Development of a Multi-Criteria based Multi-Objective Simulation – Optimization Framework Integrated with Hydrologic Model and Evolutionary Algorithm for Planning, Design, and Analysis of SuDS at a river-basin scale

Sustainable Drainage Systems (SuDS) improve storm water management by leveraging surface runoff in urban areas while limiting the negative impacts. Integrating SuDS into the urban environments requires a systematic planning and design framework across various spatial scales. Existing studies have utilized GIS-based multi-criteria methods and Spatial Decision Support Systems (SDSS) to identify suitable locations for SuDS. However, these approaches are often subjective and do not account for the hydrologic characteristics of the catchment. Integrating SDSS tools with the Simulation – Optimization (S – O) frameworks has shown potential for addressing multiple storm water management objectives. Current S – O frameworks typically focus on providing solutions based on “which element” and “where to locate” or “which element” and “how much to allocate”. However, an ideal SuDS – SDSS should answer all the three questions: “which element”, “where to locate” and “how much to allocate”.

To holistically address this problem, a novel framework integrating SDSS with a hydrologic model using an evolutionary algorithm is proposed. The framework begins with the selection of appropriate thematic layers and water balance layers obtained from a hydrologic model – Soil and Water Assessment Tool (SWAT). The weights for the selected thematic layers are calculated using Normalized Mutual Information, an objective method that quantifies how well the chosen thematic layers explain the hydrologic response of the catchment. The second part determines suitable sites for SuDS elements, viz., bio-retention cell, infiltration trench, permeable pavement, rain garden, swales through spatial overlay analysis. The S – O part of the framework involves synthetic modelling of Hydrologic Response Units (HRUs) using the Storm Water Management Model (SWMM). This model is then coupled with the Multi-Layer Green-Ampt (MLGA) based SuDS modules to simulate the runoff response for selected design storms of various return periods. Finally, the optimal combination of SuDS (“which element”) and area to be allocated are determined using the Non-dominated Sorting Genetic Algorithm (NSGA – II). The framework produces pareto-optimal solutions, enabling decision-makers to evaluate trade-offs and develop policies for planning and development.

The framework is applied to the Adyar basin, covering an area of 830 km². The optimal solutions obtained are implemented and simulated in the SWAT model to evaluate the peak flow and runoff volume reductions at the sub-basin scale. This research provides insights into how various combinations of SuDS implemented at the HRU level reduce peak flow and runoff volume at the sub-basin scale and how SuDS influence the water balance components, offering critical insights for urban planners and water resource managers.