Author(s)

MARÍA JAYA-MONTALVO, EDGAR BERREZUETA, JOSÉ CUERVAS-MONS, FERNANDO MORANTE-CARBALLO

Abstract

The Paute River basin in the south-central Andes (Ecuador), which supplies approximately 38% of the national hydroelectric power generation, is located in a region highly exposed to geological hazards associated with surface runoff erosion, active tectonics and mass-movement processes affecting the hydroelectric complex. This study proposes an integrative approach to generate potential landslide density maps for the slopes surrounding the Paute-Molino dam using Sentinel-1 SAR data (vertical transmission/horizontal reception (VH) and vertical transmission/vertical reception (VV) polarisations), satellite-derived accumulated precipitation and topographic criteria. A pre- and post-event backscatter change index was calculated to identify potentially affected areas, with results refined using slope, curvature and water-body masking. The resulting maps were validated using a local landslide inventory. This study analyses variations in the density of affected areas under VH and VV polarisations during a recorded extreme precipitation event. The VV polarisation showed a greater spatial correspondence with the mapped landslide inventory. In contrast, VH polarisation, which is more sensitive to vegetation structure, generated a broader spatial dispersion of high backscatter anomalies. VV backscatter provided a clearer response over exposed and geomorphologically disturbed surfaces, reinforcing its applicability in mountainous tropical environments such as the Ecuadorian Andes, where landslides frequently involve vegetation removal and soil exposure. However, the relatively limited number of Sentinel-1 images used in the multitemporal stacks may increase speckle noise and introduce false-positive detections associated with vegetation dynamics, soil moisture variability and hydrological changes. Therefore, the generated heatmaps should be interpreted as indicators of potential geomorphological disturbance and landslide-prone areas. Future studies should incorporate larger SAR stacks and multi-source validation strategies to improve detection accuracy and reduce uncertainty.

Keywords

GEE, mass movements, remote sensing, hazard assessment, GIS