| Soil moisture-atmosphere interaction is an important driving factor of climate,where land evaporation plays an important role.Soil moisture supplies water vapour to the atmosphere through land evaporation.The accuracy of land evaporation estimation is of great significance since it is an intermediate variable in which soil moisture affects air temperature and precipitation.In this study,reliability ensemble averaging(REA)and signal-to-noise ratio optimization(SNR-Opt)methods are used to generate a set of long-series daily land evaporation datasets with a spatial resolution of 0.25 degree in historical period and three sets of monthly datasets with a spatial resolution of 1 degree under multiple future climate change scenarios,which provide support for subsequent research.Since transpiration is the main component of land evaporation,and the changes of vegetation will affect land evaporation significantly,the response structures of land evaporation to vegetation greening under historical and future climate change scenarios have been explored with ridge regression and sensitivity analysis.The Dual Extended Kalman Filter-based generalized Partial Directed Coherent(g OPDC)was first applied to the study of soil moisture-atmosphere interaction,which realized the quantitative expression of two-way influences of soil moisture-air temperature and soil moisture-precipitation in the time-frequency domain under historical and future climate change scenarios.Finally,the sites have been used as a case to further explain the dynamic mechanism of the interaction between soil moisture and atmosphere.The causal structure characteristics of key processes in soil moisture-atmosphere interaction in different climate regions have been quantitatively described with information flow process network.Furthermore,the controlling factors of land-atmosphere coupling have been analyzed.The role of land evaporation in land-atmosphere coupling has been revealed through disaster cases.The main conclusions are as follows:(1)The quality of the merged REA land evaporation product has been significantly improved,and the merged projected datasets in multiple scenarios have reduced the uncertainty greatly.(2)Vegetation is always the dominant factor of land evaporation variation both in historical period and under multiple climate change scenarios.However,the sensitivity of land evaporation to vegetation variation decreases gradually with the increase of vegetation change rate and radiative forcing levels.(3)There are seasonal differences in the influence of soil moisture on air temperature and precipitation in China during historical periods.The influence of soil moisture on air temperature reaches the strongest in spring,while in summer on precipitation.Both of them reach the strongest in the eastern humid and semi-arid and semi-humid regions where land evaporation is relatively strong.In recent years,the influence on air temperature has weakened in the eastern humid region,and has gradually increased on precipitation.The influence strength is closely related to land evaporation,which increases first and then decreases with the increase of land evaporation.(4)Under the background of global warming,the influence of soil moisture on air temperature and precipitation changes.The influence strength of soil moisture on air temperature changed more dramatically under higher radiative forcing levels.The influence on precipitation mainly showed an increasing trend,with the increasing trend aggravated under higher radiative forcing levels.(5)The causal structure of the key processes in soil moisture-atmosphere interaction can be well expressed by information flow.There are significant differences in causal structures of different climate regions and seasons.The interaction between variables in semi-arid and semi-humid regions is more complex,and is the most complex in summer and the simplest in winter with the longer time lag.Water is the main controlling factor of land-atmosphere coupling in arid Inner Mongolia Site,and energy is the main controlling factor in other regions.The change of evaporation fraction will lead to or strengthen the disaster,which reveals the role of land evaporation in land-atmosphere coupling. |