Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
4.6
2.5
Journals
Atmosphere
Special Issues
Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution...
share announcement

Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources

A special issue of Atmosphere (ISSN 2073-4433). This special issue belongs to the section "Air Quality".

Deadline for manuscript submissions: 9 June 2025 | Viewed by 10329

Special Issue Editors

Dr. Zhenyi Xu

E-Mail Website
Guest Editor
Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei, China
Interests: mobile-source emission prediction; spatiotemporal data; deep learning; intelligent transportation
Dr. Changfa Tao

E-Mail Website
Guest Editor
School of Automotive and Transportation Engineering, Hefei University of Technology, Anhui, China
Interests: Outdoor environmental quality; Tunnel ventilation; Built environment simulation; Pollutant dispersion in street canyons; Smoke movement
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

Mobile-source emissions account for more than 80% of carbon monoxide and hydrocarbons, and more than 90% of nitrogen oxides and solid particles in urban air pollutants. Additionally, these mobile-source emissions have become the main source of urban air pollution, causing serious damage to the social-ecological environment. Therefore, it is necessary to carry out comprehensive supervision and analysis methods of urban mobile-source emissions, as the results obtained are of great significance for protecting public health and improving rational urban planning, as well as traffic conditions. Meanwhile, the temporal and spatial distribution of urban mobile-source emissions is affected by many complex factors. On the one hand, from the perspective of long-term vehicle-emission inventory calculations, it mainly depends on the city's total vehicle volume and vehicle type composition. On the other hand, in terms of short-term and real-time variations in traffic emissions, it is mainly influenced by urban road network topology, traffic flow conditions, and external meteorological factors. This series of factors has led to great challenges in achieving full-time monitoring and comprehensive supervision of urban mobile-source emissions. Summarizing the existing literature, we can find that the focus of mobile-source emission prediction tends to shift from a road segment level to urban region scale, from a single city to multiple cities, from a macro-inventory prediction to fine-grained instantaneous prediction. We propose this Special Issue, “Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources”, to collect state-of-the-art research articles in the field with the hope of sharing views, findings, strategies, and recommendations to achieve equitable access to clean air. 

Dr. Zhenyi Xu
Dr. Changfa Tao
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Atmosphere is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • mobile-source emission spatiotemporal analysis at road level
  • relationships of mobile-source emission variations across regions
  • mobile-source emission control management strategies
  • correlation analysis of air pollution and traffic emissions
  • novel analysis method for heavy-duty vehicle OBD measurement data processing

Benefits of Publishing in a Special Issue

  • Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
  • Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
  • Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
  • External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
  • Reprint: MDPI Books provides the opportunity to republish successful Special Issues in book format, both online and in print.

Further information on MDPI's Special Issue policies can be found here.

Published Papers (8 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

Jump to: Review

26 pages, 8986 KiB  
Article
Comprehensive Scale Fusion Networks with High Spatiotemporal Feature Correlation for Air Quality Prediction
by Chenyi Wu, Zhengliang Lai, Yunwu Xu, Xishun Zhu, Jianhua Wu and Guiqin Duan
Atmosphere 2025, 16(4), 429; https://doi.org/10.3390/atmos16040429 - 8 Apr 2025
Viewed by 344
Abstract
The escalation of industrialization has worsened air quality, underscoring the essential need for accurate forecasting to inform policies and protect public health. Current research has primarily emphasized individual spatiotemporal features for prediction, neglecting the interconnections between these features. To address this, we proposed [...] Read more.
The escalation of industrialization has worsened air quality, underscoring the essential need for accurate forecasting to inform policies and protect public health. Current research has primarily emphasized individual spatiotemporal features for prediction, neglecting the interconnections between these features. To address this, we proposed the generative Comprehensive Scale Spatiotemporal Fusion Air Quality Predictor (CSST-AQP). The novel dual-branch architecture combines multi-scale spatial correlation analysis with adaptive temporal modeling to capture the complex interactions in pollutant dispersion and enhanced pollution forecasting. Initially, a fusion preprocessing module based on localized high-correlation spatiotemporal features encodes multidimensional air quality indicators and geospatial data into unified spatiotemporal features. Then, the core architecture employs a dual-branch collaborative framework: a multi-scale spatial processing branch extracts features at varying granularities, and an adaptive temporal enhancement branch concurrently models local periodicities and global evolutionary trends. The feature fusion engine hierarchically integrates spatiotemporally relevant features at individual and regional scales while aggregating local spatiotemporal features from related sites. In experimental results across 14 Chinese regions, CSST-AQP achieves state-of-the-art performance compared to LSTM-based networks with RMSE 6.11–9.13 μg/m3 and R2 0.91–0.93, demonstrating highly robust 60 h forecasting capabilities for diverse pollutants. Full article
(This article belongs to the Special Issue Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources)
►▼ Show Figures

