摘要
在变化环境下,中国南方湿润地区干旱事件呈多发趋势。以高度城市化的深圳市为例,选取10个气象代表站（1965—2010年）逐日降雨数据计算不同时间尺度的标准化降水指数（Standard Precipitation Index,SPI）,采用M-K趋势分析法及小波分析法分析深圳市干旱变化趋势及周期以辨识其干湿时空变化特征;采用随机森林算法定量分析混合ENSO指数（Multivariate ENSO Index,MEI）、太平洋年代际振荡指数（Pacific Decadal Oscillation,PDO）、太阳黑子等7个影响因子对深圳市干湿变化的影响程度。结果表明：深圳市的气象干旱多发于1月份,于1—3月呈干旱趋势,8—10月则呈湿润趋势;12个月尺度的SPI序列变化周期集中在2.0~6.5 a;城市中心区为近年干旱主要发生地区,其干旱趋势明显;ENSO现象及PDO是影响深圳市干湿变化的主要影响因素。
The drought events show an increasing trend in humid areas of southern China recently under the changing environment. Shenzhen City, a highly urbanized area in southern China, was taken as a study case to discuss the spatio-temporal features of dry-wet variation. Daily precipitation data （1965--2010） from 10 meteorological stations were utilized to compute the standard precipitation in- dex （SPI） on different time scales; whereas, the Mann-Kendall statistical method and Morlet wavelet analysis were applied to analyzing the variation trends and periods of drought, respectively. Finally, Random Forest algorithm was adopted to quantitatively analyze the influence of seven factors on the dry-wet variation in Shenzhen, such as Multivariate ENSO Index （MEI） , the Pacific Decadal Oscilla- tion （PDO） , sunspot and so on. Results indicate ： the meteorological drought of Shenzhen City frequently occurred in January ; a dry- ness trend was identified from Januaiy to March while a wet trend from August to October; the variation period of SPI at 12-month scale was 2.0 - 6.5 a ; the center of the city becomes more susceptible to meteorological drought recently as the droughts events show signifi- cantly increasing trend in these areas; ENSO and PDO were demonstrated as the two most dominant indexes that greatly impact the dry- wet variation of the city.
出处
《华北水利水电大学学报（自然科学版）》
2016年第3期11-18,共8页
Journal of North China University of Water Resources and Electric Power：Natural Science Edition
基金
国家自然科学基金项目(91547202
51210013
51479216
51479217)
