基于气象—农业综合干旱指数的云南省干旱时空演变特征

Spatiotemporal evolution characteristics of drought in Yunnan Province based on meteorological-agricultural comprehensive drought index

摘要

【目的】为探寻契合云南省实际情况的干旱指数，客观地揭示云南省干旱变化规律，实现对干旱状况的有效监测。【方法】通过熵权法将气象干旱指数(标准化降水蒸散发指数)与农业干旱指数(标准化土壤湿度指数)进行结合，构建新型综合干旱指数。借助相关分析、均方根误差以及标准偏差等方法，对不同干旱指数进行评估与筛选，确定最契合云南省实际状况的干旱指数。并基于优选出的干旱指数，运用游程理论、Mann-Kendall趋势检验等方法，深入剖析云南省干旱的时空变化特征。【结果】综合干旱指数反映干旱情况优于其他干旱指数，与气象、农业干旱指数的相关性分别为64%、60%,与统计的旱情资料的吻合率高达94.11%,在监测2009—2010年冬春连旱事件时，该方法所呈现的干旱发展态势与实际发生过程高度契合；时间上，近20年，云南省月尺度与季尺度的COI值均呈现出波动下降的态势。表明，干旱发生的频次有所增加，严重程度亦趋向增强。季节差异上，春季的干旱状况最为严峻，冬季次之，夏季和秋季相对较轻。但夏季干旱正以0.034/a的速率持续增强；空间上，云南省干旱强度与严重度表现出以滇中区域为中心，向外围逐渐递减的态势。在春、冬季节，重度干旱与特重度干旱主要集中于滇西南及滇中区域；而在夏、秋季节，轻度干旱与中度干旱发生频率较高。进一步通过趋势分析可知，春、秋、冬三个季节的干旱程度整体呈现减弱趋势(p<0.05)。其中，春季与秋季约60%~70%的区域呈现出显著下降趋势。夏季则与之相反，干旱强度显著增强，约25.93%的区域表现出明显增强态势(p<0.05),且主要集中于滇中与滇南地区。冬季，滇北区域约12.96%的面积干旱程度呈现显著增强趋势。【结论】COI在云南省的适用性优于其他干旱指数，夏季干旱有明显增强趋势，滇中、滇南区域上升趋势显著。

Abstract

[Objective] This study aims to develop a drought index suitable for the actual conditions of Yunnan Province, [Objective]ly reveal drought variation patterns, and achieve effective drought monitoring. [Methods] The meteorological drought index(Standardized Precipitation-Evapotranspiration Index, SPEI) and the agricultural drought index(Standardized Soil Moisture Index, SSMI) were combined using the entropy weight method to develop a new comprehensive drought index. Various drought indices were evaluated and screened using method such as correlation analysis, root mean square error, and standard deviation to identify the most suitable drought index for Yunnan Province. Then, based on the most suitable drought index, [Methods]such as the run theory and Mann-Kendall trend test were used to thoroughly analyze the spatiotemporal variation characteristics of drought in Yunnan Province. [Results] The comprehensive drought index outperformed other drought indices in reflecting drought conditions, showing correlations of 64% and 60% with meteorological and agricultural drought indices, respectively. Additionally, it showed 94.11% consistency with statistical drought records. During the 2009—2010 winter-spring consecutive drought events, the drought evolution pattern captured by this method showed high consistency with actual drought process. In terms of time, the COI values at both monthly and seasonal scales across the province showed a fluctuating downward trend over the past 20 years, indicating increased drought frequency and severity. Seasonally, drought conditions were the most severe in spring, followed by winter, while summer and autumn showed relatively milder conditions. However, summer drought showed a persistent intensification trend at a rate of 0.034 per year. Spatially, the drought intensity and severity in Yunnan Province showed a concentric pattern, decreasing outward from the central Yunnan. Severe and extremely severe droughts were mainly concentrated in southwest and central Yunnan in spring and winter, while mild and moderate droughts occurred more frequently in summer and autumn. Trend analysis further showed a significant weakening trend(p<0.05) in drought severity during spring, autumn, and winter. Specifically, about 60% to 70% of areas in spring and autumn showed a significant decrease. In contrast, summer showed significant drought intensification. About 25.93% of areas showed a notable increasing trend(p<0.05), particularly in central and southern Yunnan. In winter, about 12.96% of the northern Yunnan exhibited a significant increase in drought severity. [Conclusion] The COI proves to be more suitable for Yunnan Province than other drought indices. Summer drought shows a distinct intensification trend, with central and southern Yunnan experiencing the most significant increases.

关键词

综合干旱指数 / 熵权法 / 游程理论 / 时空分布特征 / 趋势检验 / 影响因素

Key words

comprehensive drought index / entropy weight method / run theory / spatiotemporal distribution characteristics / trend test / influencing factors

基金资助

国家自然科学基金项目(41761081)

云南省社科重点基金项目(ZD202218)