Evaporation in the semi-arid region: sensitivity analysis and comparison of estimation methods

Abstract

Knowledge of evaporation rates is crucial for planning the use of water resources. However, its quantification requires diverse meteorological data and can be influenced differently by these factors. This study aimed to: i) analyze the sensitivity of evaporation obtained by the Class “A” Tank (TCA) and estimated by the Penman method (1948) to negative influences; and ii) investigate the correlations between the TCA evaporation values and the evaporation estimated by the methods of Linacre (1993) and Kohler (1955) in relation to those by the method of Penman (1948). Daily data obtained for a period of thirty years was used, through a meteorological station located in Juazeiro, Bahia. The E values obtained by TCA and the Penman method were correlated with the indirect variables through regression analyses. Furthermore, E values were determined using alternative methods: Linacre (1993) and Kohler (1955), which were correlated with the values obtained by the Penman method (1948), considered standard in this study. The proximity of the values was evaluated using Pearson's representation coefficient (r), the determination coefficient (r²) and Willmott's agreement index(d). The results demonstrated that the evaporation values measured by TCA and estimated by Penman's research (1948) were sensitive to vapor pressure deficiency. The evaporation obtained by the Class A Tank and estimated by the Linacre method (1993) was closer to the result obtained by the Penman model (1948). Suggest that analyzes be performed to identify the suitability of different methods on varying time scales.