Climate-Adaptive Urban Flood Hazard Framework: Risk Evaluation and Resilience Optimization

The phenomenon of urban flooding has evolved into a critical challenge in contemporary urban environments. Urban floods can occur due to rapid unplanned urbanization, frequent climatic extremes, and poor urban drainage infrastructure (URDAN). Incidents of urban flooding have become frequent in recent history. Hence, it is imperative to strengthen the flood resilience of cities. The present study proposes a holistic, multi-faceted flood resilience framework that integrates the critical elements of past urban floods, simulates existing URDAN using present and future climate extremes, and evaluates the integration of Low Impact Development (LID) in enhancing resilience of cities. Historically, landuse change and climate variability are quantified along with a dedicated assessment of previous urban floods. For the present, urban flood risk zonation and hotspot identification (UFRZHI) ascertain areas at higher flood risk. Performance of URDAN in a high flood risk zone is then evaluated using Stormwater Management Model (SWMM). The URDAN is optimized for performance using LID elements, while General Circulation Models (GCMs) integrate future rainfall projections.

The study analyzed the 2019 urban flood and found that climate change was the immediate cause of the floods, as intense rainfall overwhelmed the existing URDAN, while the week-long inundation resulted from poor management. The builtup area increased from 29.9 to 48.5 sq. km., while vegetation declined from 64.5 to 48.7 sq. km. between 1990 and 2020. The long-term historical (1950–2020) climate variability assessed using Modified Mann-Kendall and Centroidal Day (CD) shifts shows a forward shift in monsoonal and annual rainfall in recent decades. The variability in total rainfall is more pronounced post-1985, while rainfall during monsoons has intensified. An increase of 64.53 mm (18.9%) in surface runoff is observed despite decreasing rainfall trends.

A normalized multivariate approach ranks the performance of downscaled and bias-corrected GCMs. The ensemble of the top three optimal GCMs is used for future predictions for Shared Socioeconomic Pathways (SSP) (SSP245, SSP370, and SSP585) scenarios. The historical annual maximum hourly rainfall series was fitted to Generalized Extreme Value distribution and alternating block method develops design storm hyetographs. The UFRZHI analysis identifies the most flood-prone areas, the existing URDAN of which is comprehensively evaluated using SWMM for 2- (baseline), 5-, 10-, and 25-year return periods. Simulation results show that the baseline URDAN fails and the time to peak ( Tp) is 59 minutes for 2-year return period. The peak outlet discharge (Qpeak ) increases and  remains constant with higher return periods and warmer climate forcings.

Permeable pavement, bioretention cells, and green roofs included URDAN shows a sharp decrease in Qpeak and a 22-minute delay in Tp. The inclusion of LID eliminates URDAN failure and reduces the runoff in subcatchments by 40-45%. The performance of LIDs saturates under higher return periods and warmer climatic scenarios. The framework in the present study can model risk prone URDAN and alleviate the failure stress through the inclusion of LID and climate uncertainty. The proposed framework can be used to develop an urban flood resilient city under climatic extremes.