Temporal trend changes in reference evapotranspiration contrasting different land uses in southern Amazon basin

Abstract

Due to the expansion of agricultural borders, land use and land cover changes in the southern part of the Brazilian Amazon have been altering water and energy balances over the last decades. Still, Brazilian agriculture, for the most part, depends on rainfall to supply the water needs of crops and evapotranspiration is the main mean of water transferring, in the form of vapor, to the process of cloud formation and consequently, rains. Reference evapotranspiration (ETo) is a climatic variable with great potential to indicate climate changes, which can be used to detect temporal trends and calculate aridity index (AI). Thus, the general objective is to analyze the behavior of the reference evapotranspiration and the components used to estimate it in contrast to the land use in four locations in the North of Mato Grosso state, Brazil, in the southern Amazon basin. The main factors related to possible changes in ETo over a mean period of 20 years were evaluated, identifying trends for the data set studied (ETo, sunshine duration, air temperature, wind speed, relative humidity, and precipitation) and the correlation (r) between reference evapotranspiration and the other variables. Besides, the behavior of the AI was analyzed to observe the degree of rainfall insufficiency to maintain the soil in humidity conditions suitable for plant development. To identify how significant the trends were, we used the Mann-Kendall trend analysis and linear regression analysis. We found at least two different significant trends for each location. The reference evapotranspiration showed a significant trend only in the station located in the municipality of Canarana, with a negative trend, varying 8.4 mm per year. The aridity index, for the driest months of the year, presented values close to 1, characterizing the region as medium to high aridity. The variable that most influenced the ETo variability was the insolation.