基于LUR模型的PM2.5浓度空间分布监测及分析

为有效解决传统监测技术无法获取城市内部高分辨率PM_2.5浓度空间分布情况的问题，基于土地利用回归(land use regression，LUR)模型，以关中平原城市群为例模拟其PM_2.5空间分布状况，通过获取研究范围内54个监测站点的PM_2.5浓度数据，结合土地利用类型、气象、地形、植被指数、人口密度、交通和污染源等因素，分别建立春、夏、秋、冬及年均5个LUR模型。结果表明：LUR模型调整后各季节及年平均值的R²分别达到0.831 (春)、0.817 (夏)、0.874 (秋)、0.857 (冬)、0.900 (全年平均)，5种模型拟合度均较好；采取交叉互验的方法进行了精度检验，显示5种模型的平均精度均达到80.4%，说明LUR模型在模拟关中平原城市群PM_2.5浓度空间分布时适用性良好。模拟结果显示，研究区各季节的PM_2.5浓度在空间分布上大致相同，呈现出东部高、西部低的明显特征，且空间分布状况受地形因素的影响较大。但在浓度均值的季节变化上则具有夏季低、冬季高的明显差异。本研究结果可为关中平原城市群PM_2.5污染防治提供科学依据，亦可为城市内部PM_2.5浓度空间分布数据的获取提供新思路。

Monitoring and analysis of PM2.5 concentration spatial distribution based on LUR model

In order to effectively solve the problem that the traditional monitoring technology cannot obtain the high-resolution spatial distribution of PM2.5 concentration in the city, the Guanzhong plain city group was taken as an example to simulate its PM2.5 spatial distribution status based on the land use regression (LUR) model. The 5 LUR models for spring, summer, autumn, winter and an annual average were built through obtaining PM2.5 concentration data of 54 monitoring stations in the study range and combining the factors such as land use type, meteorology, terrain, vegetation index, population density, traffic and pollution sources. The results showed that the adjusted R2 of each season and annual average of the LUR model reached 0.831 (spring), 0.817 (summer), 0.874 (autumn), 0.857 (winter), 0.90 (annual average), respectively, and better fitting levels occurred for the five models. A cross-examination method was used to carry out the accuracy test, and the average accuracy of the five models reached 80.4%, indicating that the LUR model had good applicability when simulating the spatial distribution of PM2.5 concentration in the Guanzhong plain city group. The simulation results showed that the PM2.5 concentration in each season of the study area was roughly same in spatial distribution with the significant characteristics of high in the east, low in the west, and obvious distribution trends along the altitude. However, there was a clear difference of low in summer and high in winter for the seasonal change of the mean concentration. The results of this study can provide a scientific basis for the prevention and control of PM2.5 pollution in the Guanzhong plain city group, and can also provide new ideas for obtaining the spatial distribution data of PM2.5 concentration within the city.