Stratospheric influence on surface ozone pollution in China.

Events of stratospheric intrusions to the surface (SITS) can lead to severe ozone (O3) pollution. Still, to what extent SITS events impact surface O3 on a national scale over years remains a long-lasting question, mainly due to difficulty of resolving three key SITS metrics: frequency, duration and intensity. Here, we identify 27,616 SITS events over China during 2015-2022 based on spatiotemporally dense surface measurements of O3 and carbon monoxide, two effective indicators of SITS. An overview of the three metrics is presented, illustrating large influences of SITS on surface O3 in China. We find that SITS events occur preferentially in high-elevation regions, while those in plain regions are more intense. SITS enhances surface O3 by 20 ppbv on average, contributing to 30-45% of O3 during SITS periods. Nationally, SITS-induced O3 peaks in spring and autumn, while over 70% of SITS events during the warm months exacerbate O3 pollution. Over 2015-2022, SITS-induced O3 shows a declining trend. Our observation-based results can have implications for O3 mitigation policies in short and long terms.

Exploring Global Land Coarse-Mode Aerosol Changes from 2001-2021 Using a New Spatiotemporal Coaction Deep-Learning Model.

Coarse-mode aerosol optical depths (cAODs) are critical for understanding the impact of coarse particle sizes, especially dust aerosols, on climate. Currently, the limited data length and high uncertainty of satellite products diminish the applicability of cAOD for climate research. Here, we propose a spatiotemporal coaction deep-learning model (SCAM) for the retrieval of global land cAOD (500 nm) from 2001-2021. In contrast to conventional deep-learning models, the SCAM considers the impacts of spatiotemporal feature interactions and can simultaneously describe linear and nonlinear relationships for retrievals. Based on these unique characteristics, the SCAM considerably improved global daily cAOD accuracies and coverages (R = 0.82, root-mean-square error [RMSE] = 0.04). Compared to official products from the multiangle imaging spectroradiometer (MISR), the moderate resolution imaging spectroradiometer (MODIS), and the polarization and directionality of Earth's reflectances (POLDER) instrument, as well as the physical-deep learning (Phy-DL) derived cAOD, the SCAM cAOD improved the monthly R from 0.44 to 0.88 and more accurately captured over the desert regions. Based on the SCAM cAOD, daily dust cases decreased over the Sahara, Thar Desert, Gobi Desert, and Middle East during 2001-2021 (>3 × 10-3/year). The SCAM-retrieved cAOD can contribute considerably to resolving the climate change uncertainty related to coarse-mode aerosols. Our proposed method is highly valuable for reducing uncertainties regarding coarse aerosols and climate interactions.

New global aerosol fine-mode fraction data over land derived from MODIS satellite retrievals.

The space-borne measured fine-mode aerosol optical depth (fAOD) is a gross index of column-integrated anthropogenic particulate pollutants, especially over the populated land. The fAOD is the product of the AOD and the fine-mode fraction (FMF). While there exist numerous global AOD products derived from many different satellite sensors, there have been much fewer, if any, global FMF products with a quality good enough to understand their spatiotemporal variations. This is key to understanding the global distribution and spatiotemporal variations of air pollutants, as well as their impacts on global environmental and climate changes. Modifying our newly developed retrieval algorithm to the latest global-scale Moderate Resolution Imaging Spectroradiometer (MODIS) aerosol product (Collection 6.1), a global 10-year FMF product is generated and analyzed here. We first validate the product through comparisons with the FMF derived from Aerosol Robotic Network (AERONET) measurements. Among our 169,313 samples, the satellite-derived FMFs agreed with the AERONET spectral deconvolution algorithm (SDA)-retrieved FMFs with a root-mean-square error (RMSE) of 0.22. Analyzed using this new product are the global patterns and interannual and seasonal variations of the FMF over land. In general, the FMF is large (>0.80) over Mexico, Myanmar, Laos, southern China, and Africa and less than 0.5 in the Sahelian and Sudanian zones of northern Africa. Seasonally, higher FMF values occur in summer and autumn. The linear trend in the satellite-derived and AERONET FMFs for different countries was explored. The upward trend in the FMFs was particularly strong over Australia since 2008. This study provides a new global view of changes in FMFs using a new satellite product that could help improve our understanding of air pollution around the world.

A Spatial-Temporal Interpretable Deep Learning Model for improving interpretability and predictive accuracy of satellite-based PM2.5.

