影响滦河流域降水空间分布的多元环境因子作用机制

Influencing mechanism of multi-source environmental factors on the spatial distribution of precipitation in the Luanhe River Basin

为定量评估环境因子对降水空间分布的影响,实现统计降尺度环境因子的科学选取,针对现有方法存在无法避免线性假设、无法保证多自变量共线性免疫和无法有效探索因子间交互作用对降水分布影响的问题,采用地理探测器的定量分析方法,以滦河流域为研究对象,从因子探测、交互探测和生态探测3个方面探究了环境因子及其交互作用对降水空间分布的影响。结果表明:因子探测中,在年、季、月和旬时间尺度上,经度、纬度、坡度、坡向、NDVI、高程的影响力均超过1%,能够直接影响降水的分布,其中纬度、经度和高程对降水分布的影响最为显著;交互探测中,除2月高程-坡度及7月上旬NDVI-经度外,其余交互作用均达到增强的效果,且所有因子交互结果的影响力均超过10%,相较于单因子有较大幅度的提升;生态探测中,不同环境因子对降水分布的影响差异显著。

The purpose of this study is to quantitatively assess the influence of environmental factors on the spatial distribution of precipitation,to scientifically select the environmental factors for the statistical downscaling,and to address the problems of existing methods that cannot avoid the linearity assumption,ensure the immunity of multiple independent variables covariance and effectively explore the influence of factor interactions on the distribution of precipitation.By taking the Luanhe River basin as the research object,a quantitative analysis method based on GeoDetector is proposed and this study investigates the influence of environmental factors and their interactions on the spatial distribution of precipitation from three aspects:factor detection,interaction detection and ecological detection.Firstly,the factor detection results show that the influence of latitude,longitude,slope,slope direction,NDVI and elevation all exceed 1%in the annual,seasonal,monthly and tenday scales,and all these factors can directly affect the distribution of precipitation,among which latitude,longitude and elevation have the most significant influence on the distribution of precipitation.Secondly,the interaction detection results show that except for elevation-slope in February and NDVI-longitude in July,the other interactions are enhanced and the influences of all factors exceed 10%,which are more significant than that of single factor.Thirdly,the ecological detection results show that the influences of different environmental factors on the precipitation distribution are significantly different.