Landslides in the Northern Darjeeling District, India: investigating dataset choice, international partnerships, and the use of global scale forecasting models for medium range prediction

Dolan, S. (2025) Landslides in the Northern Darjeeling District, India: investigating dataset choice, international partnerships, and the use of global scale forecasting models for medium range prediction. MPhil thesis, University of Reading. doi: 10.48683/1926.00122532

Abstract/Summary

This thesis explores landslide hazards within the framework of Disaster Risk Reduction (DRR) and the United Nations' initiatives, "Warnings for All" and "Information for All," aligned with the Sustainable Development Goals (SDGs). Research Chapter I investigates the integration of global-scale landslide inventories (LSIs) to address data gaps and limitations in the Global South. LSIs are critical for hazard analysis, yet their availability and accessibility are limited. Using geospatial techniques and data integration methods, this chapter combines global LSIs to improve spatial and temporal coverage. The results demonstrate an increase in the number of recorded landslide events and provide a foundation for analysing global landslide trends and identifying vulnerable regions. However, the findings reveal significant limitations in applying global LSIs at the local scale. In the Darjeeling district of the Northeastern Indian Himalayas, gaps in spatial and temporal coverage highlight the inadequacy of current freely available datasets for localised hazard assessment. Recommendations emphasise collaborative and equiTable approaches to recording, maintaining, and sharing global landslide inventories to support DRR strategies. Research Chapter II focuses on identifying historical landslides in the data-scarce Darjeeling district using ECMWF Re-Analysis 5th Generation (ERA5) total precipitation data and predefined intensity-duration (ID) thresholds. Combining findings from Chapter I with ERA5 data, this explores the potential of integrating meteorological datasets for landslide early warning systems. The analysis identifies "wet day" events and evaluates conditions preceding historical landslides recorded in the combined LSI. However, the approach is hindered by the lack of comprehensive landslide records, which limits the skill of the ID threshold method. The study underscores the need for improved landslide monitoring to enhance predictive capabilities. In conclusion, this thesis highlights the importance of integrating datasets to advance global and local landslide understanding. By aligning with the UN’s agenda, this research advocates for more inclusive and equiTable DRR practices.

Item Type	Thesis (MPhil)
URI	https://centaur.reading.ac.uk/id/eprint/122532
Identification Number/DOI	10.48683/1926.00122532
Divisions	Science > School of Archaeology, Geography and Environmental Science
Download/View statistics	View download statistics for this item