A comprehensive review on available tools for modeling low impact development (LID) techniques in lab, small and large-scale watersheds

Applying Low Impact Development (LID) practices within urban area has become one of the most popular method for stormwater management and increasingly for urban flood management. To evaluate the performance of LID practices in each watershed a suitable mathematical model is required. In this review, eleven popular existing LID models (e.g. GIFMOD, HEC-HMS, PCSWMM, MIKE-URBAN, WBM, WIN-SLAMM, etc.) are compared and contrasted. Comparison was carried out based on general features including the model application, type of in-built GI in model, temporal data resolution, spatial data as well as hydrological and hydraulic modeling approaches. Results show that all models have different applications: a few are only suitable for design optimization of LIDs; while others are better suited for more advanced engineering design. Most of the reviewed tools can model LIDs explicitly, however, many have severe limitations. Few models (HEC-HMS, HYDRUS, PCSWMM and MIKE-URBAN) allow for full temporal resolution, while other models only accept distributed daily or lumped amounts of rainfall. Some of the models have built-in Geospatial Information Systems (GIS) capabilities, while many need extra effort to handle spatial data. Also, LIDs are treated as a new watershed element and cannot be considered as an existing sub-basin within the watershed in most of the models. In terms of hydrological analysis, wide ranges of methods for calculating water balance of LID are used in different approaches influenced by the particular model application. In addition, some of the reviewed models (PCSWMM and MIKE-URBAN) can model LID practices incorporated with complex networks and perform a full hydrodynamic numerical modeling of regular Stormwater Collection Networks (SCN). While most of the models have been recently updated to incorporate LID technologies in newer revisions, several important features are still missing, such as the ability to model multi-layer growing media for infiltration-based LIDS, tree canopy and plant uptake effects, different spatial scales, subsurface flow modeling, cold climate conditions, LID treatment trains within a watershed, automated calculation of area treated by LID practices, automated design of LIDs and SCNs, and adding model calibration tools to overcome uncertainties associated with modeling green infrastructures, soil characteristics etc.. Results from this review provide a primary insight into existing LID models to facilitate selecting the best-suited model for the required modeling objective. Also, for future further studies and development of LID modeling tools recommendations were presented.