Surface runoff estimation over heterogeneous canal commands applying medium resolution remote sensing data with the SCS-CN method

Abstract

The precise estimation of surface runoff from rainfall is critical for water resource management. In the recent past, remote sensing and Geographic Information System (GIS) technologies have been widely used in the estimation of surface runoff from watersheds, and from agricultural fields in particular. This is due to the inherent ability of remote sensing to capture spatial heterogeneity of surface parameters such as land use and land cover. This could lead to better performances of surface runoff simulation models. Surface runoff volume/rate estimation involves quantifying the amount of rainfall exceeding infiltration and initial abstractions which must be satisfied before the occurrence of runoff. The widely accepted SCS curve number method was employed to calculate surface runoff, using a combination of remotely-sensed land use/land cover and hydrometrological data in the Punjab canal command areas. Land use/Land cover maps for four cropping seasons, Rabi 2004-05, Kharif 2005, Rabi 2006-07 and Kharif 2007 were derived using red and near infrared bands of MODIS 8 day products. The existing soil map was reclassified into hydrological soil groups and rainfall data were interpolated using the inverse distance method to represent the spatial rainfall values of each canal command. The results show that CN values vary from 70 to 95 during the study period. The highest CN value of 94.4 is during the Kharif 2005 season. Meanwhile the runoff-coefficient is changing from 0.01 to 0.25 and 0.01 to 0.43, respectively, during Rabi 2004/05 and Rabi 2006/07. During Kharif 2005 and Kharif 2007, the runoff-coefficient varied from 0.01 to 43 and 0.01 to 0.45, respectively. The study shows that the SCS curve number method can be used for runoff estimation with the help of remote sensing products and GIS technologies from catchments where gauging data is not available.