Estimativa da evapotranspiração a partir de dados diários de previsão meteorológica

Abstract

The real time determination and future of evapotranspiration (ET) can contribute to more efficient management of irrigation. The objective of this study was: (i) to evaluate the accuracy of the estimate of the reference evapotranspiration (ETo) calculated on daily data of the weather forecast for seven future days (EToest), to different locations, growing seasons and periods with and without and rain; (Ii) evaluate the impact of using the EToest in the water balance of a soil cultivated with corn. The estimate ETo was performed to Santa Maria - RS and Crystalline and Morrinhos - GO. The collection of weather forecast data was carried out between June 26, 2011 to March 31 2014 was collected-daily (6:00 AM), the site: http://www.tempoagora.com.br/, the forecast data maximum temperature (Tmax) and minimum (Tmin) air, maximum relative humidity (RHmax) and minimum (RHmin) and wind speed (U2) for the next seven days. The EToest calculation was carried out by FAO-Penman-Monteith (FAO PM) and Hargreaves methods. The ETo observed day (EToday) was calculated by the FAO- PM method with data measured by automatic weather station of the National Institute of Meteorology (INMET). The values of EToest (est+1 the est+7) were compared with the values observed daily (ETodia). The accuracy of the estimated EToest was evaluated using statistical indicators (coefficient of determination - R2, slope - b, mean square error of the square root -RMSE and average error - ME). For the analysis of the water balance of a soil cultivated with corn, a field experiment was accomplished in the year 2011/12, in an area of UFSM. The experimental design was a completely randomized factorial design with four replications. The factor "A" consisted of two methods of calculation to estimate crop evapotranspiration (ETc): (i) using crop coefficient (Kc) and simple; (Ii) using dual Kc. The "D" factor consisted of four irrigation strategies: (i) 100% replacement of ETc; (ii) 75% replacement of ETc; (iii) 50% of ETc and replacement; (iv) 25% of replacement etc. The volumetric water content and available soil were simulated is compared with the measured (100% tartamento simple Kc). The results are an overestimation of EToest (>0,04 and <0,72 mm day-1) for the three sites, attributed to overestimation of Tmax by the weather forecast, a variable that showed high association with EToest (R2 0,60). This overestimation of EToest is reduced as the estimation approximates the observed daily. The differences between the ETodia and EToest, has small inlfuência the results of the soil water balance (RMSE less than 10 mm for the available water). Thus, EToest can be an alternative for determining the need for crop irrigation in future time, contributing to the management of irrigated activity and for the most efficient use of natural resources and other agricultural.