ESTIMATIVA DO 𝐊CB COM DADOS DE NDVI: AVALIAÇÃO DE EQUAÇÕES EM CULTIVO DE MELOEIRO IRRIGADO NO SEMIÁRIDO BRASILEIRO

Abstract

One of the greatest challenges facing agriculture today is to provide increased productivity coupled with the preservation and efficient management of the scarce water resources available. The increase in water consumption observed in the last decades, due to the increase of the world population, imposes the implementation of actions that aim at the efficient use of water, mainly in areas on agricultural crops. Melon cropping in the Brazilian semi-arid region is accomplished through the use of plastic coverings in the litter and polypropylene blanket aiming to reduce the insect attack and a higher quality of the fruit for export, so the loss of water to the environment occurs by transpiration. The use of remote sensing techniques allows the planning of irrigation by obtaining and processing images generated by satellites to determine the basal coefficient of cultivation that represents the water consumption by the culture derived from the sweating. In the present work, the objective was to evaluate equations for the determination of the basal coefficient of cultivation, using NDVI data for the irrigated melon in the Brazilian semi-arid region. Three cloud free images (09/01/2009, 09/17/2009 and 10/03/2009) generated by the Landsat 5 TM satellite, orbit 216 and point 63, were obtained and processed. These images were used to calculate the Normilized Diference Vegetation Index (NDVI) and determine the basal crop coefficient for each equation analyzed in the study. The Bowen ratio - energy balance method was used to validate the equations and the Penman-Monteith equation provided the reference evapotranspiration. The highest values of NDVI and Kcb were found in the last phenological stage of the melon, characterizing the phase of maximum vegetative development of the crop. The evaluated equations underestimated the basal evapotranspiration of melon at 4.16%, 5.2%, 7.5% and one of them was overestimated at 9.69%, respectively, proving the efficiency and generality of the use of spectral models for determination of the water needs of the species in question.