Event-based flood estimation in un-gauged sub-basins: a comparative assessment of SCS-UH, CWC-UH and Nash-GIUH based rainfall-runoff models in Shilabati River, Eastern India

Estimating peak discharge (Qp) and design flood in small tributary sub-basins is challenging owing to limited observed streamflow data. To address this, the synthetic unit hydrograph (SUH) concept was introduced that helps to estimate Qp of direct surface runoff (DSRO) hydrograph by routing the excess rainfall to the basin's outlet, facilitating the construction of hydraulic structures in areas lacking observed rainfall-runoff data. Therefore, the present study evaluates the performances of three types of SUHs i.e., CWC-UH, Nash-GIUH and SCS-UH in estimating DSRO hydrographs, with an emphasis on Qp and time to peak (TP) during a storm event in ten tributary sub-basins of one of the most flood-affected river, Shilabati in Eastern India. The results of these three models exhibit striking similarities in the shapes of DSRO hydrographs derived from the SCS-UH and the Nash-GIUH models compared to the CWC-UH. The Nash-GIUH model stands out as the superior model due to the strong correlation (R2 = 0.86) between the ratio of Qp and Tp and the observed flood extents (flood-inundated area) for all sub-basins. In the Nash-GIUH-based DSRO hydrographs, Sub-basin-9 witnesses the highest Qp (334.64 m3s- 1) with short Tp (19 h) (Qp/Tp ratio = 17.61) followed by Sub-basin-10 (Qp/Tp ratio = 13.50). Similarly, 31.67% and 19.51% of the total areas of Sub-basin-9 and - 10, respectively, were affected by flood inundation in the past. Therefore, the association between the shape of hydrographs and flood extents depicts that the Nash-GIUH-based rainfall-runoff model can effectively estimate floods in areas lacking streamflow data.