河西走廊沙尘暴及其影响因子的多尺度研究

Multi-scales Variation of Sand-dust Storms in Gansu Corridor

沙尘暴活跃期具有周期性，这与影响因子的周期变化有关，尝试利用具有多时间尺度和多分辨率特性的小波分析技术对其进行研究。利用Morlet小波对甘肃河西走廊18个气象站多年平均的沙尘暴日数进行分析，发现沙尘暴变化的时间序列具有多尺度振荡的特点，存在7a、13a、以及23a左右的特征时间尺度（周期变化）。通过对大风日数、干旱指数、降水、气温等沙尘暴的影响因子分析，表明沙尘暴的活跃度与大气环流的年际变化和突变有关，特别是与大风日数在时间尺度和主周期等方面具有明显的一致性，7a左右的年际变化是它们的主周期；干旱指数、降水和气温具有同步的时间序列，13～25a的年代际变化是它们的主周期，因此，它们对沙尘暴的影响主要体现在大时间尺度上。

Wavelet Analysis is a developing technology in the field of time-frequency analysis, which has the features of multi-time scales, multi-levels and multi-resolutions. Sand-dust storms periodically change in different space and time scales, and it is difficult to accurately simulate it by means of the traditional methods. Using the method of wavelet analysis, this paper discusses the series data of sand-dust storms in recent 30 years and the influencing variables including gales, drought, precipitation and air temperature. The wavelet analysis shows that sand-dust storms and climatic variables alternate obviously in multi-time scales. Time series of all the climatic variables are in-phase, but their main periods are different. Periodic oscillations of 7 years for sand-dust storms and gales dominate over other scales, correspondingly 1325 years periodic oscillations of drought indices and precipitation and temperature variations are in the dominance. According to the periodical trends of the sand-dust storms, the Gansu Corridor will receive the sand-dust storms above normal level as a whole, but the increasing capability will be quite finite. This paper reveals that variability of sand-dust storms are more complicated than drought and precipitation and temperature, and gales have the prominent controls on sand-dust storms than other climatic variables, which influence sand-dust storms mainly on the long-tern scales. The research also illustrates that wavelet analysis offers a novel method of exploring the trends of the sand-dust storms, and the method could potentially has the ability to forecast the climatic variations.