Validation of the Hydrological Processes in a Hydrological Model

Abstract

Hydrological models are often required to model watersheds where the conditions change over time. Calibration and validation of these models is a difficult process that requires validation of each of the major hydrological processes within the model. This paper presents the calibration, validation, and sensitivity analysis of the WATFLOOD hydrological model. The calibration process is usually based on streamflow and may involve an implicit validation of the hydrological processes when the internal state variables are monitored to ensure that the model operates realistically. This paper presents explicit validations of several internal state variables (soil moisture, evaporation, snow accumulation and snowmelt, and groundwater flow) and the statistical characteristics of the streamflow. The WATFLOOD model is shown to track each of these variables with sufficient accuracy for operational use of the model. In addition, several behavioral sensitivity checks are presented to show that the model behaves in a realistic manner. This paper provides a broadly based methodology for calibration and validation of a distributed hydrological model.