The characteristics of water cycle in arid region of China under the climatic background of warm and wet

As the 4th IPCC assessment report concluded, the arid region which located in the northwest part of China has the largest warming extent both in the past 50 years and under different future climate scenarios. This observational fact can lead to the great change and response of water cycle in temporal and spatial distribution in this sensitive region to global warming. In this study, we first analyze the different water cycle variables over arid region of China (ARC) during the past 60+ years (1948~2010) based upon the latest monthly dataset of Global Land Data Assimilation System (GLDAS-2) which driven by Noah land surface model 3.3 (Noah LSM) with 1 degree horizontal resolution. The results show that during the past 60+ years over ARC, there is a significant increase of precipitation and evapotranspiration. The increment of annual precipitation exceeds 4.0 mm/decade, while the increment of annual evapotranspiration approaches 4.0 mm/decade. Both precipitation and evapotranspiration increase largest increments in summer (over 2.0 mm/decade). Corresponding to the wetter background, the runoff depth increases over most part of ARC, especially in the north part and center part of Taklimakan Desert. The soil moisture in 4 depth layers divided in Noah LSM also shows the significant increasing trend in the past 60+ years. In order to investigate the mechanism of water cycle change in ARC, an offline numeric experiment by Common Land surface Model 3.5 (CLM 3.5) is taken after the data analysis based on two different meteorological forcing datasets (The observation and Princeton datasets) in finer spatial resolution. The results of soil moisture trend agree well with the GLDAS based analysis. Moreover, the trend of simulated water table depth shows the complex variations both in temporal and spatial patterns, and this change can make great influence on stream flow, especially the subsurface stream flow simulation. Finally, a basin-scale hydrological model is introducing in this study for the stream flow simulation in two typical basins in ARC, the results indicate that the stream flow can be significantly impacted by the warm and wet background in ARC.