img

官方微信

高级检索

中国沙漠 ›› 2007, Vol. 27 ›› Issue (1): 130-136.

• 天气与气候 • 上一篇    下一篇

我国北方沙尘暴与强沙尘暴过程的分形特征及趋势预测

尹晓惠1, 王式功2   

  1. 1.北京市气象局气象台, 北京 100089; 2.兰州大学大气科学学院, 甘肃 兰州 730000
  • 收稿日期:2005-08-08 修回日期:2005-10-25 出版日期:2007-01-20 发布日期:2007-01-20

Fractal Characteristics and Trend Forecast of Dust-Storms and Severe-Dust-Storms in Northern China

YIN Xiao-hui1, WANG Shi-gong2
  

  1. 1.Beijing Meteorological Bureau, Beijing 100089, China; 2.School of Atmospheric Sciences, Lanzhou University, Lanzhou 730000, China
  • Received:2005-08-08 Revised:2005-10-25 Online:2007-01-20 Published:2007-01-20

摘要: 沙尘暴是一种常见的灾害性天气现象。由于其爆发的频繁性和危害的严重性,沙尘暴趋势预测的研究已经成为了国内外专家学者关注的热点和焦点。在回顾相关研究进展的基础上,借助于分形理论在处理复杂系统方面的独特切入点,利用1954-2002年中国北方720个气象站的沙尘暴观测资料和681个气象站的强沙尘暴观测资料,分析了近49 a来,我国北方沙尘暴及强沙尘暴事件的时序分形特征,并运用R/S分析方法对其变化趋势进行科学预测。研究发现,我国北方地区沙尘暴及强沙尘暴事件具有近似一致的变化态势,即自20世纪50年代以来呈逐渐减少,但近期又相对增多的变化趋势。1954-2002年沙尘暴及强沙尘暴事件的时间序列具有分形特征,其关联维数分别是2\^64和3\^34;平均可预测时间尺度分别为7~8 a和8~9 a。Hurst指数能够较好的表征我国北方沙尘暴及强沙尘暴事件的发生规律。根据Hurst指数及其变化趋势可以预测,2002年以后的7~8 a内,我国北方沙尘暴事件将具有先增加后减少的趋势;1999年以后的8~9 a内,我国北方强沙尘暴事件也将具有先增加后减少的趋势。

关键词: 沙尘暴, 分形理论, 相空间重构, 可预报时间尺度, R/S分析

Abstract:

Dust-storm is a kind of common harmful weather phenomenon. A lot of domestic and international experts and scholar have paid close attention to the prediction of them due to the high frequency and severe calamity of dust-storms. This paper first reviewed the international progress of relevant research, and then analyzed the fractal characteristics of dust-storms and severe-dust-storms in recent 49 years in Northern China according to the observed data from 1954 to 2002. Finally, we used the R/S analytical method to predict the variation tendency of them. The results showed that the dust-storm and severe-dust-storm events have an approximate unanimous variation tendency; namely, the occurrence of dust-storm and severe-dust-storm reduced gradually since 1950s but increased relatively in recent years. The temporal sequences of dust-storms and severe-dust-storms in Northern China have fractal characteristics. The correlative dimensions are 2\^64 and 3\^34, and the average predictable periods are 7~8 years and 8~9 years respectively. The Hurst indexes can reflect the rules of dust-storms and severe- dust-storms in Northern China very well. According to the indexes and the variation tendency of dust-storm and severe-dust- storm, we can predict that the occurrence of dust-storm in Northern China will increase first and reduce next from 2002 to 2010 and the occurrence of severe-dust-storms will have a similar trend from 1999 to 2008.

Key words: dust-storm, fractal theory, phase space reconstruction, forecasting time scale, R/S analysis

中图分类号: