagu district, Karnataka, India, interferometric wide swath (IW) mode Sentinel-1 datasets are used to identify landslide locations. Interferograms from DInSAR obtained using Sentinel-1 VV polarization data are corrected using zenith total delay (ZTD) maps obtained from GACOS, which successfully assisted to identify major landslides in the study area. Sentinel-1 DInSAR analysis has limitations over dense vegetative regions and terrain conditions (particularly in mountainous regions) due to spatial and temporal decorrelation. With the aid of InSAR time series analysis, integration of deformation recognized and land use/land cover pattern helps experts in providing near real-time mitigation measures during a disaster event. Research highlights Due to heavy rainfall, Kodagu District recorded 104 landslides between 15th and 17th August 2018. Advanced InSAR approach is adopted to identify the location of landslides in Kodagu district, Karnataka to explore the advantages of utilising SAR Datasets during an event. GACOS corrected Sentinel-1 derived InSAR results assists in identification of larger landslides. Limitations are experienced due to the dense vegetation and loss of coherence between the InSAR pairs. Research suggested to use L and S band datasets along with corner reflectors installed in the mountainous regions to obtain results at higher accuracy.