Effects Of Air Temperature And Precipitation On Global Soil Moisture Variability

Soil moisture is widely recognized as a key parameter in environmental processes because of its important role in controlling the water and energy balance between the land surface and atmosphere.It has a significant impact on evapotranspiration rates,soil thermal parameters,surface albedo,and Bowen ratio and thus has an effect on the planetary boundary layer regime,clouds,and precipitation.The variability of soil moisture,especially soil moisture drying,has an important relationship with heat waves,dust outbreaks and plant productivity and survival.Therefore,studying soil moisture and its variability is crucial to understanding climate change and land-atmosphere interactions.In this study,the variability of soil moisture according to climatic region was analyzed;the quantitative contributions of air temperature and precipitation to soil moisture change,especially soil moisture drying,were highlighted;and uncertainties of soil moisture in historical simulations and future projections were emphasized.The main conclusions are summarized as follows:(1)The data from the Global Land Data Assimilation System(GLDAS)could efficiently describe the distribution and variation of soil moisture over China by comparisons with observation.We analyzed global variability of soil moisture for the period 1948–2010 using the GLDAS data.The results showed that soil moisture was dominated by negative trends,with pronounced drying over North Africa and East Asia.The contributions of soil moisture drying over the two regions to global drying were more than 60%.Spatial analysis according to climatic region revealed that the most obvious drying occurred over transitional regions between dry and wet climates.The noticeable drying first took place in the humid transitional regions and extended to the dry transitional regions,beginning in the 1980 s.Overall,a clear decreasing trend occurred over a period of 63 years for China,with pronounced drying over northeast China and north China,and the drying rate was higher than the global average trend.The long-term variations of soil moisture in different layers over the semi-arid region of Loess Plateau showed a decreasing trend with a clear interannual change.(2)The relative contributions of air temperature and precipitation to soil moisture were studied by regression analysis and the land surface model.Statistical analyses showed that variability of soil moisture was notably related to the changes in precipitation and temperature,but with different roles.For the global average,precipitation had a dominant effect on the variability of soil moisture at interannual to decadal time scales,but temperature was the main cause of the long-term trend of soil moisture on the whole.The enhanced soil drying in the transitional regions was primarily caused by global warming,which is illustrated by regression analysis and the land surface model.The qualitative analysis and quantitative contributions illustrated that soil drying over China was driven primarily by decreasing precipitation and was enhanced almost twofold by increasing temperatures.As soil moisture decreases,the positive feedback between soil moisture and temperature may result in future water shortages.Following the Representative Concentration Pathways 8.5(RCP8.5)and 4.5(RCP4.5)simulation scenarios of Coupled Model Intercomparison Project phase 5(CMIP5),the model-predicted soil moisture demonstrated a continuously decreasing trend during the 21 st century.Over semi-arid region of Loess Plateau,there was a negative correlation between soil moisture and air temperature in all layers,and with the increase of soil depth the correlation enhanced.On the contrary,a positive correlation could be observed between precipitation and soil moisture and the largest correlation coefficient appeared at the surface layer.(3)Uncertainties of soil moisture in historical simulations(1920–2005)and future projections(2006–2080)were investigated by using the outputs from the CMIP5 and Community Earth System Model.The results showed that soil moisture climatology varies greatly among models despite the good agreement between the ensemble mean of simulated soil moisture and the GLDAS data.The uncertainties caused by initial conditions and model structure showed similar spatial patterns and magnitudes,with high uncertainties in dry regions and low uncertainties in wet regions.In addition,the long-term variability of uncertainty caused by model structure rapidly decreased before 1980 and increased thereafter,but the uncertainty caused by initial conditions showed an upward trend over the entire time span.The model structure and initial conditions can cause uncertainties at all time scales.Despite these large uncertainties,almost all of the simulations showed significant decreasing linear trends in soil moisture for the 21 st century,especially in the Mediterranean region,northeast and southwest South America,southern Africa,and southwestern USA.