CS(AR)-3/99-2000 : Asymptotic estimation of runoff curve number for small hard rock watersheds from rainfall-runoff data

Abstract

One of the major activities in applied hydrology is the estimation of storm event runoff from ungauged small watersheds. Such estimates are often required in the design of small hydraulic structures and formulation of watershed management strategies. USDA, Soil Conservation Service (SCS) Curve Number method is a well accepted tool in Hydrology for the estimation of design floods for small hydraulic structures and for other rainfall-runoff analyses. This is a simple, predictable, and stable conceptual method. The curve number (CN) is a measure of retention by a given combination of soil and vegetation and varies from 0 (no runoff) to 100 (all rainfall becomes runoff). Curve number values for different combinations of soil, land use and treatment classes are given in SCS National Engineering Handbook (SCS, 1985). These tables were derived from the analyses of data from small experimental watersheds in USA. So it is preferable to have an alternate method to estimate curve number of the selected catchment. Method suggested by Hawkins (1993), where an average curve number for a watershed is estimated from observed rainfall-runoff data set, used in the present study.

Observed rainfall-runoff event data has been used to estimate the average curve number for three small hard rock catchments, namely, Melumalai, Barchi, and Dandavathi. Natural and frequency matched data sets were used for the analysis. Sensitivity of computed curve number, with rainfall and runoff values, has been tested by forcing 10% error in data set. In order to check the applicability of this method, curve number was estimated for Barchi watershed by the conventional SCS method using the soil and land use data.