降水集中度的变化特征及影响因素分析——以山西为例

An analysis of precipitation concentration variation characteristics and influential factors in Shanxi province, China

基于山西14个站点1957—2014年逐日降水资料,计算年尺度与多年尺度下的降水集中指数(CI)和极端降水指数(R_(95p)和R_(99p)),利用统计方法,研究了山西降水集中度的时空变化和影响因素。结果表明,CI值与R_(99p)显著正相关,CI值增大则发生极端强降水几率增大。多年尺度CI值体现出明显的纬度地带性和较大的局地空间差异。受温带大陆性季风气候控制,山西省降水CI值介于0.59~0.64,相对亚热带地区较小,且变化范围较小。总体上年尺度CI值呈下降趋势,五台山、右玉、五寨、运城等站点下降趋势显著。盆地区域较高山高原区的CI值更大且下降趋势不显著,更易于发生极端降水事件。较高的高程和较大的变幅可增强高程对CI值的影响。太平洋年代际振荡(PDO)与年尺度CI值显著负相关。PDO冷位相时期,西太平洋副高西进、偏强,CI值偏高,发生极端强降水的可能性增大。本研究揭示了自然地理条件和海-气相互作用对区域尺度降水集中度存在复杂的影响。

Shanxi province is located on the border of monsoon in the north of China, which is across multiple latitudes. The mountains and basins in the study area are staggered. Therefore, the influential factors of precipitation in the study area are complicated. Based on the daily precipitation data of 14 rain gauging stations in Shanxi from 1957 to 2014, the precipitation concentration index （CI） and extreme precipitation index （R95p, R99p） were calculated at annual scale and multi-year scale. Mann-Kendall method was utilized to analyze the spatio-temporal variation of precipitation concentration, and its influential factors. The results showed that the positive correlation between R99p and CI at annual scale is significant, which means high precipitation concentration will increase the possibility of extreme precipitation. Precipitation concentration shows obvious differences in latitudinal zonality and local space at multi-year scale. Due to a temperate continental monsoon type, the precipitation concentration degree is relatively low and the change range is relatively narrow （0.59-0.64）. In general, the concentration of precipitation is decreasing at inter annual scale, and the descending trend of Wutaishan, Youyu, Wuzhai and Yuncheng is signifcant at the 0.05 signifcance level. The possibility of extreme precipitation in the central basin area is greater than that in the plateau high-mountain area due to the higher CI value and insignifcant declined trends detected. The correlation between multi-year scale precipitation CI and elevation, latitude, longitude, mean annual precipitation and mean annual rainy days is not signifcant. The higher elevation and larger amplitude make the infuence of elevation on the concentration of precipitation at annual scale increase. The correlation between precipitation CI and NAO, ENSO air-sea interaction and EASMI is not signifcant. However, there is a signifcant negative correlation between annual CI and PDO. The possibility of extreme precipitation is greater when PDO is negative with the Western Pacifc Subtropical High westward and stronger. The results demonstrated geographic conditions and air-sea interactions exert a complex infuence on precipitation concentration at a regional scale.