Abstract：Based on routine meteorological observations from 15 national stations in Xilingol League spanning from 1961 to 2013,this study employed linear trend estimation to analyze the spatiotemporal distribution characteristics of evaporation,and utilized correlation and regression analyses to establish multiple regression estimation models for each station.The results indicate that evaporation in the region generally exhibits a declining trend,with 12 out of the 15 stations showing a decrease,thus confirming the existence of the "evaporation paradox."Spatially,evaporation decreases from west to east,with the highest annual average in desert steppe,the lowest in meadow steppe,and typical steppe falling in between.Correlation analysis reveals that evaporation is positively correlated with annual mean diurnal temperature range,annual sunshine hours,and annual mean 2min wind speed,while negatively correlated with annual precipitation,annual mean vapor pressure,annual mean relative humidity,and annual mean atmospheric pressure.Furthermore,the established multiple regression models passed validation tests,with the overall average error between measured and estimated values remaining within 6％,demonstrating the feasibility of this research methodology.
2026, Vol. 43 
