Advanced Landslide Inventory Mapping with Multispectral Data in Munnar, Kerala

Accurate landslide modeling is essential for effective risk management, often constrained by the limitations of existing inventory data. This study focuses on using satellite data to create a detailed landslide inventory. To validate this approach, the 2018 landslides in the Munnar region is used as a case study. The study demonstrates the potential of Sentinel-2 satellite data for generating a comprehensive landslide inventory in Munnar, Kerala. By analyzing vegetation changes with Normalized Difference Vegetation Index (NDVI), 112 precise landslide polygons were identified and mapped, surpassing the existing 28 points. This detailed inventory provides an accurate spatial understanding of the 2018 landslide hazard in Munnar, going beyond traditional methods. The study's significance lies in its contribution to our understanding of topographic influences on landslide occurrences, with over 90% of landslides occurring within a specific slope range of 10–35° for the chosen study area. The performance evaluation through a confusion matrix reveals the strengths and limitations of the NDVI method with a spatial accuracy of 0.9996, sensitivity of 0.8911, a specificity of 0.9997 and a precision of 0.6795. The study's findings have implications for landslide susceptibility assessment and risk management in the Western Ghats region, emphasizing the importance of accurate inventory data in predicting landslide susceptibility.