Operacionalização do balanço de energia e evapotranspiração em escala regional com dados de sensores orbitais.

Abstract

Basically, evapotranspiration (ET) is obtained in a timely manner or place in some experiments from data measured on the surface, which does not allow for spatial visualization on a large scale. Furthermore, models are quite complex and require data from meteorological stations, often scarce. In this sense, we intend to employ a simplified methodology and adjusted from existing data from satellite sensors available for the Northeast region of Brazil in order to operationalize obtaining the BE and evapotranspiration on a large scale only with remote sensing algorithms. MODIS/Terra data were used in the period 2002-2011 for implementation of the proposed methodology and evaluation of results. The data resulted in obtaining time series data and maps of NDVI, albedo, land surface temperature and the components of BE and evapotranspiration. The results were analyzed for various targets scattered throughout the Northeast region, selected based on the vegetation map of Brazil's IBGE 2002, and compared with literature data. The targets of the region showed average NDVI values in the range 0.35 to 0.70; albedo in the range 11-23% and Ts between 24 and 34 ºC. The assessment values of daily net radiation presented mean percentage error (MPE) and mean absolute error (MAE) of 9% and 15 Wm-2, respectively, when compared with field experiment data. The estimation of ET values according to eddy covariance method in a field experiment on the Frutacor farm presented overestimated data when compared to the simplified model, showing MPE and MAE equal to 22% and 0.79 mm.dia-1, respectively. The mean annual ET values varied from 482 to 1105 mm.yr-1 in areas of savanna and rainforest, respectively. When compared to MOD16 product data was observed that there was an overestimation by the latter, ranging from 596 to 1532 mm.yr-1 for the same areas. The results obtained are consistent with the literature and showed that the proposed methodology is a powerful tool to estimate the components of the BE and ET.