Cálculo da evaporação em reservatório tropical por diversos métodos: o caso do açude Gavião

Abstract

The evaporation process is seen by the water management sector as a "loss" that negatively impacts the region's dense network of dams, reducing the water supply in an area susceptible to drought. Studies that make it possible to obtain accurate evaporation estimates in reservoirs represent an important action for the adequate management of water resources. The objective was, therefore, to estimate evaporation in the tropical reservoir Gavião, Ceará, Brazil, between 11 October and 11 December 2019, using several methods: (i) Penman, using three stations located in different positions; (ii) water balance; and (iii) high precision sensor, which measures the difference in water level based on pressure measurements. It is also intended to verify if there is (and what) effect of the location of the weather station on the estimation of evaporation by Penman. We used meteorological data collected from a floating meteorological station installed on the reservoir lake, a station installed in a terrestrial environment, about 400 m from the water mirror of the reservoir and a station at the Pici Campus in Fortaleza, 19 km from the water mirror of the reservoir. The results of two conventional methods of quantification of evaporation were compared with the values obtained from the pressure sensor, the most accurate. Penman's method based on the data from the embarked station proved to be the most favorable in the daily and monthly scale, presenting a better adjustment. Differences in the average evaporation by these methods varied between 1 and 7 mm/day. The water balance method presented the highest evaporation rates, overestimated due to the imprecision of its calculation and the high uncertainty in the measurements of its terms. When comparing the evaporation estimates by Penman's method, it was observed that the stations around the reservoir overestimate the evaporation, probably because the stations recorded higher solar radiation. This may have contributed to the overestimation of evaporation, since the Penman method showed a strong sensitivity to variation in solar radiation. The evaporation estimated by meteorological stations with different proximity to the water mirror differ statistically from that obtained by the station on the water mirror. Conclusively, we can state, based on the data collected in this research, that the two methods widely used in the Brazilian semiarid region, which are, (i) that of the water balance with the hypothesis of negligible infiltration; and (ii) that of Penman based on data from stations located on land and distant from the lake are not reliable, especially for the daily time scale.