Figure 1

24 pages, 4783 KiB  
Article
Deep Learning for Atmospheric Modeling: A Proof of Concept Using a Fourier Neural Operator on WRF Data to Accelerate Transient Wind Forecasting at Multiple Altitudes
by Paulo Alexandre Costa Rocha, Jesse Van Griensven Thé, Victor Oliveira Santos and Bahram Gharabaghi
Atmosphere 2025, 16(4), 394; https://doi.org/10.3390/atmos16040394 - 28 Mar 2025
Viewed by 461
Abstract
This study addresses the problem of the computational cost of transient CFD simulations, which rely on iterative time-step calculations, by employing deep learning to generate optimized initial conditions for accelerating the Weather Research and Forecasting (WRF) model. To this end, we forecasted wind [...] Read more.
This study addresses the problem of the computational cost of transient CFD simulations, which rely on iterative time-step calculations, by employing deep learning to generate optimized initial conditions for accelerating the Weather Research and Forecasting (WRF) model. To this end, we forecasted wind speed for short time frames over the Houston region using the WRF model data from 2019 to 2022, training the models to predict the X-component (U) wind speed. The so-called global FNO model, trained across all atmospheric heights, was first tested, achieving competitive results. A more refined approach was tested to improve it, training separate models for each altitude level, enhancing accuracy significantly. These ad hoc models outperformed surface and middle atmosphere persistence, achieving 27.64% and 20.46% nRMSE, respectively, while remaining competitive at higher altitudes. Variable selection played a key role, revealing that different physical processes dominate at various altitudes, necessitating distinct input features. The results highlight the potential of deep learning, particularly FNO, in atmospheric modeling, suggesting that tailored models for specific altitudes may enhance forecast accuracy. Thus, this study demonstrates that a deep learning model can be designed to start the iterations of a transient simulation, reducing convergence time and enabling faster, lower-cost predictions. Full article
(This article belongs to the Special Issue Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources)
►▼ Show Figures

Figure 1

33 pages, 8248 KiB  
Article
Optimizing the Architecture of a Quantum–Classical Hybrid Machine Learning Model for Forecasting Ozone Concentrations: Air Quality Management Tool for Houston, Texas
by Victor Oliveira Santos, Paulo Alexandre Costa Rocha, Jesse Van Griensven Thé and Bahram Gharabaghi
Atmosphere 2025, 16(3), 255; https://doi.org/10.3390/atmos16030255 - 23 Feb 2025
Cited by 2 | Viewed by 982
Abstract
Keeping track of air quality is paramount to issue preemptive measures to mitigate adversarial effects on the population. This study introduces a new quantum–classical approach, combining a graph-based deep learning structure with a quantum neural network to predict ozone concentration up to 6 [...] Read more.
Keeping track of air quality is paramount to issue preemptive measures to mitigate adversarial effects on the population. This study introduces a new quantum–classical approach, combining a graph-based deep learning structure with a quantum neural network to predict ozone concentration up to 6 h ahead. The proposed architecture utilized historical data from Houston, Texas, a major urban area that frequently fails to comply with air quality regulations. Our results revealed that a smoother transition between the classical framework and its quantum counterpart enhances the model’s results. Moreover, we observed that combining min–max normalization with increased ansatz repetitions also improved the hybrid model’s performance. This was evident from evaluating the assessment metrics root mean square error (RMSE), coefficient of determination (R2) and forecast skill (FS). Values for R2 and FS for the horizons considered were 94.12% and 31.01% for the 1 h, 83.94% and 48.01% for the 3 h, and 75.62% and 57.46% for the 6 h forecasts. A comparison with the existing literature for both classical and QML models revealed that the proposed methodology could provide competitive results, and even surpass some well-established forecasting models, proving to be a valuable resource for air quality forecasting, and thus validating this approach. Full article
(This article belongs to the Special Issue Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources)
►▼ Show Figures

Figure 1

18 pages, 750 KiB  
Article
Air Pollutant Concentration Forecasting with WTMP: Wavelet Transform-Based Multilayer Perceptron
by Xiaoling Wang, Liangzhao Tao, Mingliang Fu and Qi Wang
Atmosphere 2024, 15(11), 1296; https://doi.org/10.3390/atmos15111296 - 28 Oct 2024
Viewed by 900
Abstract
Atmospheric pollutants’ real-time changes and the internal interactions among various data make it challenging to efficiently predict concentration variations. In order to extract more information from the time series of pollutants and improve the accuracy of prediction models, we propose a type of [...] Read more.
Atmospheric pollutants’ real-time changes and the internal interactions among various data make it challenging to efficiently predict concentration variations. In order to extract more information from the time series of pollutants and improve the accuracy of prediction models, we propose a type of Multilayer Perceptron model based on wavelet decomposition, named Wavelet Transform-based Multilayer Perceptron (WTMP) model. This model decomposes pollutant data through overlapping discrete wavelet transforms to extract non-stationarity and nonlinear dependencies in the time series. It combines the decomposed data with static covariate information such as data collection time and inputs them into an improved Multilayer Perceptron (MLP) model, reconstructing and outputting the prediction results. Finally, the model is validated using atmospheric pollutant data collected at a specific location in Ruian City, Zhejiang Province, China. The results indicate that the model performs well with minimal prediction errors. Full article
(This article belongs to the Special Issue Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources)
►▼ Show Figures

