Integrated PSInSAR and SBAS-InSAR analysis for landslide detection and monitoring
Sajid Hussain, Bin Pan, Wajid Hussain, Meer Muhammad Sajjad, Muhammad Ali, Zeeshan Afzal, M. Abdullah-Al-Wadud, Aqil Tariq
Abstract
The Lower Hunza section of the Karakoram Highway (KKH) in Northern Pakistan is highly prone to landslides due to gravity, tectonic stress, and erosion from glaciers and rivers, posing significant threats to infrastructure, communities, and KKH. Accurate and efficient detection and monitoring are essential for disaster monitoring and early warning. This study utilized Sentinel-1 SAR data over two years, combining Permanent Scatterers Synthetic Aperture Radar (PSInSAR) and Small Baseline Subset (SBAS-InSAR) technologies to address individual technique limitations, such as potential error accumulation of SBAS and sparse scatterer distribution of PSInSAR. The PSInSAR technique has a 12.5 % higher mean coherence than SBAS, but the density of pixels in the PSInSAR method is nine times less than that of SBAS. The combined approach revealed average displacement rates ranging from 57 mm/year (uplift) to −146 mm/year (subsidence) and identified 36 potential landslides. The landslides located in Khana Abad and Nagar Khas were found to be the most active, displacing with the mean slope deformation of 23.42 and 54.26 mm annually. The spatial distribution and deformation characteristics were meticulously analyzed, highlighting the significant influence of regional geological structures , including lithology and fault lines, on landslide activity.