Being able to monitor PM2.5 across a range of scales is incredibly important for our ability to understand and counteract air pollution. Remote monitoring PM2.5 using satellite-based data would be incredibly advantageous to this effort, but current machine learning methods lack necessary interpretability and predictive accuracy. This study details the development of a new Spatial-Temporal Interpretable Deep Learning Model (SIDLM) to improve the interpretability and predictive accuracy of satellite-based PM2.5 measurements. In contrast to traditional deep learning models, the SIDLM is both "wide" and "deep." We comprehensively evaluated the proposed model in China using different input data (top-of-atmosphere (TOA) measurements-based and aerosol optical depth (AOD)-based, with or without meteorological data) and different spatial resolutions (10 km, 3 km, and 250 m). TOA-based SIDLM PM2.5 achieved the best predictive accuracy in China, with root-mean-square errors (RMSE) of 15.30 and 15.96 µg/m3, and R2 values of 0.70 and 0.66 for PM2.5 predictions at 10 km and 3 km spatial resolutions, respectively. Additionally, we tested the SIDLM in PM2.5 retrievals at a 250 m spatial resolution over Beijing, China (RMSE = 16.01 µg/m3, R2 = 0.62). Furthermore, SIDLM demonstrated higher accuracy than five machine learning inversion methods, and also outperformed them regarding feature extraction and the interpretability of its inversion results. In particular, modeling results indicated the strong influence of the Tongzhou district on the principle PM2.5 in the Beijing urban area. SIDLM-extracted temporal characteristics revealed that summer months (June-August) might have contributed less to PM2.5 concentrations, indicating the limited accumulation of PM2.5 in these months. Our study shows that SIDLM could become an important tool for other earth observation data in deep learning-based predictions and spatiotemporal analysis.

Understanding global changes in fine-mode aerosols during 2008-2017 using statistical methods and deep learning approach.

Despite their extremely small size, fine-mode aerosols have significant impacts on the environment, climate, and human health. However, current understandings of global changes in fine-mode aerosols are limited. In this study, we employed newly developed satellite retrieval data and an attentive interpretable deep learning model to explore the status, changes, and association factors of the global fine-mode aerosol optical depth (fAOD) and aerosol fine-mode fraction (FMF) from 2008 to 2017. At the global scale, the results show a significant increasing trend in land FMF (2.34 × 10-3/year); however, the FMF over the ocean and the fAOD over land and ocean did not reveal significant trends. Between 2008 and 2017, high levels of both fAOD (>0.30) and FMF (>0.75) were identified over China, southeastern Asia, India, and Africa. Seasonally, global land FMF showed high values in summer (>0.70) and low values in spring (<0.65), while land fAOD was high in summer (>0.15) but low in winter (<0.13). Importantly, Australia and Mexico experienced significant increasing trends in FMF during all four seasons. At the regional scale, a significant decline in fAOD was identified in China, which indicates that government emission controls and reductions have been effective in recent decades. The deep learning model was used to interpret the result and showed that O3 was significantly associated with changes in both the FMF and fAOD. This finding suggests the importance of synergizing the regulations for both O3 and fine particles. Our work comprehensively examined global spatial and seasonal fAOD and FMF changes and provides a holistic understanding of global anthropogenic impacts.

New interpretable deep learning model to monitor real-time PM2.5 concentrations from satellite data.

Particulate matter with a mass concentration of particles with a diameter less than 2.5 µm (PM2.5) is a key air quality parameter. A real-time knowledge of PM2.5 is highly valuable for lowering the risk of detrimental impacts on human health. To achieve this goal, we developed a new deep learning model-EntityDenseNet to retrieve ground-level PM2.5 concentrations from Himawari-8, a geostationary satellite providing high temporal resolution data. In contrast to the traditional machine learning model, the new model has the capability to automatically extract PM2.5 spatio-temporal characteristics. Validation across mainland China demonstrates that hourly, daily and monthly PM2.5 retrievals contain the root-mean-square errors of 26.85, 25.3, and 15.34 µg/m3, respectively. In addition to a higher accuracy achievement when compared with various machine learning inversion methods (backpropagation neural network, extreme gradient boosting, light gradient boosting machine, and random forest), EntityDenseNet can "peek inside the black box" to extract the spatio-temporal features of PM2.5. This model can show, for example, that PM2.5 levels in the coastal city of Tianjin were more influenced by air from Hebei than Beijing. Further, EntityDenseNet can still extract the seasonal characteristics that demonstrate that PM2.5 is more closely related within three month groups over mainland China: (1) December, January and February, (2) March, April and May, (3) July, August and September, even without meteorological information. EntityDenseNet has the ability to obtain high temporal resolution satellite-based PM2.5 data over China in real-time. This could act as an important tool to improve our understanding of PM2.5 spatio-temporal features.

Fugitive Road Dust PM2.5 Emissions and Their Potential Health Impacts.

Fugitive road dust (FRD) particles emitted by traffic-generated turbulence are an important contributor to urban ambient fine particulate matter (PM2.5). Especially in urban areas of developing countries, FRD PM2.5 emissions are a serious environmental threat to air quality and public health. FRD PM2.5 emissions have been neglected or substantially underestimated in previous study, resulting in the underestimation of modeling PM concentrations and estimating their health impacts. This study constructed the FRD PM2.5 emissions inventory in a major inland city in China (Lanzhou) in 2017 at high-resolution (500 × 500 m2), investigated the spatiotemporal characteristics of the FRD emissions in different urban function zones, and quantified their health impacts. The FRD PM2.5 emission was approximately 1141 ± 71 kg d-1, accounting for 24.6% of total PM2.5 emission in urban Lanzhou. Spatially, high emissions exceeding 3 × 104 µg m-2 d-1 occurred over areas with smaller particle sizes, larger traffic intensities, and more frequent construction activities. The estimated premature mortality burden induced by FRD PM2.5 exposure was 234.5 deaths in Lanzhou in 2017. Reducing FRD emissions are an important step forward to protect public health in many developing urban regions.