Department of Building, Civil and Environmental Engineering, Concordia University, Montreal, Quebec H3G 1M8, Canada.
Environ Pollut. 2013 Oct;181:172-81. doi: 10.1016/j.envpol.2013.06.037. Epub 2013 Jul 15.
This paper investigates the biases associated with the ground-level nitrogen dioxide (NO2) concentrations derived from the satellite Ozone Monitoring Instrument (OMI) NO2 data through comparisons with the modeling and the monitoring results for the state of California in 2008. The seasonal and annual average ground-level NO2 concentrations are both analyzed from the OMI using the local NO2 profile obtained from the GEOS-Chem simulation. The OMI-derived ground-level NO2 concentrations are then compared with the NO2 concentrations predicted by a GIS-Based Multi-Source and Multi-Box model (GMSMB) and the in-situ measurements, correlation coefficients among the three sets of results are all above 0.84 with an average slope of 0.81 ± 0.04. Particularly, various biases associated with the three data sets have been analyzed, and the OMI-derived NO2 concentrations and the GMSMB modeling results have been proven to be essential for assessing regional air pollutant exposure risks with the aid of the extensive remote sensing database.
本研究通过与建模和监测结果的比较，调查了卫星臭氧监测仪（OMI）NO2 数据得出的近地面二氧化氮（NO2）浓度的偏差，该研究针对的是 2008 年加利福尼亚州的情况。本研究利用从 GEOS-Chem 模拟中获取的本地 NO2 廓线，从 OMI 中分析了季节性和年度平均近地面 NO2 浓度。然后，将 OMI 得出的近地面 NO2 浓度与基于 GIS 的多源多箱模型（GMSMB）预测的 NO2 浓度以及现场测量值进行了比较，这三组结果之间的相关系数均高于 0.84，平均斜率为 0.81±0.04。特别是，分析了这三组数据的各种偏差，结果证明，借助广泛的遥感数据库，OMI 得出的 NO2 浓度和 GMSMB 建模结果对于评估区域空气污染物暴露风险非常重要。
