Parameterizing unit hydrographs (UH) to account for rainfall derived inflow and infiltration (RDII) for multiple land uses under tropical climate

Rainfall-derived inflow and infiltration (RDII) is the flow entering a drainage system during and sometimes, long after, cessation of rainfall. RDII can be an important source of sanitary backup into basements and Sanitary Sewerage Overflows (SSO) into receiving water bodies. Accounting for RDII, however, is not always straightforward and is subject to a number of uncertainties, including topographic conditions, land use land cover (LULC), soil types, etc. The hydrology component of PCSWMM uses a triangular Unit Hydrograph  (UH) approach to account for RDII flows where the UH shape is described by three parameters: R (fraction  of  rainfall volume that  becomes  inflow/infiltration, %), T (time from the onset of rainfall to the peak of unit hydrograph, hrs), and K (the  ratio of the  time  to  recession of the  unit  hydrograph to  the time to peak). Literature parameterizing R, T, and K, is limited yet predominantly associated with areas with temperate climate conditions, e.g., North America. Little has been made known about how to estimate these parameters in tropical climate conditions. As such, this study examined the R T K characteristics for different sewersheds in Singapore. In total, we instrumented and modelled 12 sub-catchments, representing different LULC, e.g. residential, institutional, cemetery, farmland, forests and mixed urban during the period 2014 to 2017. Catchment properties, e.g., pervious percentages, Manning’s N values, slopes were obtained from literature review, other studies of ours, and actual field measurements across the country. The main calibrating process primarily employs UH where the R² between the simulated and observed data regarding event totals and maximums range from 0.701 and 0.934. We observed that UH parameters vary temporally and spatially. For instance, the K values distinguish forest areas being the most delayed from other sites. Time wise, R values are the highest and the most varied in July and August whereas T values are highest in August, October, and November. Different biophysical conditions, such as soil types, slopes, can explain the geographical variability in calibrating R, T, K. The temporal variation of these three parameters relates to the double monsoon climatic regime of Singapore. The values of the UH parameters derived from this research could constitute practical estimation guidelines for engineers and modellers that are involved in complex RDII problems.