Obtenção da evapotranspiração real diária através da aplicação de técnicas de sensoriamento remoto no semiárido brasileiro.

Abstract

The loss ofwatervaporfrom the surface tothe atmosphereis known as evapotranspiration, which releases largevolumes of waterby evaporationfromsoiland vegetationbytranspiration.Estimatingwater usein irrigated areasisimportant forgood planning, management andrational useof this resource, as well as to improve production. For this studywill be processedsatellite imagesofQuixeré-CE, with interestat the FarmFrutacorwhere there arepredominance ofbanana's growing. Therefore, the mainobjective of this studywas to evaluatethe accuracy andoperational of thealgorithmS-SEBI andSSEBin relation toSEBALfor estimating thedailyactual evapotranspiration(ETa) in asemiarid regionof northeastern Brazil; containingareas ofbanana orchard, caatingaandexposed soil. The modelsS-SEBI andSSEBexhibitedstrong correlation, r>0.93, with a significance of5% and more than86% ofvariance explained. It was verifiedthat theSSEBunderestimated theETain all analyzes, andin generalthe S-SEBI overestimated. TheS-SEBI exhibitedless than12% errorsin the orchardandcaatinga, theSSEBexhibitederrorsgreater than22%, however the soil exposedtoboth modelsshowed largediscrepancies in theSEBALwitherror greater than36%. Therefore, among the two algorithmscomparedwithSEBAL, the S-SEBI showed best performance to estimate ETa with minor discrepancies