PDF(Pubmed)
Abstract：
The concentration of gases in the atmosphere is a topic of growing concern due to its effects on health, ecosystems etc. Its monitoring is commonly carried out through ground stations which offer high precision and temporal resolution. However, in countries with few stations, such as Ecuador, these data fail to adequately describe the spatial variability of pollutant concentrations. Remote sensing data have great potential to solve this complication. This study evaluates the spatiotemporal distribution of nitrogen dioxide (NO2) and ozone (O3) concentrations in Quito and Cuenca, using data obtained from ground-based and Sentinel-5 Precursor mission sources during the years 2019 and 2020. Moreover, a Linear Regression Model (LRM) was employed to analyze the correlation between ground-based and satellite datasets, revealing positive associations for O3 (R2 = 0.83, RMSE = 0.18) and NO2 (R2 = 0.83, RMSE = 0.25) in Quito; and O3 (R2 = 0.74, RMSE = 0.23) and NO2, (R2 = 0.73, RMSE = 0.23) for Cuenca. The agreement between ground-based and satellite datasets was analyzed by employing the intra-class correlation coefficient (ICC), reflecting good agreement between them (ICC ≥0.57); and using Bland and Altman coefficients, which showed low bias and that more than 95% of the differences are within the limits of agreement. Furthermore, the study investigated the impact of COVID-19 pandemic-related restrictions, such as social distancing and isolation, on atmospheric conditions. This was categorized into three periods for 2019 and 2020: before (from January 1st to March 15th), during (from March 16th to May 17th), and after (from March 18th to December 31st). A 51% decrease in NO2 concentrations was recorded for Cuenca, while Quito experienced a 14.7% decrease. The tropospheric column decreased by 27.3% in Cuenca and 15.1% in Quito. O3 showed an increasing trend, with tropospheric concentrations rising by 0.42% and 0.11% for Cuenca and Quito respectively, while the concentration in Cuenca decreased by 14.4%. Quito experienced an increase of 10.5%. Finally, the reduction of chemical species in the atmosphere as a consequence of mobility restrictions is highlighted. This study compared satellite and ground station data for NO2 and O3 concentrations. Despite differing units preventing data validation, it verified the Sentinel-5P satellite\'s effectiveness in anomaly detection. Our research\'s value lies in its applicability to developing countries, which may lack extensive monitoring networks, demonstrating the potential use of satellite technology in urban planning.
摘要：
大气中的气体浓度由于其对健康的影响而成为人们日益关注的话题，生态系统等。它的监测通常通过地面站进行，地面站提供高精度和时间分辨率。然而,在车站很少的国家，比如厄瓜多尔，这些数据未能充分描述污染物浓度的空间变异性。遥感数据具有解决这一难题的巨大潜力。这项研究评估了基多和昆卡的二氧化氮（NO2）和臭氧（O3）浓度的时空分布，使用从2019年和2020年的地面和Sentinel-5前体任务来源获得的数据。此外，线性回归模型(LRM)用于分析地基数据集和卫星数据集之间的相关性，在基多中，O3（R2=0.83，RMSE=0.18）和NO2（R2=0.83，RMSE=0.25）呈正相关；Cuenca的O3（R2=0.74，RMSE=0.23）和NO2（R2=0.73，RMSE=0.23）。采用类内相关系数(ICC)分析了地面数据集和卫星数据集之间的一致性，反映了它们之间的良好一致性(ICC≥0.57)；并使用Bland和Altman系数，这显示了低偏差，超过95%的差异在一致的范围内。此外，该研究调查了COVID-19大流行相关限制的影响，比如社交距离和孤立，大气条件。这分为2019年和2020年的三个时期：之前（1月1日至3月15日），期间（3月16日至5月17日），之后（从3月18日到12月31日）。Cuenca的NO2浓度下降了51%，而基多则下降了14.7%。Cuenca的对流层柱减少了27.3％，基多的对流层柱减少了15.1％。O3呈增加趋势,Cuenca和Quito的对流层浓度分别上升了0.42%和0.11%，而Cuenca中的浓度下降了14.4%。基多经历了10.5%的增长。最后,强调了由于流动性限制而导致的大气中化学物质的减少。这项研究比较了卫星和地面站NO2和O3浓度的数据。尽管不同的单位阻止数据验证，它验证了Sentinel-5P卫星在异常检测中的有效性。我们的研究价值在于它对发展中国家的适用性，可能缺乏广泛的监控网络，展示了卫星技术在城市规划中的潜在用途。
