摘要:用1979/1980～2017/2018年冬季逐日长江下游气温资料研究长江下游冬季低温日数与温度低频振荡的联系。结果表明，冬季长江下游逐日气温存在较显著的季节内振荡周期（15～25 d、25～40 d和50～70 d振荡），其中与12～2月低温日数关系最密切的是25～40 d振荡。基于2001～2018年逐日长江下游实时气温的25～40 d低频分量和东亚地区850 hPa低频温度主成分，建立了冬季长江下游气温低频分量的延伸期预测的时变扩展复数自回归模型（ECAR）。其中，采用基于T-EOF（temporal empirical orthogonal functions）延拓的实时奇异谱（SSA，singular spectrum analysis）滤波，很好地抑制经典SSA滤波的边界效应，得到较稳定的实时低频振荡信号。对2001/2002～2017/2018年12～2月长江下游温度低频分量进行独立的实时延伸期预报试验的结果表明，这种数据驱动的简化的复数预测模型对25～40 d时间尺度的长江下游冬季低频温度的预测时效可达26 d左右，预报能力显著优于经典自回归模型（AR），能为提前3～4周预报长江下游地区冬季持续低温过程提供有价值的预测背景信息。
关键词:长江下游低频温度/
实时奇异谱分析/
延伸期预报模型/
低温天气/
冬季
Abstract:Basing from observational data, this study analyzed the variations in ISO (intraseasonal oscillation) of the daily temperatures and its relationships to the low temperature in December–February of 1979/1980–2017/2018 over the lower reaches of the Yangtze River (LYR). The daily temperatures over the LYR in December–February are mainly of periodic oscillations of 15–25 d, 25–40 d, and 50–70 d, and the interannual variation in the intensity of its 25–40-day oscillation has a strong positive correlation with the number of low-temperature days in December–February. The real-time low-frequency components of daily temperature in the LYR and the principal components of 850-hPa low-frequency temperature in eastern Asia from 2001 to 2018 were used to establish the time-varying extended complex autoregressive (ECAR) model on an extended-range forecast of the 25–40-day low-frequency temperature over the LYR in winter. Real-time SSA (singular spectrum analysis) filtering with the T-EOF (temporal empirical orthogonal functions) extension can effectively inhibit the end effects of the traditional SSA and enhance the real-time signal of ISO. A 17-year independent real-time extended-range forecast was conducted on the extended-range forecast of the low-frequency component of the temperature over the LYR in December–February of 2001/2002–2017/2018. Experimental results show that the data-driven ECAR model has a good forecast skill at the lead time of approximately 26 days, and its forecast ability is superior to that of the traditional autoregressive model. Hence, the development and variation of the leading 25–40-day modes for the low-frequency temperatures at 850 hPa in eastern Asia and temporal evolutions of their relationships to the low-frequency components of the temperature over the LYR in winter supplied the valuable predicting background to determine the extended-range weather process in the persistent low temperature over the LYR at the 3–4-week leads.
Key words:Low frequency temperature over the lower reaches of Yangtze River/
Real-time singular spectrum analysis/
Extended-range forecasting model/
Low temperature weather/
Winter
PDF全文下载地址:
http://www.iapjournals.ac.cn/dqkx/article/exportPdf?id=7a55b9db-6464-4cf7-b8e7-6d4053609d93
