Abstract:
This report is aimed at derivation of a spatially distributed unit hydrograph for Temur
watershed at railway bridge no. 293. The method for distributed unit hydrograph
computation allows for spatial non-uniformity of excess rainfall. Consequently, it is
based on the time-area method (Clark, 1945) derived using GIS. The GIS allows
development of a watershed's channel network for calculation of realistic travel times, it handles the distributed excess rainfall in calculating local surface runoff rates as inputs for channel flow and it compiles the time-area diagram from which distributed unit hydrograph is derived.
Simulation results shows that the errors between observed and simulated peak discharge are from -5.8% to 22.1% which is well within acceptable limits for designing small structures on this stream. However, the errors between observed and simulated time to
peak is large compared to errors in peak discharge simulation. Graphical comparison of observed and simulated discharge shows that the rising limb of computed hydrograph match well with rising limb of observed hydrograph in all cases, however, the recession shape of observed and simulated hydrograph do not match very well. This could be attributed to the fact that a very simple method for velocity simulation was used and no calibration of parameters has been performed. Also a pure translation model is used in this study which could be responsible for lack of agreement in the shape of hydrograph. As can be seen from topographic map of the catchment, the catchment is having mild slopes in majority of area and there is very likelihood of storage effects due to this mild slope. From this study it can be concluded that the method work well for simulation of peak discharge. However, for overall shape and time to peak discharge simulation, further refinements in the method are required. It is, therefore, appropriate to simulate this catchment with model, which take both translation and storage effects into consideration for further study.
