Coupling patterns of climatic variables in the formation of evaporation from open water bodies still are not clear due to the uncertainty involved in the process. The main goal of this study was the detection of such patterns via second-order sensitivity analysis. The Partial Deviations method, based on the Artificial Neural Network, was utilized to reveal coupling patterns. The new method was tested at two neighboring sites (Ahvaz and Isfahan) in Iran. We found that at Ahvaz station coupling between one day-lagged evaporation with air temperature and humidity with magnitudes of 26.37% and 25.21%, respectively had a major effect on the evaporation gradient. Similarly, the major effects on the evaporation rate at Isfahan station belonged to the coupling one day-lagged evaporation with air temperature and wind speed with magnitudes of 36.97% and 18.98%, respectively. The interaction patterns showed that the rate of evaporation reversed for both stations in the warm seasons of the year mainly because of an increase in atmospheric humidity. The climatic variables on their own domain (mostly their high values), aroused the effect of other variables, such as temperature, one day-lagged evaporation, wind speed and radiation which in interaction with other variables caused inverse the rate of evaporation in some cases. Even though adjacent climates have the most impact on each other, their coupling patterns are significantly different. Our study highlights the importance to include the reversal of the evaporation rate in modeling evaporation from open water bodies.