摘要
采用多元线性回归分析方法,以近期某年中国31个省级PM2.5排放数据为因变量,选取省级的高速公路长度、干线公路长度、支线公路长度、人口、GDP和工业总产值6个因子为自变量构建模型,对PM2.5排放数据进行空间分布模拟,得到近期中国PM2.5排放数据公里格网分布图,然后在此基础上分析了PM2.5排放数据强度的空间分布规律。结果表明:(1)大部分省的相对误差小于30%,模拟结果精度较高,表明多元线性回归模型可以准确的模拟PM2.5排放数据的空间分布;(2)中国PM2.5排放强度区域分布差异明显,整体表现为从东部向中西部逐渐降低,同时还存在几个明显的高值和低值区域。
Based on the idea of attribute data spatialization,we first built a multiple regression model with the length of expressway,the length of trunk highway,the length of branch highway,population,GDP and gross industrial output value as independent variables,PM2.5 emission data as dependent variables.Then the PM2.5 emission data was spatialized from provincial level to the kilometer grid level with the method of combining model computation with error correction.Finally we analyzed the spatial distribution patterns of PM2.5 emission data.The results showed:(1)The mean relative errors of most provinces are less than 30%,which indicates the model is characterized by higher precision;(2)There were distinct regional differences distribution characteristics,which mainly showed that the intensity of PM2.5 emission data gradually reduced from the eastern areas to the middle-western areas,and there were several obvious high-value regions and low-value regions in China.
作者
王晖
夏既胜
WANG Hui;XIA Ji-sheng(School of Environment and Earth Science,Yunnan University,Kunming 650091,Yunnan,China)
出处
《云南地理环境研究》
2018年第1期46-51,共6页
Yunnan Geographic Environment Research
基金
国家自然科学基金项目(41461103)
