Mr Shashank Kumar Anshu will present his APS as per the details below:

Date: 26th February 2025 (Wednesday)  

Time:   1030 - 1130 hrs.    

Venue: C-TARA Conference Room No. 2   

Title:  Community-Based Landslides Early Warning System along the Western Ghats of Maharashtra   

Guide: Prof. Parmeshwar D. Udmale   

RPC Members: Prof. Pankaj Sekhsaria, Prof. Raaj Ramsankaran   

Abstract :

Landslides severely threaten life, infrastructure, and the environment, disrupting ecosystems and socio-economic stability. The United Nations Agenda 2030 highlights disaster risk reduction as crucial, recognizing that hazards like landslides exacerbate community vulnerabilities. The 2023 UN SDG Report states that over 1.3 billion people live in landslide-prone areas, with millions affected annually. India, among the top five most vulnerable countries, records over 200 landslide-related deaths each year, with 12.6% of its land area and 30% of its population at risk—especially in the Himalayas and Western Ghats. The National Disaster Management Authority (NDMA) emphasizes the need for effective Early Warning Systems (EWS) and risk management strategies. Maharashtra’s Western Ghats, with its steep terrain and increasing extreme weather events reported by the Indian Meteorological Department (IMD), faces heightened landslide risks. Limited monitoring and community preparedness further worsen vulnerabilities. In this context, it is essential to understand existing landslide disaster management approaches and the associated challenges and opportunities adopted locally to strengthen the management strategies further.

This research aims to review and design a comprehensive Community-based Landslide EWS (CLEWS) tailored to local communities in Maharashtra. The first step involves identifying high-risk zones utilizing multi-source datasets, including satellite imagery, IoT-based ground monitoring, and historical landslide records. A dynamic landslide susceptibility map will be created using advanced machine learning algorithms such as Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Convolutional Neural Networks (CNNs). These models will be optimized to improve the accuracy of landslide predictions by incorporating various factors such as terrain, hydrology, geology, and human activities.

The innovative aspect of this research lies in its intention to integrate Indigenous knowledge and community-reported indicators into the early warning framework. This will be done through community workshops and consultations, where local knowledge holders will be invited to share their insights and observations. These insights will be systematically incorporated into the EWS, ensuring they are scientifically robust and feasible to adopt. Effective risk communication strategies will also be emphasized to ensure that early warnings reach low-literacy and remote populations through accessible methods such as colour-coded flags, visual pictograms, Interactive Voice Response-based voice alerts, and community radios.

The pilot study areas—Malin, Taliye, and Irshalwadi—have been selected based on their historical vulnerability to landslides and socio-economic challenges. The initial phases of the research will focus on real-time testing of the proposed CLEWS framework in collaboration with stakeholders such as the State Disaster Management Authority (SDMA), the Geological Survey of India (GSI), and the Indian Meteorological Department (IMD). By bridging scientific advancements with community-driven approaches and involving key stakeholders, this research aims to lay the groundwork for an inclusive and adaptive EWS that has the potential to significantly enhance disaster resilience and ultimately save lives in landslide-prone regions of Maharashtra.

Keywords: Community-Based Disaster Management, Crowdsourcing, Early Warning System, Geospatial Data, Landslides, Machine Learning.