Evaluation of Land–Atmosphere Coupling Over Eastern China in the Two Versions of FGOALS

This study describes the performances of version two of the grid-point (FGOALS-g2) and spectral (FGOALS-s2) Flexible Global Ocean–Atmosphere–Land System Models in simulating land–atmosphere coupling over eastern China. The evaluation focuses on the soil moisture feedbacks on spring and summer mean surface air temperature occur 1979–2004. The Climatic Prediction Center (CPC) dataset shows that significant negative feedbacks on spring mean surface air temperature occur mainly over many areas of northeastern China and the ecological and climatic transition zone of northern China (transition zone). Negative feedbacks on summer mean surface air temperature occur mainly over the transition zone and North China with many grid cells achieving the 90 % significance level. Positive feedbacks also occur over some areas, but much fewer grid cells achieving the 90 % significance level. The soil moisture feedbacks over eastern China from the CPC dataset generally agree with those from the Global Land Data Assimilation System (GLDAS) dataset. In general, both FGOALS-g2 and FGOALS-s2 models capture relatively well the negative soil moisture feedback features over many areas of northern China but overestimate the amplitude, except for FGOALS-g2 in spring. Both models have biases in simulating positive soil moisture feedbacks over some areas. For the mean percentage of the variances of surface air temperature owing to soil moisture feedbacks, both models overestimate the negative percentage amplitude, except for FGOALS-g2 in spring. In general, FGOALS-g2 matches the CPC dataset better than FGOALS-s2 does regarding soil moisture feedbacks on surface air temperature over eastern China.