Estimativa da evapotranspiração real em região semiárida, utilizando o algoritmo SAFER

Abstract

The effects of global climate change are diverse, such as increased pollution, deforestation, to water and food shortages. In this sense, it is essential to use techniques to estimate the amount of water needed for the development of agricultural crops. Therefore, the main objective of this work is using remote sensing techniques and the SAFER algorithm (Simple Algorithm for Evapotranspiration Retrieving) to estimate evapotranspiration for the central-north region of the state of Pernambuco and southwest of Paraíba. It is a semi-arid area, except for the city of Triunfo (PE), characterized as an altitude swamp. For estimation, was used surface data from the meteorological station of Serra Talhada (PE), belonging to the National Institute of Meteorology (INMET), as well as images from the Landsat8 satellite. From the observed data, the reference evapotranspiration was calculated using the Penman-Monteith (PM) method, necessary for the calculation of the Real SAFER evapotranspiration. Using the QGIS Software, mosaics were made with the images of the quadrants 216/065 and 216/066 and the clipping of the region of interest, with a radius of 50 km, starting from the meteorological station. After processing and analyzing the images, four dates were selected for the study, which are referenced in the text by the season of the year, to better understand their differences: 01/02/2014 (summer), 05/26/2014 (autumn), 07/18/2016 (winter) and 11/05/2015 (spring). The variables estimated by SAFER were analyzed: albedo, surface temperature and Normalized Difference Vegetation Index - NDVI. The results obtained for the albedo showed higher values for spring, the hottest and driest season analyzed, and lower values for autumn, the wettest period. These results are in agreement with those of NDVI, which are inversely related to albedo. Higher NDVI values were estimated in the rainy season, while lower indices were found for the driest season, for which the vegetation is less robust. They also corroborate the results for surface temperature (𝑇𝑆 ), which suggest higher temperatures during spring when compared to autumn. Comparing the estimates of the average 𝑇𝑆 of the study area with the data observed at the Serra Talhada station, it was verified that there was an overestimation of the variable, with a greater difference in spring, while for autumn the 𝑇𝑆 was underestimated. As for the estimation of real evapotranspiration by SAFER, it was verified that the average value was higher in autumn, period with higher precipitation index, and lower in spring, hot period and with lower water availability. Subsequently, an analysis was carried out for a smaller area, measuring 750 meters by 750 meters, with the meteorological station in the center, and the variables were analyzed to verify whether a more homogeneous area would present adverse results compared to the previous ones. The results showed slightly different estimates, however, within the standard deviation. Comparing the SAFER evapotranspiration estimates with the data collected at the station, the 𝐸𝑇𝑅 𝑆𝐴𝐹𝐸𝑅 and 𝐾𝐶𝑆𝐴𝐹𝐸𝑅 (SAFER evapotranspiration fraction) quotient was calculated for the original study area; quotient of 𝐸𝑇𝑅 𝑆𝐴𝐹𝐸𝑅 and 𝐾𝐶𝑆𝐴𝐹𝐸𝑅 for the smaller area; and 𝐸𝑇𝑅 𝑆𝐴𝐹𝐸𝑅 from the clipping and 𝐾𝐶𝑆𝐴𝐹𝐸𝑅 from the total area, verifying that the values of 𝐸𝑇𝑅 𝑆𝐴𝐹𝐸𝑅 are close to 𝐸𝑇0 𝑃𝑀, due to dependence from 𝐸𝑇0 𝑃𝑀 to estimate by SAFER. Greater differences occur when using the 𝐸𝑇𝑅 𝑆𝐴𝐹𝐸𝑅 value of the clipped area and the 𝐾𝐶𝑆𝐴𝐹𝐸𝑅 of the total area, with greater variations in autumn and spring, due to the dependence of surface temperature on the evapotranspiration estimate. It was concluded that SAFER, even though it was developed to estimate 𝐸𝑇 for crops, is efficient for estimating in heterogeneous areas, and can be used as a tool to aid irrigation management in different areas.