论文部分内容阅读
On multi time-scale of intraseasonal,interannual and interdecadal,soil moisture is the dominant land surface parameter affecting the global atmosphere.As there is a limit for climate prediction,interannual variability of seasonal mean fields arising from variability within the season is essentially unpredictable on seasonal,or longer time scales.A variance decomposition method could estimate the signal of intraseasonal (unpredictable) and predictable (slow) component,respectively.This methodology provides a way to better identify and understand the source of predictable skill as well as the sources of uncertainty in climate variability.In this study,the Community Land Model (CLM3.5) was used to simulate regional soil moisture in China during 1958-2010 with the atmospheric forcing field constructed from station observations.Based on the variance decomposition method,the seasonal potential predictability of soil moisture in China mainland was given,the spatial patterns of predictable and unpredictable component were identified,the predictors for seasonal soil moisture and their possible mechanisms were discussed.Furthermore,seasonally predictable and unpredictable modes of soil moisture simulated by Coupled Model Intercomparison Project (CMIP5) were assessed,and projected changes of soil moisture were given.