Evaluation on summer temperature and precipitation predictions in Huai River Basin by CFSv2 model

Based on CFSv2 model reforested data supplied by the National Climate Center and observed temperature and precipitation data at 172 stations in the Huai River Basin, the skills of model prediction for summer temperature and precipitation are evaluated.Results show that simulated climatological mean temperature is similar to observation, while simulated precipitation is clearly less than observation mainly due to the underestimation in July and August, despite of the similar northern dry and southern wet spatial distribution.The negative bias is also found in root mean square deviation.From the view of frequency distribution, reforested temperature is similar to observation, but there are great differences for precipitation both in a negative bias and a low frequency of extreme events.Analysis of ROC further indicates that the skill of predicting high-temperature events is better than that of predicting low-temperature events, while the skill of predicting abnormal precipitation depends on leading months.The main spatial structures are well displayed in the main modes.The warming trend is well simulated but with a positive bias in warming speed.There is some skill in predicting annual variability of the main precipitation EOF mode, though the decadal change cannot be embodied.The results from this article can serve as a reference to climate prediction and model interpretation and application.