{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,5]],"date-time":"2026-03-05T12:00:43Z","timestamp":1772712043622,"version":"3.50.1"},"reference-count":49,"publisher":"MDPI AG","issue":"3","license":[{"start":{"date-parts":[[2026,2,28]],"date-time":"2026-02-28T00:00:00Z","timestamp":1772236800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Hainan Provincial Natural Science Foundation of China","award":["623RC455"],"award-info":[{"award-number":["623RC455"]}]},{"name":"Hainan Provincial Natural Science Foundation of China","award":["623RC457"],"award-info":[{"award-number":["623RC457"]}]},{"name":"Hainan Provincial Natural Science Foundation of China","award":["425QN244"],"award-info":[{"award-number":["425QN244"]}]},{"name":"Scientific Research Fund of Hainan University","award":["KYQD (ZR)-22096"],"award-info":[{"award-number":["KYQD (ZR)-22096"]}]},{"name":"Scientific Research Fund of Hainan University","award":["KYQD(ZR)-22097"],"award-info":[{"award-number":["KYQD(ZR)-22097"]}]},{"name":"Lanzhou University-Hainan University Technical Service Project","award":["HD-KYH-2024424"],"award-info":[{"award-number":["HD-KYH-2024424"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IJGI"],"abstract":"<jats:p>Ground-level ozone pollution poses significant risks to public health and ecosystems and remains a major environmental challenge worldwide. Accurate forecasting is difficult due to the nonlinear formation mechanisms of ozone and its strong dependence on meteorological conditions. This study proposes a Wind Speed and Direction-Based Dynamic Spatiotemporal Graph Attention Network (WSDST-GAT) for multi-step hourly ground-level ozone prediction. The model integrates a wind-aware dynamic graph to represent anisotropic pollutant transport and a Transformer-based temporal encoder to capture long-range dependencies. Meteorological variables are incorporated to enhance physical interpretability and predictive robustness. A co-kriging module is further employed to reconstruct continuous spatial ozone fields with quantified uncertainty. Using hourly observations from 35 monitoring stations in Beijing, WSDST-GAT achieves a Coefficient of Determination of 0.957, with a Mean Absolute Error of 5.25 \u03bcg\/m3, and a Root Mean Square Error of 9.58 \u03bcg\/m3. The prediction intervals demonstrate strong reliability with a Prediction Interval Coverage Probability of 94.01% and a Prediction Interval Normalized Average Width of 0.174. These results indicate that the proposed framework provides an accurate and physically informed solution for ozone forecasting and air quality management.<\/jats:p>","DOI":"10.3390\/ijgi15030101","type":"journal-article","created":{"date-parts":[[2026,3,2]],"date-time":"2026-03-02T08:51:57Z","timestamp":1772441517000},"page":"101","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["A Novel Wind-Aware Dynamic Graph Neural Network for Urban Ground-Level Ozone Concentration Prediction"],"prefix":"10.3390","volume":"15","author":[{"ORCID":"https:\/\/orcid.org\/0009-0005-1476-1999","authenticated-orcid":false,"given":"Wenjie","family":"Wu","sequence":"first","affiliation":[{"name":"School of Cyberspace Security (School of Cryptology), Hainan University, Haikou 570228, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4685-3645","authenticated-orcid":false,"given":"Xinyue","family":"Mo","sequence":"additional","affiliation":[{"name":"School of Cyberspace Security (School of Cryptology), Hainan University, Haikou 570228, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7683-2589","authenticated-orcid":false,"given":"Huan","family":"Li","sequence":"additional","affiliation":[{"name":"School of Cyberspace Security (School of Cryptology), Hainan University, Haikou 570228, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,2,28]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"73","DOI":"10.1007\/s11270-024-07702-4","article-title":"Water quality, air pollution, and climate change: Investigating the environmental impacts of industrialization and urbanization","volume":"236","author":"Saxena","year":"2025","journal-title":"Water Air Soil Pollut."},{"key":"ref_2","first-page":"151","article-title":"Environmental pollution causes and consequences: A study","volume":"3","author":"Appannagari","year":"2017","journal-title":"North Asian Int. Res. J. Soc. Sci. Humanit."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"4607","DOI":"10.1016\/j.aej.2021.10.021","article-title":"Comprehensive comparison of various machine learning algorithms for short-term ozone concentration prediction","volume":"61","author":"Yafouz","year":"2022","journal-title":"Alex. Eng. J."},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Mo, X., Li, H., and Zhang, L. (2022). Design a regional and multistep air quality forecast model based on deep learning and domain knowledge. Front. Earth Sci., 10.","DOI":"10.3389\/feart.2022.995843"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"2162","DOI":"10.1016\/S0140-6736(24)00933-4","article-title":"Global burden and strength of evidence for 88 risk factors in 204 countries and 811 subnational locations, 1990\u20132021: A systematic analysis for the Global Burden of Disease Study 2021","volume":"403","author":"Brauer","year":"2024","journal-title":"Lancet"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"148","DOI":"10.1016\/j.scitotenv.2016.10.231","article-title":"Haze, public health and mitigation measures in China: A review of the current evidence for further policy response","volume":"578","author":"Gao","year":"2017","journal-title":"Sci. Total Environ."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"e009493","DOI":"10.1136\/bmjopen-2015-009493","article-title":"Long-term exposure to ambient ozone and mortality: A quantitative systematic review and meta-analysis of evidence from cohort studies","volume":"6","author":"Atkinson","year":"2016","journal-title":"BMJ Open"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"112136","DOI":"10.1016\/j.rse.2020.112136","article-title":"Reconstructing 1-km-resolution high-quality PM2.5 data records from 2000 to 2018 in China: Spatiotemporal variations and policy implications","volume":"252","author":"Wei","year":"2021","journal-title":"Remote Sens. Environ."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"1571","DOI":"10.1007\/s11869-021-01040-8","article-title":"Environmental impact estimation of PM2.5 in representative regions of China from 2015 to 2019: Policy validity, disaster threat, health risk, and economic loss","volume":"14","author":"Mo","year":"2021","journal-title":"Air Qual. Atmos. Health"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"105741","DOI":"10.1016\/j.scs.2024.105741","article-title":"A bivariate simultaneous pollutant forecasting approach by Unified Spectro-Spatial Graph Neural Network (USSGNN) and its application in prediction of O3 and NO2 for New Delhi, India","volume":"114","author":"Mandal","year":"2024","journal-title":"Sustain. Cities Soc."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"125071","DOI":"10.1016\/j.envpol.2024.125071","article-title":"Predicting plateau atmospheric ozone concentrations by a machine learning approach: A case study of a typical city on the southwestern plateau of China","volume":"363","author":"Wang","year":"2024","journal-title":"Environ. Pollut."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"174229","DOI":"10.1016\/j.scitotenv.2024.174229","article-title":"A deep learning model integrating a wind direction-based dynamic graph network for ozone prediction","volume":"946","author":"Wang","year":"2024","journal-title":"Sci. Total Environ."},{"key":"ref_13","first-page":"25","article-title":"Prediction of PM10, SO2, NO2, O3, and CO Concentrations in Guadalajara Using ARIMA and Open Data with Python","volume":"16","author":"Silva","year":"2025","journal-title":"Int. J. Comb. Optim. Probl. Inform."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"123183","DOI":"10.1016\/j.envpol.2023.123183","article-title":"A review of the CAMx, CMAQ, WRF-Chem and NAQPMS models: Application, evaluation and uncertainty factors","volume":"343","author":"Gao","year":"2024","journal-title":"Environ. Pollut."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"6241","DOI":"10.1016\/j.atmosenv.2011.06.071","article-title":"Application of WRF\/Chem-MADRID for real-time air quality forecasting over the Southeastern United States","volume":"45","author":"Chuang","year":"2011","journal-title":"Atmos. Environ."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"3843","DOI":"10.1007\/s11869-025-01810-8","article-title":"Sensitivity analysis of NO2 and O3 concentrations modeled with WRF-CMAQ to boundary and initial conditions and first model layer height in the Metropolitan Area of Buenos Aires, Argentina","volume":"18","author":"Luque","year":"2025","journal-title":"Air Qual. Atmos. Health"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"102862","DOI":"10.1016\/j.apr.2025.102862","article-title":"Attribution of a Typical Ozone Pollution Episode in Handan: WRF-CMAQ modeling and Process Analysis Based on a Local Emission Inventory","volume":"16","author":"Fang","year":"2025","journal-title":"Atmos. Pollut. Res."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"102733","DOI":"10.1016\/j.uclim.2025.102733","article-title":"Application of WRF-Chem for predicting air quality resulting from the formation of photochemical compounds in a subtropical urban environment","volume":"65","author":"Deroubaix","year":"2026","journal-title":"Urban Clim."