The Predictability Of Eurasian Snow In The NCEP Climate Forecast System Version 2 Reforecasts

Eurasian snow cover fraction(SCF) and snow water equivalent(SWE) prediction and predictability are analyzed using the recently developed NCEP Climate Forecast System, version 2(CFSv2) monthly retrospective forecasts for 1983-2010. The main conclusions are presented as follows:The CFSv2 is capable of representing the seasonal cycle of Eurasian SCF and SWE. Generally, the prediction skill for the Eurasian SCF is higher than SWE and the prediction during snowmelt period is better than snow-accumulation period.This study focuses on the prediction skill and predictability of Eurasian SCF in snowmelt and snow-accumulation periods since the intensive variability of SCF occurs in the two periods. The CFSv2 reasonably predicts the interannual variations, long-term trend and leading pattern of Eurasian SCF in snowmelt period several months ahead. In comparison with the snowmelt season, the CFSv2 shows a better prediction skill in climatological values but a worse skill in the interannual variability in snow-accumulation period. Additionally, the observed SCF in the snow-accumulation period exhibits increasing tendency while the CFSv2 does not successfully simulate it but shows opposite linear trend. Analysis shows that the biases of Eurasian SCF in the snowmelt and snow-accumulation periods are significantly related with those of temperature and precipitation in the CFSv2. The forecasted cooler and wetter atmosphere is suggestive of the overestimation of the mean SCF. Meanwhile, the underestimation in the variability of both temperature and precipitation in the CFSv2 may be the important factor for the underestimated variability of SCF, especially for the damped variability of SCF in the snow-accumulation period. Generally, the CFSv2 shows a higher and more stable prediction skill after late-1990 s than before in the two periods. The change in the initial condition in the CFSv2 and the observed SCF in late-1990 s might be the plausible reason for it.The CFSv2 is capable of representing the climatological distribution of observed Eurasian spring SWE. However, owing to the smaller snowmelt rate and late snow-ablation in the model, it produces more SWE in spring. The correlation analysis shows significant relationship between the model forecasts ahead of one to five months and the observation. That means the CFSv2 could generally predict the interannual variations of Eurasian spring SWE ahead of one to five months while shows poor skill for longer lead month. But the CFSv2 reforecasts ahead of one to five months also displays overlarge downtrend and variability compared with the observation. Results show this phenomenon is probably caused by the strong downward trend and intensive variability of their initial conditions in the CFSR since there exists rather high correlation between the model results and their corresponding initial conditions. Further, the signal–noise ratio analysis demonstrates that the CFSv2 has predictability in the Eurasian spring SWE about one to five months in advance.