Effect of potential evapotranspiration estimates on effective parameters and performance of the MIKE SHE-code applied to a medium-size catchment

The effect of the method of calculating a time series of crop potential evapotranspiration (ET_p) on both the model performance and the magnitude of the main effective parameters of a distributed hydrological model was assessed on the basis of an independent multi-calibration. The ET_p was estimated as a function of the reference evapotranspiration (ET_o) by means of a crop coefficient approach. Three methods for deriving ET_o estimates were used in modelling a medium size catchment in Belgium: (A) Food and Agriculture Organisation (FAO) FAO-24 standard approach; (B) FAO-24 approach but used with coefficients, for both the wind and the Stefan-Boltzmann equations, differing from the standard formulation; and (C) FAO-56 standard approach. For assessing the performance of the estimation approaches, the point scale ET_o outputs of the generation methods were compared to local point scale ET_o guidelines derived from previous research. The outputs of generation method B and of the generation method C showed a close agreement with the local point scale ET_o guidelines. Their effect (to a catchment scale) on the performance of the hydrological model also seemed to be comparable. The best model performance was obtained by using the higher ET_o data generated by method A. The research also revealed a significant dependency of some of the effective parameters on the different ET_p estimates; especially of those parameters related to the computation of the actual evapotranspiration (ET_act).