Abstract:Based on the prediction results of Beijing Climate Center-Climate Prediction System version 3-Subseasonal to Seasonal version 2 (BCC-CPSv3-S2Sv2), various evaluation and test methods are used to test the prediction effect of the model in the flood season across the Yangtze River Basin, and to evaluate the prediction skills of subseasonal daily/dekad precipitation in the flood season in the Yangtze River Basin. The model error characteristics and the available prediction lead time of model precipitation are analyzed. The results show that the model systematically overestimates the precipitation in flood season in the Yangtze River Basin as a whole, and its prediction skill in the middle and lower reaches of the Yangtze River is higher than that in the upper reaches of the Yangtze River. The dekad prediction skill of the model is improved with the initial time approaching, but the BS predicted 5 d ahead is better than that predicted 1 d ahead. The effective prediction time of the model for the daily quantitative prediction in the flood season in the Yangtze River Basin is about ten days, and its qualitative prediction of the precipitation anomaly in the flood season is similar. The skill of model prediction 1 dekad ahead is obviously higher than that 2 dekads ahead. The analysis results of the probabilistic prediction of precipitation anomalies during the flood season show that the prediction 2 to 3 dekads ahead also has some reference values. In addition, the prediction skill of model under less rain scenario is better than that under excessive rain scenario. There is still much room for improving the prediction capability for the moderate rainfall and above.