Calibration of the Monthly Time Scale Runoff Model

Abstract

The stream runoff model developed by Mohseni and Stefan (1996) has a monthly time scale and is based on the water budget theory. Its function is to make mean monthly runoff projections under different climate scenarios. The model uses 6 climate variables, 11 watershed and soil parameters, and 3 parameters related to both climate and runoff. Some of the parameters are measurable and, therefore, obtainable as model input. The model lumps all watershed and soil parameters both vertically and horizontally. A nonsystematic calibration procedure gives different results, depending on the initial values chosen for some of the calibration parameters. The calibration parameters of the model are related to the two processes which are the most difficult to quantity and where the most information is required: direct runoff and snowmelt runoff. A systematic calibration procedure has been added to the original model to avoid inconsistencies in the results. The systematic calibration procedure is selected for the direct runoff parameters. For the snowmelt runoff, only some modifications in input are implemented. Base flow algorithm also required some changes in estimating the hydraulic conductivity of the storage below the root zone in order to better fit the water budget theory and Darcy's Law. For testing, the modified model is applied to two watersheds in two different climate regions, one in northern Minnesota and one in southwestern Oklahoma.