长江三角洲PM_(10)质量浓度遥感估算与时空分布特征

Estimate PM_(10) concentration over the Yangtze Delta using remote sensing:analysis of spatial and temporal variations

卫星遥感反演气溶胶光学厚度(AOT)已被广泛地应用于地面PM10遥感监测。为遥感监测长江三角洲地区PM10,利用2013年的MODIS/Terra AOT产品,考虑研究区36个空气质量监测站点的风速、温度、湿度和边界层高度等气象条件,构建了基于MODIS AOT产品估算PM10的模型。利用17个空气质量监测站点数据对模型进行散点拟合验证,结果表明,模型估算精度较高,春夏秋冬4个季节PM10质量浓度的模型估算值与地面监测值的相关系数R2值分别为0.72、0.76、0.69和0.72。利用模型估算的长时间序列PM10时空分布数据进行时空变化特征分析,结果表明:2000—2013年研究区PM10质量浓度呈增长趋势,月均增长量为0.077μg/m3,最大值出现在2月,为(107.2±22.0)μg/m3,最小值出现在8月,为(40.5±12.0)μg/m3;研究区PM10质量浓度空间分布差异显著,南部低,北部高,高值主要出现在由上海、杭州和南京构成的三角形区域的城市群中,而低值主要出现在南部远离城市的森林区域。结果表明,基于MODIS/Terra AOT产品和地面观测气象数据估算PM10的多元线性回归模型能较好地应用于区域PM10监测。

Satellite remote sensing retrieved aerosol optical thickness（ AOT） data have been widely used in the monitoring of surface particulate matter（ PM） concentrations,specifically PM10. To monitor PM10 by remote sensing techniques in the Yangtze Delta,an estimation model for PM10 concentrations was constructed based on Moderate Resolution Imaging Spectroradiometer（ MODIS） AOT data,PM10 concentration data collected at 36ground-based air quality observation sites,and meteorological data from 2013. Afterwards,the model estimated PM10 was validated with PM10 concentration data from 17 ground-based air quality observation sites,and the results showed that the model estimates had high degrees of accuracy. The correlation coefficient values（ R^2） between model estimates of PM10 concentrations and ground-based monitoring data in the spring,summer,autumn,and winter were 0. 72,0. 76,0. 69,and 0. 72,respectively. The results of the variation characteristics from temporal and spatial analyses based on the long-term PM10 data set estimated with the model showed that PM10 concentrations were associated with an increasing trend from 2000 to 2013,and the average monthly growth rate was 0. 077 μg / m^3; moreover,the maximum concentration of PM10was（ 107. 2 ± 22. 0） μg / m^3 in February and the minimum standard deviation was detected in December for the corresponding concentration of（ 40. 5 ± 12. 0）μg /m^3. Spatial differences were also found in the PM10 data with greater differences in the south and lesser differences in the north. Mass concentrations of PM10 were the highest in urban areas,which are distributed as a deltashaped region consisting of Shanghai,Hangzhou,and Nanjing. Conversely,the concentrations of PM10 were the lowest in the forested regions that are located away from city centers. Overall,these results suggest that MODIS AOT and meteorological data can be used to monitor regional PM10 concentrations with a multi-linear regression model.