Improved hourly estimate of aerosol optical thickness over Asian land by fusing geostationary satellites Fengyun-4B and Himawari-9.

Asian over-land aerosols are complexities due to a mixture of anthropogenic air pollutants and natural dust. The accuracy of the aerosol optical thickness (AOT) retrieved from the satellite is crucial to their application in the aerosol data assimilation system. Fusion of AOTs with high spatiotemporal resolution from next-generation geostationary satellites such as Fengyun-4B (FY-4B) and Himawari-9, provides a new high-quality dataset capturing the aerosol spatiotemporal variability for data assimilation. This study develops a complete fusion algorithm to estimate the optimal AOT over-land in Asia from September 2022 to August 2023 at 10 km × 10 km resolution with high efficiency. The data fusion involves four steps: (1) investigating the spatiotemporal variability of FY-4B AOT within the past 1 h and 12 km radius calculation domain; (2) utilizing the aerosol spatiotemporal variability characteristics to estimate FY-4B pure and hourly merged AOTs; (3) performing bias corrections for FY-4B and Himwari-9 hourly merged AOT for different observation times and seasons considering pixel-level errors for each satellite; (4) fusing the bias-corrected FY-4B and Himawari-9 hourly merged AOT based on maximum-likelihood estimation (MLE) method. Compared to the original FY-4B AOT, validation with AERONET observation confirms that the root mean square error (RMSE) of hourly merged FY-4B AOT decreases by around 40.6 % and the correlation coefficient (CORR) increases by about 27.8 %. Compared to FY-4B and Himawari-9 merged AOT, the fused AOT significantly decreases (increases) RMSE (CORR) by around 24.7 % (7.3 %) and 20.2 % (5.6 %). In addition, fused AOT is double the number of single-sensor merged AOT. Fusion aerosol map accurately describes the spatial and temporal variations in Asian regions controlled by air pollution and dust storms. Further studies are required for other landscapes with different satellite combinations to promote the application in the data assimilation system.

Long-term (2006-2015) variations and relations of multiple atmospheric pollutants based on multi-remote sensing data over the North China Plain.

In this analysis, the Aqua/MODIS aerosol optical thickness (AOD), Aura/OMI tropospheric NO2 and SO2 column concentration from 2006 to 2015 were used to statistically analyze the spatial distribution characteristics and variation trends of three polluted parameters from three temporal scales of monthly, seasonal and annual average. The results showed that the minimum values of NO2 and SO2 column concentrations both appeared in July and August, and the maximum values appeared in December and January, which was contrary to the variations in AOD. The highly polluted levels were mainly distributed in Shijiazhuang, Xingtai, and Yancheng cities of Hebei Province, and gradually transported to Zhengzhou, Henan Province, north and southwest of Shandong Province, and Tianjin, along the main line of Taiyuan-Linyi, Shanxi Province. AOD and NO2 had significant differences on the seasonal average scale, whereas SO2 had little changes. These pollutants had declined year by year since 2011, in the 10-year period, AOD and SO2 respectively decreased by 17.14% and 10.57%, and only NO2 rose from 8.69 × 1015 molecules/cm2 in 2006 to 9.10 × 1015 molecules/cm2 in 2015 with the increase rate of 4.79%. Integrated with MODIS-released fire products and the Multi-resolution Emission Inventory for China (MEIC), high AOD values in summer were usually accompanied by frequent biomass burning, and heavy heating demand of coal burning led to largest NO2 and SO2 levels in winter. Both inter-annual variations of MEIC NOx and OMI-observed NO2 responded to emission reductions of vehicle exhaustions positively, but vehicle population in Henan and Shandong provinces need to be further controlled. The significant decline of SO2 is mainly attributed to the enforcement of de-sulfurization devices in power plants. Our study found that in the treatment of complex atmospheric pollution, in addition to strict control of common sources of emissions from AOD, NO2 and SO2, it is also necessary to consider their individual characteristics.

Assessment of satellite-estimated near-surface sulfate and nitrate concentrations and their precursor emissions over China from 2006 to 2014.

