[go: up one dir, main page]
 
 
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
Applications of Artificial Intelligence in Atmospheric Sciences
share announcement

Applications of Artificial Intelligence in Atmospheric Sciences

A special issue of Atmosphere (ISSN 2073-4433). This special issue belongs to the section "Atmospheric Techniques, Instruments, and Modeling".

Deadline for manuscript submissions: 31 October 2024 | Viewed by 1006

Special Issue Editors

Dr. Erick G. Sperandio Nascimento

E-Mail Website
Guest Editor
Surrey Institute for People-Centred AI and Global Centre for Clean Air Research (GCARE), Institute for Sustainability, School of Computer Science and Electronic Engineering, University of Surrey, Guildford GU2 7XH, UK
Interests: smart buildings; smart homes; indoor air quality; airborne dispersion; nature-based solutions; low-cost sensors; air monitoring; big data; artificial intelligence; computational modeling
Special Issues, Collections and Topics in MDPI journals
Dr. Taciana Toledo De Almeida Albuquerque

E-Mail Website
Guest Editor
Department of Sanitary and Environmental Engineering, Federal University of Minas Gerais, Av. Antônio Carlos, 6.627, Belo Horizonte, MG 31270-901, Brazil
Interests: air pollution; air particulate matter; air quality; air quality modeling; air pollution control and modeling applications
Special Issues, Collections and Topics in MDPI journals
Prof. Dr. Prashant Kumar

grade E-Mail Website
Guest Editor
Global Center for Clean Air Research (GCARE), School of Sustainability, Civil and Environmental Engineering, Faculty of Engineering and Physical Sciences, University of Surrey, Surrey GU2 7XH, UK
Interests: low-cost sensing; air pollution modelling; pollution mitigation; atmospheric science
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

Current state-of-the-art (SOTA) atmospheric models, such as numerical weather prediction (NWP) models, usually require large computing power since they rely on complex physical equations and parametrisations to simulate and understand spatiotemporal atmospheric phenomena. Currently, artificial intelligence (AI) techniques are used for this purpose with improved forecasting performance, but with a fraction of the computational cost of traditional techniques, leveraging large volumes of historical atmospheric data and advanced AI techniques to build atmospheric models for different spatiotemporal scales. In fact, even world-leading weather agencies, such as the MetOffice in the UK and the European Centre for Medium-Range Weather Forecasts (ECMWF), are developing AI solutions to improve atmospheric modelling performance, presenting competitive performance with SOTA NWP.

Therefore, this Special Issue aims to explore the intersection of AI and atmospheric sciences to tackle pressing challenges in climate change, weather forecasting, clean air, and renewable energy, among others, providing a platform for researchers to showcase cutting-edge research and to foster the development and adoption of AI solutions to address key challenges in atmospheric sciences, with the potential to help achieve the United Nation’s Sustainable Development Goals (UNSDG) 3, 7, 11, and 13. Authors are invited to submit original research articles and reviews that highlight the transformative potential of novel AI techniques in various aspects of atmospheric sciences, including (but not limited to) the following:

  • Weather and extreme weather event forecasting;
  • Air pollution monitoring, management, and forecasting;
  • Renewable energy prediction and optimisation;
  • Regional downscaling;
  • Physics-informed neural networks to simulate atmospheric flow;
  • Foundation models for atmospheric challenges;
  • Climate change and resilience;
  • Indoor and outdoor modelling;
  • The airborne dispersion of contaminants and their impact on indoor and outdoor environments;
  • The inventory estimation of emissions;
  • Land use change assessment;
  • Impacts of air quality on human health;
  • Other related areas.

Dr. Erick G. Sperandio Nascimento
Dr. Taciana Toledo De Almeida Albuquerque
Prof. Dr. Prashant Kumar
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

  • artificial intelligence
  • atmospheric science
  • machine learning
  • deep learning
  • climate change
  • air pollution
  • clean air
  • renewable energy
  • clean energy
  • weather
  • extreme weather
  • physics-informed neural networks

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.
  • e-Book format: Special Issues with more than 10 articles can be published as dedicated e-books, ensuring wide and rapid dissemination.

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

Published Papers (2 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

27 pages, 18384 KiB  
Article
Calibration of Typhoon Track Forecasts Based on Deep Learning Methods
by Chengchen Tao, Zhizu Wang, Yilun Tian, Yaoyao Han, Keke Wang, Qiang Li and Juncheng Zuo
Atmosphere 2024, 15(9), 1125; https://doi.org/10.3390/atmos15091125 (registering DOI) - 17 Sep 2024
Viewed by 182
Abstract
An accurate forecast of typhoon tracks is crucial for disaster warning and mitigation. However, existing numerical weather prediction models, such as the Weather Research and Forecasting (WRF) model, still exhibit significant errors in track forecasts. This study aims to improve forecast accuracy by [...] Read more.
An accurate forecast of typhoon tracks is crucial for disaster warning and mitigation. However, existing numerical weather prediction models, such as the Weather Research and Forecasting (WRF) model, still exhibit significant errors in track forecasts. This study aims to improve forecast accuracy by correcting WRF-forecasted tracks using deep learning models, including Bidirectional Long Short-Term Memory (BiLSTM) + Convolutional Long Short-Term Memory (ConvLSTM) + Wide and Deep Learning (WDL), BiLSTM + Convolutional Gated Recurrent Unit (ConvGRU) + WDL, and BiLSTM + ConvLSTM + Extreme Deep Factorization Machine (xDeepFM), with a comparison to the Kalman Filter. The results demonstrate that the BiLSTM + ConvLSTM + WDL model reduces the 72 h track prediction error (TPE) from 255.18 km to 159.23 km, representing a 37.6% improvement over the original WRF model, and exhibits significant advantages across all evaluation metrics, particularly in key indicators such as Bias2, Mean Squared Error (MSE), and Sequence. The decomposition of MSE further validates the importance of the BiLSTM, ConvLSTM, WDL, and Temporal Normalization (TN) layers in enhancing the model’s spatio-temporal feature-capturing ability. Full article
(This article belongs to the Special Issue Applications of Artificial Intelligence in Atmospheric Sciences)
►▼ Show Figures

