AN INTEGRATED FRAMEWORK TO QUANTIFY FLOOD RISKS AND DECODE IMPACTS ON INTERDEPENDENT INFRASTRUCTURE SYSTEMS OVER FLOOD PRONE MOUNTAINOUS LANDSCAPES

Abstract

Flooding is increasingly becoming more frequent and severe in terms of magnitude and intensity globally. Although advanced flood models are developed by experts worldwide, more research is required to thoroughly comprehend how well they perform, particularly over mountainous terrains, which requires comprehensive attention. The less-addressed question gains importance for flood prone, resource-constrained nations, as adopting capital resources, including procurement of profit making software models with appropriate technical capacity, is strenuous. Bhutan, also famously known as the ‘carbon-negative country’, is mountainous in Asia, experiencing unprecedented flooding due to its delicate topography and the effects of changes in the climate. Unfortunately, a thorough data-driven modelling approach is missing to identify potential flooding risks in this region. The current study quantifies flood hazards and associated exposures while considering a robust hydrodynamic flow model over Chamkhar communities along the Chamkhar Chhu River basin, a flood-prone area The U.S. Army Corps of Engineers recently released open-source HEC RAS v6.3, whose effectiveness for flood inundation modelling has yet to be explored, is used to compute a set of flood inundation and hazard maps. The linked 1D-2D flow model structure is built to simulate multiple flooding situations corresponding to design discharge and rainfalls for 1 in 50, 1 in 100, and 1 in 200 years return periods. A rectified high-quality Digital Elevation Model (DEM) obtained using the ALOS-PALSAR satellite product was used to minimize ambiguities in the ultimate set of flood information. The simulated flooding hazards for the communities along the Chamkhar Chhu River are expressed in terms of flooding depth, velocity, and product of depth and velocity. A list of statistical performance values is calculated by evaluating the model's effectiveness of the simulated inundation maps with the historical flood map available in the Global Flood Database collected by MODIS and Terra satellites. To establish the efficiency of HEC-RAS in reproducing flood hazards, the inundation statistics are compared with TUFLOW model outputs. at 1D and 1D 2D coupled levels. Comparing these performance statistics underscores the applicability of the HEC-RAS flood model to represent the flood hazards in mountain terrains. It was observed that a substantial portion of the central region is at a potential threat of very high flood risk as the simulated flood velocity exceeds 1.6 m/s and the flood depth surpasses over 3m. The current study also evaluates the degree of population exposure. With increasing flood return periods, a systematic increase in the exposure of the population to high and very high flood hazards was noticed. In contrast, the number of people exposed to low and very low flood hazards simultaneously decreased. The framework of flood hazard mapping was extended to determine the impacts on the interdependent systems falling within the flood-prone zone of the Chamkar Chhu watershed. It becomes apparent that modern infrastructure systems are becoming progressively more dependent on one another to operate and provide required services effectively. The failure of one component in the network system can result in disastrous cascading effects on other interconnected infrastructures. This study puts forth a network-based strategy to examine the risks of riverine flooding to the critical infrastructure of an airport in a hilly region. The Strongest Path Method (SPM) was executed through Python-based programming, representing a complex infrastructure network system, and providing a means to prioritize risk mitigation measures. Python libraries like ‘NetworkX’ and the ‘Igraph’ aid in solving algorithms for determining the strongest path in a network. The Strongest Path Method is developed in the algorithm using Python codes to suit the best needs of determining the associated flooding risks of each infrastructure. The results suggest that, in comparison to the 1 in 50 and 1 in 100-year return periods, the risk index values of most structures increase substantially over the return period of 1 in 200 years. The proposed framework of flood hazard mapping and impacts on interdependent infrastructure systems is novel and promises to contribute significantly to effective flood management. The study provides a platform for deciding affordable non-structural and structural flood risk intervention strategies for the communities that will minimize the impact of recurrent flooding in the future. Given the effective and efficient numerical computation performance of HEC-RAS v6.3 over a sensitive topography and robust program-based risk quantification practice, the study encourages adopting the model to precisely identify flooding risks over any similar mountainous regions for effective flood management.