Unveiling global land fine- and coarse-mode aerosol dynamics from 2005 to 2020 using enhanced satellite-based monthly inversion data.

Accurate fine-mode and coarse-mode aerosol knowledge is crucial for understanding their impacts on the climate and Earth's ecosystems. However, current satellite-based Fine-Mode Aerosol Optical Depth (FAOD) and Coarse-Mode Aerosol Optical Depth (CAOD) methods have drawbacks including inaccuracies, low spatial coverage, and limited temporal duration. To overcome these issues, we developed new global-scale FAOD and CAOD from 2005 to 2020 using a novel deep learning model capable of the synergistic retrieval of two aerosol sizes. After validation with the aerosol robotic network (AERONET) and sky radiometer network (SKYNET), the new monthly FAOD and CAOD showed significant improvements in accuracy and spatial coverage. From 2005 to 2020, the new data showed that China had the greatest decrease in FAOD and CAOD. In contrast, India and South Latin America had a significant increase in FAOD versus North Africa in CAOD. FAOD in the regions of Argentina, Paraguay, and Uruguay in South America have shown an upward trend. The results reveal that FAOD and CAOD display distinct patterns of change in the same regions, particularly on the west coast of the United States where FAOD is increasing, while CAOD is decreasing. Aside from the year 2020 due to the global COVID-19 pandemic, the analysis showed that although China has seen at least an +85% increase in energy consumption and urban expansion in 2019 compared to 2005 due to the needs of development and construction, the implementation of China's air pollution control policies has led to a significant decrease in FAOD (-46%) and CAOD (-65%) after 2013. This research enriches our comprehension of global fine and coarse aerosol patterns, additional investigations are needed to determine the potential global implications of these changes.

Combining physical mechanisms and deep learning models for hourly surface ozone retrieval in China.

As surface ozone (O3) gains increasing attention, there is an urgent need for high temporal resolution and accurate O3 monitoring. By taking advantage of the progress in artificial intelligence, deep learning models have been applied to satellite based O3 retrieval. However, the underlying physical mechanisms that influence surface O3 into model construction have rarely been considered. To overcome this issue, we considered the physical mechanisms influencing surface O3 and used them to select relevant variable features for developing a novel deep learning model. We used a wide and deep model architecture to account for linear and non-linear relationships between the variables and surface O3. Using the developed model, we performed hourly inversions of surface O3 retrieval over China from 2017 to 2019 (9:00-17:00, local time). The validation results based on sample-based (site-based) methods yielded an R2 of 0.94 (0.86) and an RMSE of 12.79 (19.13) µg/m3, indicating the accuracy of the models. The average surface O3 concentrations in China in 2017, 2018, and 2019 were 82, 78, and 87 µg/m3, respectively. There was a diurnal pattern in surface O3 in China, with levels rising significantly from 55 µg/m3 at 9:00 a.m. to 96 µg/m3 at 15:00. Between 15:00 and 16:00, the O3 concentration remained stable at 95 µg/m3 and decreased slightly thereafter (16:00-17:00). The results of this study contribute to a deeper understanding of the physical mechanisms of ozone and facilitate further studies on ozone monitoring, thereby enhancing our understanding of the spatiotemporal characteristics of ozone.

Cooperative simultaneous inversion of satellite-based real-time PM2.5 and ozone levels using an improved deep learning model with attention mechanism.

Ground-level fine particulate matter (PM2.5) and ozone (O3) are air pollutants that can pose severe health risks. Surface PM2.5 and O3 concentrations can be monitored from satellites, but most retrieval methods retrieve PM2.5 or O3 separately and disregard the shared information between the two air pollutants, for example due to common emission sources. Using surface observations across China spanning 2014-2021, we found a strong relationship between PM2.5 and O3 with distinct spatiotemporal characteristics. Thus, in this study, we propose a new deep learning model called the Simultaneous Ozone and PM2.5 inversion deep neural Network (SOPiNet), which allows for daily real-time monitoring and full coverage of PM2.5 and O3 simultaneously at a spatial resolution of 5 km. SOPiNet employs the multi-head attention mechanism to better capture the temporal variations in PM2.5 and O3 based on previous days' conditions. Applying SOPiNet to MODIS data over China in 2022, using 2019-2021 to construct the network, we found that simultaneous retrievals of PM2.5 and O3 improved the performance compared with retrieving them independently: the temporal R2 increased from 0.66 to 0.72 for PM2.5, and from 0.79 to 0.82 for O3. The results suggest that near-real time satellite-based air quality monitoring can be improved by simultaneous retrieval of different but related pollutants. The codes of SOPiNet and its user guide are freely available online at https://github.com/RegiusQuant/ESIDLM.

Structure-Aware Multimodal Deep Learning for Drug-Protein Interaction Prediction.

Identifying drug-protein interactions (DPIs) is crucial in drug discovery, and a number of machine learning methods have been developed to predict DPIs. Existing methods usually use unrealistic data sets with hidden bias, which will limit the accuracy of virtual screening methods. Meanwhile, most DPI prediction methods pay more attention to molecular representation but lack effective research on protein representation and high-level associations between different instances. To this end, we present the novel structure-aware multimodal deep DPI prediction model, STAMP-DPI, which was trained on a curated industry-scale benchmark data set. We built a high-quality benchmark data set named GalaxyDB for DPI prediction. This industry-scale data set along with an unbiased training procedure resulted in a more robust benchmark study. For informative protein representation, we constructed a structure-aware graph neural network method from the protein sequence by combining predicted contact maps and graph neural networks. Through further integration of structure-based representation and high-level pretrained embeddings for molecules and proteins, our model effectively captures the feature representation of the interactions between them. As a result, STAMP-DPI outperformed state-of-the-art DPI prediction methods by decreasing 7.00% mean square error (MSE) in the Davis data set and improving 8.89% area under the curve (AUC) in the GalaxyDB data set. Moreover, our model is an interpretable model with the transformer-based interaction mechanism, which can accurately reveal the binding sites between molecules and proteins.

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.

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.