IIT Mandi’s New Landslide Warning System Aims to Save Lives Across the Himalayas
As climate change continues to trigger more frequent and intense natural disasters worldwide, landslides have emerged as a growing threat in the fragile Himalayan region. In a significant step towards disaster preparedness, researchers at the Indian Institute of Technology (IIT) Mandi have developed a fully operational Landslide Early Warning System (LEWS) designed specifically for the Indian Himalayan Region (IHR).
The innovative project has been led by Prof. Dericks Praise Shukla from IIT Mandi’s School of Civil and Environmental Engineering, along with research scholars Ankit Singh and Nitesh Dhiman. The system is expected to play a crucial role in reducing the loss of life and property caused by landslides, especially during the monsoon season.
A Landslide Early Warning System works by continuously analyzing terrain susceptibility and real-time rainfall data to predict the likelihood of landslides. Based on these assessments, warnings are issued for areas facing elevated risk, allowing authorities and communities to take preventive measures before disaster strikes.
Explaining its importance, Prof. Shukla said the system begins providing daily landslide forecasts as soon as the monsoon season starts. Through a dedicated web-based platform, it identifies regions at higher risk, enabling timely evacuations and better disaster management planning.
He emphasized that satellite-based warning systems are among the most effective tools for disaster risk reduction because they convert scientific observations into practical, real-time decisions. According to him, a forecasting platform covering the entire Himalayan belt can significantly strengthen preparedness, speed up response efforts, and improve coordination among disaster management agencies.
What makes IIT Mandi’s LEWS unique is its scale. While several landslide warning systems exist in India, most are limited to specific locations or smaller geographical areas. IIT Mandi’s system covers the entire Indian Himalayan Region, making it one of the most comprehensive landslide forecasting platforms developed in the country.
The researchers built the system through a multi-stage scientific process. They first analyzed nearly 26,000 recorded landslides from the Geological Survey of India database to prepare a detailed landslide susceptibility map. Various environmental and geological factors associated with landslide occurrence were then integrated using advanced ensemble machine-learning techniques.
The team next developed the P-RIL (Probability of Rainfall-Induced Landslides) model using data from NASA’s Global Landslide Catalogue and rainfall information obtained from IMERG satellite datasets. Since rainfall patterns constantly change, the model dynamically considers rainfall conditions from the preceding 15 days, making forecasts more responsive to evolving weather situations.
The final daily forecast is generated by combining the static susceptibility map with the dynamic rainfall-induced landslide model. To make the information easier to understand, the system categorizes forecasts into different risk levels using percentile-based classifications.
To ensure easy access for government agencies, disaster managers, and other stakeholders, the IIT Mandi team has also launched a Google Earth Engine-based web portal. Users can view landslide forecasts for the current day as well as the previous three days. The platform also allows users to download forecast bulletins in PDF format and receive location-specific alerts through WhatsApp.
Researchers believe the operational Landslide Early Warning System will significantly strengthen disaster preparedness efforts across the Himalayan region. By delivering timely, location-specific warnings, the technology has the potential to reduce economic losses, improve emergency response, and most importantly, save lives in landslide-prone areas.
The Landslide Early Warning System can be accessed online at:
https://dexter-499110.projects.earthengine.app/view/lews
