INFLUENCE OF THE ESTIMATED GLOBAL SOLAR RADIATION ON THE REFERENCE EVAPOTRANSPIRATION OBTAINED THROUGH THE PENMAN-MONTEITH FAO 56 METHOD

Abstract

Agriculture accounts for a significant portion of the state of Minas Gerais' GDP, which is one of the most productive Brazilian states in the agricultural sector. Given this scenario, it is essential to know the elements that affect agricultural productivity. The data manipulated in this work were obtained from the Network of Automatic Meteorological Stations (EMA's) from the National Institute of Meteorology (INMET) located in 51 cities from the State of Minas Gerais. There is a lack of studies to determine the most appropriate model to estimate evapotranspiration with less error in the state of Minas Gerais. Such studies are necessary because the state in question has different climatic and hydrological conditions. The objective of this paper was to evaluate whether the estimation of global solar radiation can influence the daily reference evapotranspiration by the Penman-Monteith FAO 56 method, for the state of Minas Gerais, Brazil. The estimated radiation data were applied to the Penman-Monteith FAO equation for reference evapotranspiration estimation, and these results were compared to the reference evapotranspiration data obtained by the Penman-Monteith FAO equation with radiation data measured at the stations. The performance indexes used were: Mean Relative Error (MRE), Root Mean Square Error (RMSE) and Willmott's d index. The solar radiation values used, when calibrated for the study site, tended to overestimate the evapotranspiration values, presented dispersion of 0.76 mm d-1 on average and the degree of precision of the calibrated models had an average of 0.65 mm d-1. The best solar radiation estimation model for the cities studied was the Thorton and Running model.