LAUSR

The study aims at calibration of the storm water management model (SWMM) with non-dominated sorting genetic algorithm-III (NSGA-III) for urban catchment in Hyderabad, India. The SWMM parameters calibrated were Manning's roughness coefficient (N), depression storage for pervious and impervious areas (DP and Di), sub-catchment width (W), curve number (CN), drying time (dry) of soil and percentage of imperviousness (I). The efficacy of calibration was evaluated by comparing the observed and simulated peak flows and runoff using goodness-of-fit indices. The calibration takes into consideration eight event rainfalls resulting in eight calibrated sets. Weights of goodness-of-fit indices were estimated and the best calibrated set was further validated for five continuous rainfalls/runoffs. Simulated runoff volume and peak runoff over the five continuous rainfalls deviated by 7–22% and 2–20% with respect to observed data. Results indicated that parameters calibrated for an event rainfall could be used for continuous rainfall-runoff modelling. The effect of catchment delineation scale on runoff was also studied. The study indicated that output of the model was sensitive to variation in parameter values of infiltration and imperviousness.