- DExtER_Lab-IIT Mandi
- Sep 5, 2025
- 2 min read
One month back, on 5th August 2025, a massive flash flood, accompanied by debris flows, occurred in the Kheer Ganga catchment, severely affecting Dharali village and nearby localities in Uttarkashi district, Uttarakhand, India. The disaster caused extensive damage, including the destruction of homes and buildings, and resulted in significant loss of life. Approximately 64 people were reported dead, with several others still missing. In response, large-scale search and rescue operations were launched by the Indian Army, Indo-Tibetan Border Police (ITBP), State Disaster Response Force (SDRF), and National Disaster Response Force (NDRF).
Various organizations offered differing perspectives on the cause of the disaster. The general consensus, based on expert opinions from eminent scientists, was that the event was triggered by a landslide or debris flow due to a cloudburst, with no evidence of glacier activity. However, no substantial evidence of a landslide capable of initiating such a catastrophic event was found even after 1 month of the event.
This hypothesis was explored by a study conducted by the Disaster Extreme and Environmental Remote Sensing (DEXTER) Lab at IIT Mandi assessed landslide susceptibility and risk across the Indian Himalayan Region (IHR). The study utilized a landslide susceptibility (LSZ) map prepared for the entire Indian Himalayan Range, available at this link (https://lnkd.in/gtUKPFfU). The map was generated at a 30m spatial resolution using the Random Forest method, which incorporated 14 geospatial layers. Approximately 26,000 landslide training points were used, resulting in an impressive AUC (Area Under the Curve) of 87.50%, demonstrating the model's strong predictive accuracy in identifying landslide-prone areas.
Interestingly, while Dharali village, situated along the banks of the Bhagirathi river, falls within a high landslide susceptibility zone, the upper catchment of the stream is categorized as a moderate to low landslide susceptibility zone. Furthermore, a landslide risk map, which takes into account factors like population, GDP, and other relevant attributes, indicates that the region falls within a low landslide risk zone, as shown in the figure below.
These findings suggest that, although the region exhibits moderate to high landslide susceptibility, the overall landslide risk remains low. Therefore, it is likely that this disaster did not have its origins in a landslide. Instead, it could be attributed to other factors, such as glacial activity, possibly the breaching of a temporary glacial lake, or the melting of snow or subsurface ice (permafrost) due to the higher-than-usual rainfall in the area.
For more information on landslide susceptibility and risk in the Indian Himalayan Region, visit this link (https://lnkd.in/gtUKPFfU)