China is the largest anthropogenic aerosol-generating country worldwide; however, few studies have analyzed the PM2.5 chemical components and their underlying precursor emissions over long periods and across the national domain. First, global 3-D tropospheric chemistry and transport model (GEOS-Chem)-integrated satellite-retrieved aerosol optical depth (AOD) and vertical profiles were used to estimate near-surface sulfate and nitrate levels at 10-km resolution over China from 2006 to 2014. Ground measurement validation of our satellite model yielded correlation coefficients (r) of 0.7 and 0.73 and normalized mean bias (NMB) values of -37.96% and - 32.73% for sulfate and nitrate, respectively. Second, analyses of the spatiotemporal distributions of sulfate and nitrate as well as the vertical density Ozone Monitoring Instrument (OMI)-measured SO2 (PBL_SO2) and NO2 (TVCD_NO2) indicated that the highest nitrate and sulfate levels occurred in the North China Plain (~25 µg/m3) and Sichuan Basin (SCB) (~30 µg/m3), respectively. The long-term variations in the estimated components and precursor gases indicated that the large sulfate decline was positively correlated with the SO2 emission reduction due to the mandatory desulfurization implemented in 2007. The annual growth rate of sulfate relative to the national mean was -6.19%/yr, and the concentration decreased by 17.10% from 2011 to 2014. Energy consumption increases and a lack of control measures for NO2 resulted in persistent increases in NO2 emissions and nitrate concentrations from 2006 to 2010, particularly in the SCB. With energy consumption structure advancements, reductions in NO2 emissions and corresponding nitrate levels over three typical regions were prominent after 2012. Third, the estimated national-scale uncertainties of satellite datasets at 0.1°â€¯× 0.1° were 26.88% for sulfate and 25.55% for nitrate. Differences in the spatial distributions and temporal trends between our estimated components and precursor gases were mainly attributed to the dataset accuracy, the data pre-processing strategy, inconsistent column density and near-surface mass concentration, meteorological variables and complex chemical reactions.

Assessment of Retrieved N2O, NO2, and HF Profiles from the Atmospheric Infrared Ultraspectral Sounder Based on Simulated Spectra.

The Atmospheric Infrared Ultraspectral Sounder (AIUS), the first high-resolution (0.02 cm−1) solar occultation sounder, aboard GF5, was launched in May 2018 from China. However, relevant studies about vertical profiles of atmospheric constituents based on its operational data were not conducted until half a year later. Due to an urgent need for Hin-orbit tests, the real spectra (called reference spectra hereafter) were substituted with simulated spectra calculated from the reference forward model (RFM) plus different random noises at different altitudes. In the generation process of the reference spectra for N2O, NO2, and HF species, ACE-FTS (Atmospheric Chemistry Experiment⁻Fourier Transform Spectrometer instrument on the SCISAT satellite) level 2 products replace corresponding profiles included in the atmospheric background profiles. The optimal estimation method is employed to extract N2O, NO2, and HF profiles in this study. Comparing the retrieved results with ACE-FTS level 2 products, the relative deviations for these three species are calculated. For N2O, the average relative deviation is less than 6% at altitudes below 25 km, while larger deviations are observed in the range of 25⁻45 km, with the maximum being at ~25%. Additionally, the difference for NO2 is less than 5% in the 20⁻45 km range, with a larger discrepancy found below 20 km and above 45 km; the maximum deviation reaches ±40%. For HF, the relative deviation is less than 6% for all tangent heights, implying satisfactory retrieval. The vertical resolution, averaging kernel, and number of degrees of freedom are used to assess the retrieval algorithm, which indicate that the retrieved information content is much more attributable to the reference spectra contribution than to the a priori profile. Finally, a large number of retrieval tests are performed for N2O, NO2, and HF in selected areas covering the Arctic region, northern middle latitude, tropics, southern middle latitude, and Antarctic region, and reliable results are obtained. Thus, to a great extent, the algorithm adopted in the AIUS system can process retrievals reliably and precisely.

[Haze Spectral Analysis and Detection Algorithm Using Satellite Remote Sensing Data].

Frequent occurring of haze pollution events and high fine particulate matter (PM(2.5)) concentration in China have attracted more and more attention in the world. Satellite remote sensing can be used to characterize the air pollution. However, haze is usually misidentified as fog, thin cloud or bright surface in NASA's Moderate Resolution Imaging Spectrometer (MODIS) cloud and clear days' aerosol products, and the retrieval of its optical properties is not included in MODIS cloud detection and dark target algorithm. This approach first studies the spectral characters of cloud, fog, haze, and land cover pixels. Second, following the previous cloud detection and aerosol retrieval literatures, a threshold algorithm is developed to distinguish haze from other pixels based on MODIS multi-band apparent reflectance and brightness temperature. This algorithm is used to detect the haze distribution over North China Plain in 2008 spring and summer. Our result shows a good agreement with the true-color satellite images, which enhances MODIS's ability to monitor the severe air pollution episodes. In addition, the high AOD data from Beijing and Xiang Aerosol Robotic NETwork (AERONET) sites indicate nearly 80% haze days are detected by our approach. Finally, we analyze the errors and uncertainties in haze detection algorithm, and put forward the potential improvements.