Figure 1

Figure 1
<p>A visualization of all forecasted typhoons. Adapted from the figure by Xu [<a href="#B34-atmosphere-15-01125" class="html-bibr">34</a>] et al.</p> Full article ">Figure 2
<p>The three-dimensional time-series structure of a typhoon.</p> Full article ">Figure 3
<p>Methodological Workflow Diagram.</p> Full article ">Figure 4
<p>Typhoon Track Prediction Model Based on the Wide and Deep Framework.</p> Full article ">Figure 5
<p>Typhoon Track Prediction Model Based on the Neural Factorization Machine Framework.</p> Full article ">Figure 6
<p>Typhoon Track Prediction Model Based on the Extreme Deep Factorization Machine Framework.</p> Full article ">Figure 7
<p>Comparative Analysis of Typhoon TPE and Latitude–Longitude RMSE at Different Integration Times in WRF Forecast Results.</p> Full article ">Figure 8
<p>Temporal Trends of WRF-Forecasted Typhoon TPE from 2000 to 2022.</p> Full article ">Figure 9
<p>Scatter Plot Comparing Latitude and Longitude of WRF Forecasts with Best Track at Different Integration Times; the red line indicates the regression line, reflecting the linear relationship between observed and predicted values. (<b>a</b>–<b>c</b>) represent scatter plots for 72 h, 48 h, and 24 h Integration Times along the longitude, and (<b>a1</b>–<b>c1</b>) represent the same along the latitude.</p> Full article ">Figure 10
<p>Comparative Analysis of Overall MSE ((<b>a</b>,<b>a1</b>), Unit: °<sup>2</sup>), Bias<sup>2</sup> ((<b>b</b>,<b>b1</b>), Unit: °<sup>2</sup>), Distribution ((<b>c</b>,<b>c1</b>), Unit: °<sup>2</sup>), and Sequence ((<b>d</b>,<b>d1</b>), Unit: °<sup>2</sup>) for Latitude and Longitude Directions Across Different Models on the 2018–2022 Test Set. The top panels show metrics for the latitude direction, while the bottom panels show metrics for the longitude direction. The models labeled in the figure are BiLSTM + ConvGRU + WDL (A), BiLSTM + ConvLSTM + WDL (B), WRF (C), BiLSTM + ConvLSTM + xDeepFM (D), and Kalman Filter (E). In each subplot, the curves of different colors correspond to 72-h, 48-h, and 24-h integration times. (<b>a</b>–<b>d</b>) represent latitudinal variables, and (<b>a1</b>–<b>d1</b>) represent longitudinal variables.</p> Full article ">Figure 11
<p>Spatial Distribution Comparison of MSE (Unit: °<sup>2</sup>) for Latitude and Longitude Directions Across BiLSTM + ConvGRU + WDL, BiLSTM + ConvLSTM + WDL, BiLSTM + ConvLSTM + DeepFM, WRF, and Kalman Filter on the 2018–2022 Typhoon Test Set. The (<b>top panels</b>) show MSE distribution for latitude, while the (<b>bottom panels</b>) show MSE distribution for longitude.</p> Full article ">Figure 12
<p>Spatial Distribution Comparison of Bias<sup>2</sup> (Unit: °<sup>2</sup>) for Latitude and Longitude Directions Across BiLSTM + ConvGRU + WDL, BiLSTM + ConvLSTM + WDL, BiLSTM + ConvLSTM + DeepFM, WRF, and Kalman Filter on the 2018–2022 Typhoon Test Set. The (<b>top panels</b>) show Bias<sup>2</sup> distribution for latitude, while the (<b>bottom panels</b>) show Bias<sup>2</sup> distribution for longitude.</p> Full article ">Figure 13
<p>Spatial Distribution Comparison of Distribution (Unit: °<sup>2</sup>) for Latitude and Longitude Directions Across BiLSTM + ConvGRU + WDL, BiLSTM + ConvLSTM + WDL, BiLSTM + ConvLSTM + DeepFM, WRF, and Kalman Filter on the 2018–2022 Typhoon Test Set. The (<b>top panels</b>) show the distribution metric for latitude, while the (<b>bottom panels</b>) show the distribution metric for longitude.</p> Full article ">Figure 14
<p>Spatial Distribution Comparison of Sequence (Unit: °<sup>2</sup>) for Latitude and Longitude Directions Across BiLSTM + ConvGRU + WDL, BiLSTM + ConvLSTM + WDL, BiLSTM + ConvLSTM + DeepFM, WRF, and Kalman Filter on the 2018–2022 Typhoon Test Set. The (<b>top panels</b>) show the sequence metric for latitude, while the (<b>bottom panels</b>) show the sequence metric for longitude.</p> Full article ">Figure 15
<p>Comparison of Track Prediction Performance in Latitude and Longitude Directions Between the WRF Model and Correction Models (BiLSTM + ConvLSTM + WDL and BiLSTM + ConvGRU + WDL) on the 2018–2022 Typhoon Test Set. The red line indicates the regression line, reflecting the linear relationship between observed and predicted values; the red shaded area represents the 95% confidence interval of the regression line.</p> Full article ">Figure 16
