Integrating cost-effectiveness optimization and robustness analysis for low impact development practices design

Current low impact development (LID) practices may suffer challenge when tackling extraordinarily variable rainfall. New framework was proposed by integrating robustness analysis and cost-effectiveness optimization under various rainfall events. The cost-effectiveness LID schemes were generated by coupling Storm Water Management Model (SWMM) with genetic algorithm. Stochastic rainfall algorithm was proposed for evaluating robustness of each scheme via joint distribution function of daily precipitation and antecedent dry days. The results show that current studies might overestimate nonpoint source pollution (NPS) based on limited events, resulting in less cost-effectiveness LIDs. This new framework, which generated 1087 additional events especially high- and low- frequency ones, extracted robust LIDs among several initial schemes. In addition, heavy rain pattern was suggested as the design criteria to increase robustness of LIDs. Large impervious areas were also suggested as priority area. This study provides guidance for designing LIDs, especially in response to rainfall events with high variability.