Evapotranspiration is an important component and key link of river basin water cycles and plant hydrological processes, and is a core issue in global climate change research. It is not only an important way to understand the energy and water consumption of permafrost regions, but also is an important channel to master the water cycle and energy balance in cold regions. In this paper, multiple linear regression analysis method and weighted comprehensive analysis of major factors method were used to investigate the variation characteristics and impact factors of evapotranspiration in the Genhe River Basin. The results showed the following: (1) The monthly average evapotranspiration in the Genhe River Basin during the freezing-thawing periods in 1980-2017 was 28.29 mm. Compared with the freezing-thawing periods, the total evapotranspiration in the growing seasons was much higher than that in the freezing-thawing periods, with monthly average evapotranspiration of 67.71 mm; (2) The main factors affecting evapotranspiration in the Genhe River Basin were precipitation and temperature. During the freezing-thawing periods, the variation in evapotranspiration in May was mainly determined by temperature. In the growing season, precipitation was the main factors affecting evapotranspiration in June. This will lay a foundation for clarifying the relationship between permafrost-climate change-hydrologic cycle in the permafrost active layer during the land surface process, so as to provide some basic data and important scientific basis for the comprehensive study of the hydrologic process and its impact on climate, ecology, water resources and environment in the permafrost area.

This study investigates the response mechanisms between soil water-heat transfer and environmental factors during freeze-thaw periods and establishes soil water-heat transfer functions in a cold region. Based on field-measured values of soil temperature and liquid-phase water content collected at an automatic weather station in the black soil area of the Songnen plain, the influence of the cumulative negative temperature on the soil freezing depth was analyzed under different snow cover conditions. A gray correlation analysis method was used to screen the environmental factors and determine those with the most influence on changes in soil water-heat transfer processes. Then, soil water-heat transfer functions were established between the selected environmental factors and soil temperature, the liquid-phase soil water content. The results showed that during the freezing and thawing period, snow cover hindered the effects of the cumulative temperature on the thickness of the frozen soil layer. Additionally, the time of occurrence of the maximum freezing depth under natural snow (NS), compacted snow (CS) and thickened snow (TS) treatments was delayed 7, 12 and 20 days, respectively, compared with that under bare land (BL). The correlation between atmospheric temperature, total radiation and soil temperature was relatively high, and this effect decreased with the increasing of snow cover. The main driving factors of variations in the liquid-phase water content were ambient humidity and saturated vapor pressure, and the effects of these factors decreased with increasing soil depth and snow cover thickness, similarly. In the active frozen layer, the correlation coefficients of the soil water-heat transfer functions were relatively high, and the function model can be tested by the significance (P < 0.05) test. However, the R-2 values of functions below the active layer were relatively low, and the soil water-heat transfer in the area below the active layer was less affected by the environment. This study reveals the characteristics of energy transfer and mass transfer in a composite system of atmospheric factors and frozen soil under snow cover conditions. It provides a reference for accurate forecasting and the efficient utilization of soil water and heat resources in cold and arid regions.