TR(BR)-2/99-2000 : A modified SCS-CN-based hydrologic model

Abstract

Long-term hydrologic simulation studies provide a useful and important input to water resources planning and watershed management practices. The Soil Conservation Service (SCS, 1956) Curve Number (SCS-CN) method is a widely used event based rainfall-runoff method. In this report, the SCS-CN method is used for simulating daily rainfall-runoff data of two catchments, viz., Hemavati (area=600 sq. km) and Ramganga (area=3134 sq. km) catchments falling in sub-humid regions of India. The model formulation is based on the conversion of precipitation to rainfall excess using SCS-CN method and its routing by single linear reservoir and linear regression techniques. The base flow that is assumed to be a fraction of the infiltration amount is routed using the lag and route method. The variation of SCS-CN parameter potential maximum retention S is governed by evapotranspiration and antecedent moisture condition. This model when applied to Hemavati has shown efficiency of 75.31% and 82.03% in calibration and validation, respectively, and when applied to Ramganga, these are 58.34 and 67.20% respectively in calibration and validation. The stability of the computed parameters is examined by reversing the data sets of calibration and validation and it is found that a greater length of data will be required to stabilise the model parameters. The application of single linear reservoir technique to Hemavati data is found to conserve the mass successfully in Hemavati application whereas the application of linear regression fails to conserve the mass of Ramganga runoff The computed values of the initial abstractions, rainfall excess, infiltration, and base flow are also presented in tabular form along with their annual and seasonal statistics. Initial abstraction values are found to be much higher in non-monsoon season than those in monsoon season.