Figure 1

21 pages, 1583 KiB  
Article
Two-Stream Networks for COPERT Correction Model with Time-Frequency Features Fusion
by Zhenyi Xu, Ruibin Wang, Kai Pan, Jiaren Li and Qilai Wu
Atmosphere 2023, 14(12), 1766; https://doi.org/10.3390/atmos14121766 - 29 Nov 2023
Cited by 3 | Viewed by 1602
Abstract
Emission factors serve as a valuable tool for quantifying the release of pollutants from road vehicles and predicting emissions within a specific time or area. In order to overcome the limitation of the computer program to calculate emissions from the road transport (COPERT) [...] Read more.
Emission factors serve as a valuable tool for quantifying the release of pollutants from road vehicles and predicting emissions within a specific time or area. In order to overcome the limitation of the computer program to calculate emissions from the road transport (COPERT) model in directly obtaining precise emission factors from on-board diagnostic (OBD) data, we propose a novel two-stream network that combines time-series features and time-frequency features to enhance the accuracy of the COPERT model. Firstly, for the instantaneous emission factors of NOx from multiple driving segments provided by heavy-duty diesel vehicles in actual driving, we select the monitored attributes with a high correlation to the emission factor of NOx considering the data scale and employing Spearman rank correlation analysis to obtain the final dataset composed of them and emission factors. Subsequently, we construct an information matrix to capture the impact of past data on emission factors. Each attribute of the time series is then converted into a time-frequency matrix using the continuous wavelet transform. These individual time-frequency matrices are combined to create a multi-channel time-frequency matrix, which represents the historical information. Finally, the historical information matrix and the time-frequency matrix are inputted into a two-stream parallel model that consists of ResNet50 and a convolutional block attention module. This model effectively integrates time-series features and time-frequency features, thereby enhancing the representation of emission characteristics. The reliability and accuracy of the proposed method were validated through a comparative analysis with existing mainstream models. Full article
(This article belongs to the Special Issue Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources)
►▼ Show Figures

Figure 1

25 pages, 17062 KiB  
Article
A Hybrid Autoformer Network for Air Pollution Forecasting Based on External Factor Optimization
by Kai Pan, Jiang Lu, Jiaren Li and Zhenyi Xu
Atmosphere 2023, 14(5), 869; https://doi.org/10.3390/atmos14050869 - 14 May 2023
Cited by 6 | Viewed by 2717
Abstract
Exposure to air pollution will pose a serious threat to human health. Accurate air pollution forecasting can help people to reduce exposure risks and promote environmental pollution control, and it is also an extremely important part of smart city management. However, the current [...] Read more.
Exposure to air pollution will pose a serious threat to human health. Accurate air pollution forecasting can help people to reduce exposure risks and promote environmental pollution control, and it is also an extremely important part of smart city management. However, the current deep-learning-based models for air pollution forecasting usually focus on prediction accuracy improvement without considering the model interpretability. These models usually fail to explain the complex relationships between prediction targets and external factors (e.g., ozone concentration (O3), wind speed, temperature variation, etc.) The relationships between variables in air pollution time series prediction problems are very complex, with intricate relationships between different types of variables, often with nonlinear multivariate dependencies. To address these problems mentioned above, we proposed a hybrid autoformer network with a genetic algorithm optimization to predict air pollution temporal variation as well as establish interpretable relationships between pollutants and external variables. Furthermore, an elite variable voting operator was designed to better filter out more important external factors such as elite variables, so as to perform a more refined search for elite variables. Moreover, we designed an archive storage operator to reduce the effect of neural network model initialization on the search for external variables. Finally, we conducted comprehensive experiments on the Ma’anshan air pollution dataset to verify the proposed model, where the prediction accuracy was improved by 2–8%, and the selection of model influencing factors was more interpretable. Full article
(This article belongs to the Special Issue Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources)
►▼ Show Figures

