Avaliação da irrigação por fusão de imagens de satélite usando computação em nuvem: um estudo na região semiárida do Brasil.

Abstract

Irrigation evaluation in arid and semiarid regions is imperative to ensure the sustainable use of limited water resources and to guarantee food production. This study aimed to evaluate the consumptive use of water for irrigation in a pilot irrigated area of the Petrolina/Juazeiro, by fusing orbital and multispectral images (Landsat series and MODIS sensors) and few surface meteorological elements. To this end, cloud computing was used to process the orbital images and surface meteorological data to determine evapotranspiration (ET), using SEBAL, as input for the soil water balance. In the analysis of the best set of equations to be used in SEBAL, two models to calculate downwelling shortwave radiation (Rsw), ten clear sky atmospheric emissivity models to calculate downwelling longwave radiation (Rlw) and two models to determine the daily net radiation (Rn,24) were evaluated The accuracy of each model was evaluated with radiation measurements obtained from research quality sensors installed in micrometeorological towers. The best performances were found for the Allen model, Duarte model, and De Bruin model for Rsw, Rlw, and Rn,24, respectively. The accuracy of using SEBAL and STARFM models in estimating key variables and the net radiation itself and in determining ET on daily and monthly scales was evaluated. Image fusion indicated RMSE of Rn, Ts, NDVI and albedo equal to 17.00 W m-2, 2.28 K, 0.07 and 0.01, respectively. Daily (ET24) and monthly (ETm) ET as a result of the combination of STARFM with SEBAL, showed low mean errors (PBIAS = -2.75% and RMSE = 0.97 mm d-1 and 16.66 mm month-1) and high coefficient of determination (0.87 for ET24 and 0.91 for ETm) compared to using SEBAL only on Landsat images (PBIAS = -5.25%, RMSE = 0.97 mm d-1 and 17.66 mm month-1, r² = 0.92). The addition of fused images caused a better fit of the estimated cumulative ET24 curve relative to measured ET24. The water balance indicated that the cultivated sugarcane suffered water stress in the middle and at the end of the growing season, a circumstance that was better represented by estimates using the ET24 curve with the addition of fused images, than using Landsat images alone. Although this increase in the temporal resolution of the estimated ET24 data indicated a higher water consumption, it reports a sufficient quantity to meet the water demand of the crops. Certainly, the need for systematic monitoring of water use for irrigation in semiarid regions, particularly in the spatial and temporal context, is not only evident, but imperative. In this context, ET24 modeling with orbital data is one of the techniques that are intended to assist the range of information required for proper irrigation management, and that can contribute to ensure better planning of public policies for water rights and use, with a view to achieving a more efficient use of this resource.