CLIMRESILINFRA: ADAPTIVE DESIGN OPTIMISATION AND GIS-BASED DECISION SUPPORT FOR CLIMATE-RESILIENT INDIAN ROAD INFRASTRUCTURE

Abstract

The compounding impacts of climate stressors like extreme flooding, heat-waves and increased monsoon variability are driving infrastructure degradation at rates which are higher than currently assumed in design of India's national highway network. Vulnerability characterisation and degradation forecasting are diagnostic, but the ability to translate those lessons into actions for adaptive design and evidence-based policy is an unmet need. In this paper, the adaptive engineering and governance modules of ClimResilInfra are presented, with only (GA–ANFIS Adaptive Design Optimisation) and (GIS Based Decision Support and Policy Framework) being discussed. Five adaptive design strategies are compared to a novel Fuzzy Resilience Index (FRI) using a Genetic Algorithm (GA) and Adaptive Neuro-Fuzzy Inference System (ANFIS) which optimizes the cost-resilience frontier for the new climate scenarios, SSP2-4.5 and SSP5-8.5. The GA optimised hybrid strategy (S4) increases FRI by 107% compared to baseline, and lowers the normalised lifecycle cost by 3%. Life Cycle Assessment (LCA) shows a co-benefit – S4 has 16% less embodied GHG emissions than baseline. For the estimated annual NH maintenance expense, ANFIS-derived maintenance scheduling decreases the expense by 18-22% from fixed 5-year cycles to risk-stratified schedules of 3.6-11.8 years. The GIS-based Decision Support System (DSS) integrates FRI scores with climate vulnerability indices, exposed population data and lifecycle costs of 20 NH interventions across the country and ranks them using MCDA through the generation of a Priority Score. The highest national action (PS = 0.847) is biorial NH scour protection. With SSP5-8.5, 35% of NH segments will exceed critical thresholds by 2050. There are three key amendments to IRC:SP-84 design standards identified as policy gaps.