干旱区生态环境知识资源中心

The initial abstraction ratio (I(a)/S) in the original SCS-CN equation is generally assumed to be 0.20. However, mounting evidence indicates the I(a)/S ratio is not constant and is typically less than 0.20. The impact of this value on runoff prediction is considerable particularly in semi-arid climates where typical 6-month design storms are less than 2.54 cm of total effective precipitation. In addition, in cold climates, the presence of snow cover at the onset of rainfall influences the abstraction ratio and increases runoff volumes due to rapid snow melting. In this study, the validity of the currently used 0.20 value in semi-arid cold climate regions where most design storm runoff events are generated as a result of rain-on-snow events with frozen ground conditions was investigated. To accomplish this, a rainfall simulator system was developed to simulate rainfall on a 1.22 m wide x 2.44 m long plot. The programmable rainfall simulator produced rain intensities and durations according to the design hyetographs encountered in semi-arid cold regions. The artificial rainfall intensities were less than 3.5 mm/hr and durations were up to 30 hours. Runoff responses were simulated under different snow depths (2.5-10.0 cm), snow water equivalents (0.5-5.58 cm) and plot slopes (0-5%). These three parameters were used to propose an alternative methodology for assigning I(a)/S ratios. Data from 19 test runs were used to develop a multiple regression equation for application in Eastern Washington to demonstrate the utility of the methodology. The new methodology improves estimates of initial losses during rain-on-snow events.