Calibration of Empirical Equations to Estimate the Daily Reference Evapotranspiration and Comparison with ANNs Models under the Lack of Meteorological Data



Accurate estimation of reference evapotranspiration plays a key role in water resources management and irrigation scheduling. The FAO Penman-Monteith (F-P-M) is a reference method for calculation of evapotranspiration that needs temperature, humidity, wind speed and solar radiation. In many stations, it is not an easy task to measure the whole weather variables. Thus, in this situation, one must use some of the equations that require less climatic data. ANNs are an effective tools to model nonlinear phenomena with an interaction relationship. In this study, Seven ANNs with different input combinations have been implemented and compared with ten locally calibrated empirical and semi-empirical ETo equations (with estimated solar radiation and humidity parameters as inputs) in Gorgan synoptic station. The comparisons have been based on the basis of statistical error techniques, using F-P-M daily ETo values as a reference. ANNs have obtained better results than the locally calibrated ETo equations in the three groups: temperature and/or relative humidity based methods ANN 3model and HS equation with RMSE equal to 0.422 and 0.551mm d-1, respectively; solar radiation-based methods ANN 4 model and T equation with RMSE equal to 0.372 and 0.518 mm d-1, respectively and methods based on similar requirements to those of PM56 except for the estimation of solar radiation and/ or relative humidity ANN 6 and F-P-MRHest equation with RMSE equal to 0.208 and 0.310 mm d-1, respectively.