A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Development and Sustainability

Thesis (M. Sc.) - Asian Institute of Technology, 2026

Urban flooding increasingly jeopardizes transportation in rapidly expanding Himalayan cities; yet, there is limited evidence of road damage and economic losses on a corridor-wide scale. This study analyzes the impact of flooding on road infrastructure in the Kathmandu Valley, Nepal, by combining comprehensive flood mapping with detailed damage and cost assessments of the Nakkhu flood incident. The study aimed to identify flood-prone sites in the Kathmandu Valley, assess flood-related damage to existing roadways, and evaluate the direct and indirect economic losses associated with such events. Methodologically, the first objective used Sentinel-1 SAR imagery to map the 2024 flood through threshold-based change detection and Random Forest classification. The second and third objectives focused on the completed section of the Nakkhu corridor and nearby bridge approaches, where inundation scenarios were translated into infrastructure damage using an adapted relative depth-damage function for transport assets, together with unit-cost replacement for direct losses and factor-based estimation for indirect losses. These analyses were supported by secondary spatial, hydraulic, and infrastructure data within a GIS-based framework. The results indicate that flood effects in the Kathmandu Valley are spatially uneven and primarily influenced by terrain, rivers, and urban land use. The analysis using the specified thresholds yielded approximately 9.38 km² of inundated regions, whereas a more thorough assessment of hydrological alterations in the landscape was achievable with a more precise Random Forest classification method. In the Nakkhu area, damages and losses exhibited a nonlinear relationship with flood depth, with bridges and adjacent roadways being particularly susceptible to flooding.