Unit Hydrograph Modeling using Geomorphological Instantaneous Unit Hydrograph (GIUH) Method

Flood forecasting at Wonogiri Reservoir is restricted on the availability of hydrologic data due to limited monitoring gauges. This issue triggers study of unit hydrograph modeling using Geomorphological Instantaneous Unit Hydrograph (GIUH) which is based on Geographic Information System (GIS). Analysis of physical watershed parameters was conducted on Digital Elevation Model (DEM) data using software Watershed Modeling System (WMS) 10.1 and ArcGIS. Nash model and S-curve method were used to process triangular GIUH into hourly Instantaneous Unit Hydrograph (IUH) and Unit Hydrograph (UH) and then was compared with the observed UH of Collins method. A sensitivity analysis was conducted on parameter of RL and Nash-model k. Evaluation of accuracy of the simulated GIUH runoff hydrograph was also conducted. The GIUH model generated UH with smaller peak discharge Qp, also slower and longer of tp and tb values than the observed UH. Accuracy test of the simulated GIUH runoff hydrograph using Nash-Sutcliffe Efficiency (NSE) shows that Keduang watershed gives a satisfying result, while Wiroko watershed gives less satisfactory result. The inaccuracies occur due to limited flood events used to derive the observed UH and stream tributaries that were not properly modeled based on Strahler method.

Development of geomorphological instantaneous unit hydrograph (GIUH) model for a new un-gauged watershed

A geomorphological instantaneous unit hydrograph (GIUH) model was developed for a watershed of Damodar valley corporation, Hazaribagh, using Nash (1959) and Itrube (1982) methods to compute peak discharge (qpeak) and time to peak (tpeak). The model was calibrated and validated for five storm events, i.e. June 24-25 (1992), October 12-13 (1993), November 2-3 (1993), June 28 (1994) and August 6 (1996) by comparing their ordinates with the ordinates of instantaneous unit hydrograph (IUH). The GIUH was tested with absolute prediction errors (APE) of the ordinate of peak discharge. On comparison, it was found that, most of the GIUH models overestimated the runoff at initial stage, while underestimated at the latter stage in comparison to the IUHs, which was mainly due to consideration of const ant value of Ф-index, for computation of effective rainfall. The absolute prediction errors (APE) were computed to be 5.97, 18.09, 23.32, 9.64 and 7.52% of the ordinates of peak discharge for the storm events of June 24-25 (1992), October 12-13 (1993), November 2-3 (1993), June 28 (1994) and August 6 (1996) respectively.