Development of a Cell-Based Parsimonious Catchment Model

One of the crucial problems in conceptual rainfall runoff models is over-parameterization. In most cases, as the number of parameters to be calibrated increases, model performance either does not improve or, it may even decrease due to poorly defined parameters. Simple daily-based rainfall-runoff models are generally lumped models that average catchment heterogeneity. To increase the spatial resolution of such models, the computational resolution may be changed from catchment scale to a cell-based scale. This change may require a modification of the model structure by addition of new parameters. Such an approach may lead to two problems: increase in the complexity of model structure and over-parameterization. In the study presented, a cell-based, parsimonious, distributed, continuous, conceptual daily rainfall-runoff model (DRRSM) with a minimum amount of parameters is introduced. The data needs of the developed model include only daily rainfall, pan evaporation, DEM (digital elevation map), land cover distribution and soil properties which can be obtained easily from various sources through the Internet. The model has proved to be successful in cases where available data on catchment characteristics are insufficient to meet the calibration needs of over-parameterized complex models.