Determination of runoff response to variation in overland flow area by flow routes using UAV imagery

Abstract

Surface runoff can be routed to both pervious areas (PAs) and drainage systems during an overland flow. Excessive runoff from an impervious area (IA) flowing into a drainage system causes an overload, which can be relieved by diverting runoff to PAs. However, the hydrological link between IAs and PAs, especially the runoff response to variation in overland flow areas (OFAs), has not well been considered in runoff simulations due to the complexity of routing. To understand how the OFA within an IA contributes to runoff generation, a novel classification approach was applied to categorize the IA in a study area in Nanjing University, Xianlin campus, China into directly connected impervious area (DCIA) and indirectly connected impervious area (ICIA) by flow routes using high-resolution ground-based images (0.5 m) from an unmanned aerial vehicle. The OFAs then include DCIA and the total impervious area (TIA), which is the sum of DCIA and ICIA. The runoff simulations were supported by Storm Water Management Model (SWMM) calibrated with observed rainfall and runoff data. The resulting proportions of DCIA and ICIA of the study area were 34.13% and 10.99%, respectively. The spatial distributions of DCIA and ICIA are characterized by the subcatchment landscape heterogeneity resulting from vegetation, imperviousness, and slope. The observed runoff coefficients and peak flows were positively correlated with the percentages of DCIA and TIA. The runoff coefficient was significantly correlated with the DCIA in a light rainfall event of 18.4 mm (R2 =0.82) and with the TIA in a heavy rainfall event of 119 mm (R2 =0.92). Runoff generation is affected by both the characteristics of the rainfall event and the accompaning variations in OFAs. Results indicate that increasing of the flow connectivity from IAs to PAs and increasing the water retention capacity of PAs may be effective strategies for optimizing landscape patterns for stormwater management.