Discussion of 'Forecasting Weekly Evapotranspiration with ARIMA and Artificial Neural Network Models' by G. Landeras, A. Ortiz-Barredo, and J. J. López

Abstract

[EN] The discussers wish to express their appreciation to the authors for their efforts in the field of evapotranspiration forecasting. The development of prediction models based on the evolution of the output variable throughout the time is a powerful and interesting procedure to deal with important planning, design, and management tasks of irrigation resource systems. Moreover, this approach has hardly been tackled in the evapotranspiration ANN community, which mainly focuses on real-time evapotranspiration estimation. Nevertheless, there are some methodological aspects regarding the generation of the ETo targets for the implementation of the ARIMA and ANN models that need to be discussed, basically the time series management and the local calibration of the Hargreaves-Samani equation