| 天山北坡经济带城市PM2.5质量浓度时空分布及模拟分析 |
| |
| 引用本文： | 刘琳,张正勇,刘芬,徐丽萍.天山北坡经济带城市PM2.5质量浓度时空分布及模拟分析[J].环境科学研究,2018,31(11):1849-1857. |
| |
| 作者姓名： | 刘琳 张正勇 刘芬 徐丽萍 |
| |
| 作者单位： | 1.石河子大学理学院, 新疆 石河子 832000 |
| |
| 基金项目： | 国家自然科学基金项目（No.41461086,41761030）;石河子大学"3152"高层次人才培养支持计划项目（No.CZ0227） |
| |
| 摘 要： | PM2.5扩散/积聚有着复杂的物理、化学过程,其时空变化与前体污染物及气象条件密切相关.为探究区域城市ρ（PM2.5）的变化规律,以天山北坡经济带城市为研究区,采用2015年1月-2017年10月大气污染物监测数据和气象数据,结合数理统计和GIS空间分析技术,分析研究区ρ（PM2.5）的时空变化特征;借助普通OLS（最小二乘法）分析影响因子的多重相关性,通过PLS（偏最小二乘）法构建ρ（PM2.5）估算模型.结果表明:①研究区各城市小时ρ（PM2.5）呈"W"型双峰变化;各月份ρ（PM2.5）呈"U"型特征,月均值为59.5 μg/m3,2月和9月分别为ρ（PM2.5）最高月和最低月;各季节ρ（PM2.5）排序为冬季（146.6 μg/m3）＞秋季（35.2 μg/m3）＞春季（34.1 μg/m3）＞夏季（26.8 μg/m3）;空间上ρ（PM2.5）由西北部克拉玛依市向东南部乌鲁木齐市逐渐增大.②PLS法构建模型能有效克服自变量多重相关性的问题,模型可解释自变量95.7%和因变量80.1%的变异信息,年均模拟值与实测值偏差为9.82%.③研究区各城市ρ（PM2.5）与ρ（CO）的相关性极显著,与气象因子中的风速和气温的相关性较显著,而与相对湿度的相关性不显著.研究显示,基于前体污染物和气象因子的PLS法构建模型是模拟ρ（PM2.5）的有效方法.
|
| 关 键 词： | PM2.5 前体物质 气象条件 PLS 模拟分析 |
| 收稿时间： | 2018/1/1 0:00:00 |
| 修稿时间： | 2018/3/18 0:00:00 |
Spatial-Temporal Distribution and Simulation Analysis of PM2.5 Concentration of the Cities in the Northern Slope Economic Zone of Tianshan Mountain |
| |
| LIU Lin,ZHANG Zhengyong,LIU Fen and XU Liping.Spatial-Temporal Distribution and Simulation Analysis of PM2.5 Concentration of the Cities in the Northern Slope Economic Zone of Tianshan Mountain[J].Research of Environmental Sciences,2018,31(11):1849-1857. |
| |
| Authors: | LIU Lin ZHANG Zhengyong LIU Fen and XU Liping |
| |
| Affiliation: | 1.College of Science, Shihezi University, Shihezi 832000, China2.College of Water and Architectural Engineering, Shihezi University, Shihezi 832000, China |
| |
| Abstract: | The diffusion and accumulation of PM2.5 is a complex physical and chemical process, which is closely related to the concentration of precursor pollutants and meteorological conditions. To research the concentration variation features of PM2.5 in the study area, combining mathematical statistics and GIS spatial analysis, the monitoring data on air pollutants and meteorological elements from January 2015 to October 2017 were used to explore the temporal and spatial distribution of PM2.5 concentrations in cities of the northern slope economic zone of Tianshan Mountain. Multiple correlations of influencing factors were analyzed by the ordinary least squares (OLS) method, and a PM2.5 estimation model was constructed by the Partial Least-Squares (PLS) method. The results showed that:(1) Among all cities in the study area, PM2.5 followed a 'W' curve in hourly concentration while the monthly mean density obeyed a 'U' curve. The monthly average concentration of PM2.5 was 59.5 μg/m3, with the highest in February and the lowest in September. The average concentrations of PM2.5 in different seasons decreased in the order winter, autumn, spring and summer, at 146.6, 35.2, 34.1 and 26.8 μg/m3, respectively. The concentration of PM2.5 in each season increased gradually from Karamay in the northwest to Urumqi in the southeast. (2) An evaluation model was set up using the method of PLS, which can easily solve multiple correlation problems. The model can explain the spatial variation of 95.7% of the independent variables and 80.1% of the dependent variables. The average annual simulated value and the measured deviation was about 9.82%. (3) The concentration of PM2.5 was most affected by pollutants such as CO, and the wind speed and temperature also had significant influences, but the correlation with relative humidity was not obvious. Research shows that PLS modeling based on precursory pollutants and meteorological factors is an effective way to simulate the temporal and spatial distribution of PM2.5 concentrations. |
| |
| Keywords: | PM2 5 precursor substances meteorological conditions PLS simulation analysis |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《环境科学研究》浏览原始摘要信息 |
|
点击此处可从《环境科学研究》下载全文 |
|
