A hybrid calibration method for improving hydrological systems using ground-based and remotely-sensed observations

Confidence in the simulation of hydrological models to accurately represent the inherent variability of a watershed is dependent on proper calibration. Calibration of hydrological models using streamflow and actual evapotranspiration (AET) data, which the latter has a remarkable spatial and temporal distribution in the watershed scale, gives more confidence on the partitioning of water between soil storage and aquifer recharge. Therefore, this study aims to improve hydrological response of the agricultural-based Neishaboor watershed, located at the northeast of Iran, using the Soil Water Assessment Tool (SWAT). To this end, monthly streamflow data at three hydrometric stations (1997–2010) and monthly AET values, computed by SEBAL (Surface Energy Balance Algorithm for Land) algorithm using MODIS products (2001−2010), were employed for the SWAT hybrid calibration and validation. The results showed that RCHRG_DP, SOL_AWC, ESCO, EPCO and CANMX are the most sensitive parameters to AET estimation. The proposed hybrid calibration method had reasonable performance on AET estimation in which NS and RMSE values varied between 0.7–0.88 and 7.89–9.96 mm.month⁻¹ in calibration and 0.86–0.93 and 6.79–7.91 mm.month⁻¹ in validation periods, respectively. In comparison to SEBAL-based AET, a general under-prediction tendency was observed by the SWAT-based AET with the proposed method. Highest error rate of the SWAT model for the calibration (average RMSE = 17.18) and validation (average RMSE = 9.44) periods was obtained in the range of measured AET values which were higher than 50 mm.month⁻¹. In the watershed scale, the calibrated SWAT model with the proposed hybrid calibration method provides efficient tool for hydrological monitoring, especially in arid and semi-arid regions.