Forecasting daily potential evapotranspiration using machine learning and limited climatic data

▶ Forecasting ET o models for water managing purposes (1985 Hargreaves ET o). ▶ 2 ET o forecasting approaches using air temperature and Multivariate Relevance Vector Machine. ▶ Ind. Approach provides better and larger forecast lags than Dir. Approach. ▶ Better ET o forecast when using forecasted weather variables rather than ET o values. ▶ In study case, ET o is forecasted of up to 4 days using the Ind. Approach. Anticipating, or forecasting near-term irrigation demands is a requirement for improved management of conveyance and delivery systems. The most important component of a forecasting regime for irrigation is a simple, yet reliable, approach for forecasting crop water demands, which in this paper is represented by the reference or potential evapotranspiration (ET o). In most cases, weather data in the area is limited to a reduced number of variables measured, therefore current or future ET o estimation is restricted. This paper summarizes the results of testing of two proposed forecasting ET o schemes under the mentioned conditions. The first or “direct” approach involved forecasting ET o using historically computed ET o values. The second or “indirect” approach involved forecasting the required weather parameters for the ET o calculation based on historical data and then computing ET o. An statistical machine learning algorithm, the Multivariate Relevance Vector Machine (MVRVM) is applied to both of the forecastings schemes. The general ET o model used is the 1985 Hargreaves Equation which requires only minimum and maximum daily air temperatures and is thus well suited to regions lacking more comprehensive climatic data. The utility and practicality of the forecasting methodology is demonstrated with an application to an irrigation project in Central Utah. To determine the advantage and suitability of the applied algorithm, another learning machine, the Multilayer Perceptron (MLP), is used for comparison purposes. The robustness and stability of the proposed schemes are tested by the application of the bootstrap analysis.