杨秋明. 2014. 基于20—30d振荡的长江下游地区夏季低频降水延伸期预报方法研究[J]. 气象学报, 72(3):494-507, doi:10.11676/qxxb2014.028
基于20—30d振荡的长江下游地区夏季低频降水延伸期预报方法研究
Study of the method of the extended-range forecast for the low frequency rainfall over the lower reaches of the Yangtze River in summer based on the 20-30 day oscillation
投稿时间:2013-07-18  修订日期:2013-12-27
DOI:10.11676/qxxb2014.028
中文关键词:  MLR/PC-CAR混合预报模型  长江下游  20—30 d低频降水  预测
英文关键词:A hybrid forecasting model of MLR/PC-CAR  The lower reaches of the Yangtze River (LYRV)  Low frequency rainfall on the time scale of 20-30 days  Prediction
基金项目:国家自然科学基金项目(41175082)。
作者单位
杨秋明 江苏省气象科学研究所, 南京, 210009 
摘要点击次数: 3508
全文下载次数: 2551
中文摘要:
      用长江下游降水低频分量和环流低频主成分,构造多变量时滞回归模型(MLR)和主成分复数自回归模型(PC-CAR)的混合预报模型(MLR/PC-CAR),对长江下游降水低频分量进行延伸期逐日变化预报,延长预报时效。通过2011年6—8月预测试验表明,20—30 d时间尺度的长江下游低频降水预测时效可达50 d左右,采用南半球中高纬度地区850 hPa 低频经向风的主成分作为预测因子的模型的预测精度明显高于东亚地区低频经向风作为预测因子的模型。这表明在20—30 d时间尺度上,长江下游降水与南半球中纬度绕球遥相关(SCGT)型有关的主分量的时滞相关更加密切。进一步对于较强20—30 d振荡的多年资料构建的MLR/PC-CAR混合模型预测试验表明,SCGT是预测夏季长江下游低频降水未来50 d变化的显著信号。基于SCGT的发展和演变,对于把握类似长江下游地区2011 年6月初旱涝急转和7月中旬持续降水和强降水过程异常变化过程很有帮助,SCGT可以作为夏季长江下游20—30 d低频降水和强降水过程进行延伸期预报的主要可预报性来源之一。
英文摘要:
      Low-frequency rainfall over the lower reaches of the Yangtze River (valley) (LYRV) and the principal component of the circulation are adopted to establish a hybrid forecasting model with the multivariable lagged regressive model (MLR) and the principal component-complex autoregressive model (PC-CAR) combined, called MLR/PC-CAR model, which is applied to the daily forecasting of low frequency rainfall over LYRV for the extended range with the forecast period of validity prolonged. By many forecast experiments in June-August of 2011, this forecasting model has good predictive skill up to 50 days for the 20-30 day rainfalls over LYRV. And the predicted low frequency rainfalls over LYRV with the predictor of the principal component of 850 hPa meridional wind anomalies over the middle and high latitudes of the Southern Hemisphere is more accurate than that over East Asia, suggesting that, on the time scale of 20-30 days, the rainfall over LYRV are more tied to the principal components associated with the SCGT for the lag time. Moreover, the forecasting experiments for many years with the stronger 20-30 day oscillation also show that the SCGT is a key signal for the prediction of the low frequency rainfall in LYRV over the next 50 days with this hybrid forecasting model of MLR/PC-CAR. Based on the development and evolution of the SCGT, it will help us to hold the process of the anomaly change of a sharp turn from drought to flood in early June and the lasting heavy rainfall in mid-July of 2011 over LYRV. Hence, the variability of the SCGT is one of the main sources of the predictability for the extended range forecast of the 20-30 day rainfall and severe rainfall over LYRV in summer.
HTML   查看全文   查看/发表评论  下载PDF阅读器
分享按钮