统计预报海温场驱动的CAM3.1模式预报试验

基于动力气候模式进行月一季尺度预报的“两步法”思想，提出一种新的预报海温场统计模型，并以该统计模型预报的海温场驱动NCAR CAM3.1模式对1981-2000年月时间尺度的东亚500 hPa高度距平场进行客观回报试验；在此基础上，提出了对预报结果的订正方法。结果表明:统计预报海温模型的预报海温场能够反映出全球海温空间分布的基本特征，并对表征ENSO事件的Ni?o3.4区海温变化的预报能力较强。该统计模型预报的海温场驱动的CAM3.1模式可以较好地预报出东亚500 hPa环流的主要分布特征，试验表明:适当的统计订正方法可以在一定程度上提高CAM3.1模式对东亚夏季500 hPa环流背景的预报技巧。

Forecasting Experiments of CAM 3.1 Model Using Statistic Forecast SST Data

Based on the ＂two-step method＂ in month/season forecast, a new statistic model is designed to predict the global SST field. The global SST during 1981-2000 are predicted using the model. Comparing with the NCEP SST field, the results indicate that the forecast global SST field from the statistic forecast SST model could reflect primary patterns of the global SST, and can predict the variability of SST anomaly in Ni?o3.4 which represents E1 Ni?o/La Ni?a event. The correlation coefficient between forecast SST and NCEP SST in Ni?o3.4 is 0.596. Furthermore, the numerical experiments based on the forecast global SST boundary conditions for forecasting the monthly anomalies of 500 hPa geopotential height over east-Asia (1981-2000) are performed using the NCAR CAM3.1 model. The first six eigenvectors which represent typical spatial patterns of CAM3.1 500 hPa and NCEP 500 hPa are decomposed from the CAM3.1 500 hPa and NCEP 500 hPa fields using EOF respectively. The results show that primary patterns of 500 hPa height anomalies in east Asia can be forecasted by driving CAM3.1 with the forecast global SST. However, the forecast ability of 500 hPa geopotential height over east Asia in summer is faulty and unstable in forecast experiments. A statistic revision method is provided to revise the forecast results of CAM3.1 model. The root mean square error and correlation coefficients of 500 hPa height anomalies between CAM3.1 forecast (revised before and after) and NCEP in summer over the east Asia are compared, the root mean square error between the two fields decreases every year, and the mean correlation coefficient during 1981-2000 increases. These results indicate that the forecasting skill of 500 hPa height anomalies over east Asia in summer is improved using the statistic revision method to a certain extent. The experiments show this statistical method can be used to improve the forecasting skill of the dynamic model CAM3.1.