Figure 1

Review

Jump to: Research

53 pages, 4091 KiB  
Review
Deep Learning in Airborne Particulate Matter Sensing and Surface Plasmon Resonance for Environmental Monitoring
by Balendra V. S. Chauhan, Sneha Verma, B. M. Azizur Rahman and Kevin P. Wyche
Atmosphere 2025, 16(4), 359; https://doi.org/10.3390/atmos16040359 - 22 Mar 2025
Viewed by 472
Abstract
This review explores advanced sensing technologies and deep learning (DL) methodologies for monitoring airborne particulate matter (PM), which is critical for environmental health assessments. It begins with discussing the significance of PM monitoring and introduces surface plasmon resonance (SPR) as a promising technique [...] Read more.
This review explores advanced sensing technologies and deep learning (DL) methodologies for monitoring airborne particulate matter (PM), which is critical for environmental health assessments. It begins with discussing the significance of PM monitoring and introduces surface plasmon resonance (SPR) as a promising technique in environmental applications, alongside the role of DL neural networks in enhancing these technologies. This review analyzes advancements in airborne PM sensing technologies and the integration of DL methodologies for environmental monitoring. This review emphasizes the importance of PM monitoring for public health, environmental policy, and scientific research. Traditional PM sensing methods, including their principles, advantages, and limitations, are discussed, covering gravimetric techniques, continuous monitoring, optical and electrical methods, and microscopy. The integration of DL with PM sensing offers potential for enhancing monitoring accuracy, efficiency, and data interpretation. DL techniques, such as convolutional neural networks (CNNs), autoencoders, recurrent neural networks (RNNs), and their variants, are examined for applications like PM estimation from satellite data, air quality prediction, and sensor calibration. This review highlights the data acquisition and quality challenges in developing effective DL models for air quality monitoring. Techniques for handling large and noisy datasets are explored, emphasizing the importance of data quality for model performance, generalizability, and interpretability. The emergence of low-cost sensor technologies and hybrid systems for PM monitoring is discussed, acknowledging their promise while recognizing the need for addressing data quality, standardization, and integration issues. This review identifies areas for future research, including the development of robust DL models, advanced data fusion techniques, applications of deep reinforcement learning, and considerations of ethical implications. Full article
(This article belongs to the Special Issue Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources)
►▼ Show Figures

Figure 1

35 pages, 3573 KiB  
Review
Analytical Methods for Atmospheric Carbonyl Compounds: A Review
by Xiaoshuai Gao, Xin Zhang, Yan Nie, Jiemeng Bao, Junling Li, Rui Gao, Yunfeng Li, Wei Wei, Xiaoyu Yan, Yongxin Yan and Hong Li
Atmosphere 2025, 16(1), 107; https://doi.org/10.3390/atmos16010107 - 19 Jan 2025
Viewed by 1331
Abstract
Atmospheric carbonyl compounds have significant impacts on the atmospheric environment and human health, making the selection of appropriate analytical techniques crucial for accurately detecting these compounds in specific environments. Based on extensive literature research, this study summarized the development history, relevant features, and [...] Read more.
Atmospheric carbonyl compounds have significant impacts on the atmospheric environment and human health, making the selection of appropriate analytical techniques crucial for accurately detecting these compounds in specific environments. Based on extensive literature research, this study summarized the development history, relevant features, and applicable scenarios of the main analytical techniques for atmospheric carbonyl compounds; pointed out the main problems and challenges in this field; and discussed the needs and prospects of future research and application. It was found that the direct sampling methods of atmospheric carbonyl compounds were applicable to low-molecular-weight carbonyl species with low reactivity, low boiling points, high polarity, and high volatility, while indirect sampling methods were suitable for a wider range and various types and phases of species. For formaldehyde, offline detection was primarily influenced by chemical reagents and reaction conditions, whereas online monitoring relied on sufficiently stable operating environments. For multiple carbonyl compounds, offline detection results were greatly influenced by detectors coupled with chromatography, whereas online monitoring techniques were applicable to all types of volatile organic compounds (VOCs), including some carbonyl compounds, providing higher temporal resolution and improved isomer identification with the development of online mass spectrometry. The combined use of proton transfer reaction-mass spectrometry (PTR-MS) and liquid chromatography-mass spectrometry (GC-MS) was suitable for the detection of carbonyl compounds in atmospheric photochemical smog chamber simulation studies. Currently, offline analytical techniques for carbonyl compounds require significant time and advanced experimental skills for multiple optimization experiments to detect a broader range of species. Online monitoring techniques face challenges such as poor stability and limited species coverage. In smog chamber simulation studies, the detection of carbonyl compounds is heavily influenced by both the sampling system and the chamber itself. Future efforts should focus on improving the environmental adaptability and automation of carbonyl compound analytical techniques, the synergistic use of various techniques, developing new sampling systems, and reducing the impact of the chamber itself on carbonyl compound detection, in order to enhance detection sensitivity, selectivity, time resolution, accuracy, and operability. Full article
(This article belongs to the Special Issue Applying Deep Learning Technology for Spatiotemporal Prediction of Air Pollution from Urban Mobile Sources)
►▼ Show Figures

Figure 1

Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-8
Back to TopTop