摘要
无标度性广泛存在于自然界系统包括气候系统中,其特征之一是可观测量存在幂函数关系,它揭示了气候系统的复杂性。为探索气候可预测性的客观基础,介绍了去趋势涨落分析(DFA)方法,并应用其对北京2003年逐日观测资料序列进行了分析。结果表明,北京日平均气温和降水量均可划分为多个标度不变区域。在特定的标度域内,它们都表现出正长程相关的性质,为制作这些量的短期气候预测提供了理论基础。
In order to investigate the observational basis of the climate predictability, with the detrended fluctuation analysis （DFA） the long-term variations of the temperature and the precipitation in Beijing have been investigated. Non-scaling exists widely in the nature system, including the climate system; one of its features is a power function existing in an observable quantity, which represents the complex of the climate system. Based on the DFA, the calculations for daily temperatures and precipitations during 2003 were conducted, and it is found that there are several ranges where their scaling is unchangeable; in the specific ranges, the time series of the temperature and the precipitation exhibit a positive long -term correlation feature, that provides a theoretical basis for climate predictions.
出处
《气象科学》
CSCD
北大核心
2008年第4期421-425,共5页
Journal of the Meteorological Sciences
基金
北京自然科学基金(8042012)
空军司令部气象局"精细预报系统"课题
关键词
北京
气候变化
无标度性
去趋势涨落分析
Beijing Climate change Scaling uncorrelation Detrended fluctuation analysis
