Desempenho de métodos empíricos na estimativa da evapotranspiração de referência: um estudo no Ceará

Abstract

The estimation of reference evapotranspiration (𝐸𝑇௢) is essential for water planning and agricultural resource management, especially in semi-arid regions such as Ceará, where water scarcity is a reality. This study aims to compare different empirical methods for calculating 𝐸𝑇௢ in municipalities of Ceará, using the Penman-Monteith method, parameterized by the Food and Agriculture Organization (FAO), as a standard. The comparison seeks to evaluate the performance of the methods and identify the most suitable approaches for each municipality, considering their specific climatic conditions. The comparative analysis was conducted using statistical metrics such as the correlation coefficients: Pearson’s (𝑟), Willmott’s concordance index (𝑑), and Camargo and Sentelhas' confidence index (𝑐). The value of the confidence index (𝑐) was taken as a performance measure to screen and select only the most promising methodologies. Calculations and graphs were developed using Microsoft Excel (version 2016), while the maps were generated in QGIS Desktop (version 3.28.0) using the Kernel density estimator. To synthesize the results, a performance matrix and a graph were created, considering the confidence index (𝑐). Based on this, the methods were ranked both in general terms, taking into account their mean values and standard deviations across the entire sample, and specifically, by analyzing the results for each municipality. Overall, the Blaney-Criddle, Linacre, and Penman methods showed the best performances. When analyzing the results for individual municipalities, the comparative analyses highlighted distinct performances, reinforcing the importance of understanding local climatic conditions and how each methodology incorporates the influence of climatic variables in the evapotranspiration estimation process. Factors such as relative humidity, solar radiation, and thermal variation directly influence the obtained values, which should be considered when selecting an 𝐸𝑇௢ estimation method. Taking specific climatic conditions into account, the best results were obtained with Priestley & Taylor in the humid municipalities and Linacre in the dry municipalities.