<p>Path Comparison of Typhoon In-fa (2021), Showing the Differences Between the WRF, BiLSTM + ConvLSTM + WDL, and BiLSTM + ConvGRU + WDL Methods and Historical Typhoon Tracks. Adapted from the figure by Xu [<a href="#B34-atmosphere-15-01125" class="html-bibr">34</a>] et al.</p> Full article ">Figure 17
<p>Path Comparison of Typhoon Chanthu (2021), Showing the Differences Between the WRF, BiLSTM + ConvLSTM + WDL, and BiLSTM + ConvGRU + WDL Methods and Historical Typhoon Tracks. Adapted from the figure by Xu [<a href="#B34-atmosphere-15-01125" class="html-bibr">34</a>] et al.</p> Full article ">
14 pages, 29945 KiB  
Article
Improving Air Quality Prediction via Self-Supervision Masked Air Modeling
by Shuang Chen, Li He, Shinan Shen, Yan Zhang and Weichun Ma
Atmosphere 2024, 15(7), 856; https://doi.org/10.3390/atmos15070856 - 19 Jul 2024
Viewed by 483
Abstract
Presently, the harm to human health created by air pollution has greatly drawn public attention, in particular, vehicle emissions including nitrogen oxides as well as particulate matter. How to predict air quality, e.g., pollutant concentration, efficiently and accurately is a core problem in [...] Read more.
Presently, the harm to human health created by air pollution has greatly drawn public attention, in particular, vehicle emissions including nitrogen oxides as well as particulate matter. How to predict air quality, e.g., pollutant concentration, efficiently and accurately is a core problem in environmental research. Developing a robust air quality predictive model has become an increasingly important task, holding practical significance in the formulation of effective control policies. Recently, deep learning has progressed significantly in air quality prediction. In this paper, we go one step further and present a neat scheme of masked autoencoders, termed as masked air modeling (MAM), for sequence data self-supervised learning, which addresses the challenges posed by missing data. Specifically, the front end of our pipeline integrates a WRF-CAMx numerical model, which can simulate the process of emission, diffusion, transformation, and removal of pollutants based on atmospheric physics and chemical reactions. Then, the predicted results of WRF-CAMx are concatenated into a time series, and fed into an asymmetric Transformer-based encoder–decoder architecture for pre-training via random masking. Finally, we fine-tune an additional regression network, based on the pre-trained encoder, to predict ozone (O 3) concentration. Coupling these two designs enables us to consider the atmospheric physics and chemical reactions of pollutants while inheriting the long-range dependency modeling capabilities of the Transformer. The experimental results indicated that our approach effectively enhances the WRF-CAMx model’s predictive capabilities and outperforms pure supervised network solutions. Overall, using advanced self-supervision approaches, our work provides a novel perspective for further improving air quality forecasting, which allows us to increase the smartness and resilience of the air prediction systems. This is due to the fact that accurate prediction of air pollutant concentrations is essential for detecting pollution events and implementing effective response strategies, thereby promoting environmentally sustainable development. Full article
(This article belongs to the Special Issue Applications of Artificial Intelligence in Atmospheric Sciences)
►▼ Show Figures

Figure 1

Figure 1
<p><b>Left</b>: Traditional air prediction pipeline. <b>Right</b>: The proposed masked air modeling framework for improving air quality prediction.</p> Full article ">Figure 2
<p>Schematic illustration of the Transformer-based masked air modeling.</p> Full article ">Figure 3
<p><b>Left</b>: The location of the YRD. <b>Right</b>: The spatial distribution of air quality monitoring sites.</p> Full article ">Figure 4
<p>Scatter density plots of cross-validation results for the WRF-CAMx model (<b>left</b>) and our MAM model (<b>right</b>). Cells with aggregate counts up to 1% of the total will be colored. Each row from top to bottom represents the simulation results in January, April, July, and October, respectively.</p> Full article ">Figure 5
<p>Time series comparison. From top to bottom: Shanghai, Zhejiang, Jiangsu, and Anhui. From left to right: January, April, July, and October.</p> Full article ">Figure 6
<p>Air quality prediction accuracy in different geographic locations.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-2
Back to TopTop