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"105599","DOI":"10.1016\/j.atmosres.2021.105599","article-title":"A comprehensive investigation of surface ozone pollution in China, 2015\u20132019: Separating the contributions from meteorology and precursor emissions","volume":"257","author":"Mousavinezhad","year":"2021","journal-title":"Atmos. Res."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"3","DOI":"10.3934\/environsci.2024020","article-title":"Multi-step ahead ozone level forecasting using a component-based technique: A case study in Lima, Peru","volume":"11","author":"Quispe","year":"2024","journal-title":"AIMS Environ. Sci."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"1349","DOI":"10.56028\/aetr.14.1.1349.2025","article-title":"Comparative Analysis of ARMA and ARIMA Models for Air Quality Prediction","volume":"14","author":"Yan","year":"2025","journal-title":"Adv. Eng. Technol. Res."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"4331","DOI":"10.1007\/s00521-018-3345-0","article-title":"Predicting hourly ozone concentrations using wavelets and ARIMA models","volume":"31","author":"Salazar","year":"2019","journal-title":"Neural Comput. Appl."},{"key":"ref_23","first-page":"106","article-title":"Different approaches of multiple linear regression (MLR) model in predicting ozone (O3) concentration in industrial area","volume":"15","author":"Napi","year":"2023","journal-title":"Int. J. Integr. Eng."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"114459","DOI":"10.1016\/j.rse.2024.114459","article-title":"Two-decade surface ozone (O3) pollution in China: Enhanced fine-scale estimations and environmental health implications","volume":"317","author":"Yang","year":"2025","journal-title":"Remote Sens. Environ."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"015006","DOI":"10.1088\/2632-2153\/ae2f8b","article-title":"GNADET: A geospatial neural advection\u2013diffusion equation framework with graph transformer for surface temperature forecasting","volume":"7","author":"Lu","year":"2026","journal-title":"Mach. Learn. Sci. Technol."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"2183","DOI":"10.1007\/s40808-021-01220-6","article-title":"Prediction of tropospheric ozone using artificial neural network (ANN) and feature selection techniques","volume":"8","author":"Kapadia","year":"2022","journal-title":"Model. Earth Syst. Environ."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"51","DOI":"10.1016\/j.apr.2020.02.024","article-title":"Prediction of ozone hourly concentrations by support vector machine and kernel extreme learning machine using wavelet transformation and partial least squares methods","volume":"11","author":"Su","year":"2020","journal-title":"Atmos. Pollut. Res."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"120145","DOI":"10.1016\/j.atmosenv.2023.120145","article-title":"Using machine learning to improve the estimate of US background ozone","volume":"316","author":"Hosseinpour","year":"2024","journal-title":"Atmos. Environ."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"13517","DOI":"10.1007\/s10462-023-10466-8","article-title":"Deep learning modelling techniques: Current progress, applications, advantages, and challenges","volume":"56","author":"Ahmed","year":"2023","journal-title":"Artif. Intell. Rev."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"8783","DOI":"10.1007\/s00521-019-04282-x","article-title":"A real-time hourly ozone prediction system using deep convolutional neural network","volume":"32","author":"Eslami","year":"2020","journal-title":"Neural Comput. Appl."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"120987","DOI":"10.1016\/j.atmosenv.2024.120987","article-title":"Temporal CNN-based 72-h ozone forecasting in South Korea: Explainability and uncertainty quantification","volume":"343","author":"Salman","year":"2025","journal-title":"Atmos. Environ."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"106048","DOI":"10.1016\/j.envsoft.2024.106048","article-title":"Transfer learning in environmental data-driven models: A study of ozone forecast in the Alpine region","volume":"177","author":"Sangiorgio","year":"2024","journal-title":"Environ. Model. Softw."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"3977","DOI":"10.1016\/j.egyr.2025.11.016","article-title":"A novel point-interval prediction model for wind speed based on hybrid deep learning and RIME optimization algorithms","volume":"14","author":"Huang","year":"2025","journal-title":"Energy Rep."},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Zhang, D., Hou, Y., Zhao, B., and Wang, X. (2024, January 13\u201316). A prediction method for ozone concentration based on GCN-BiLSTM-Attention. Proceedings of the 2024 10th International Conference on Computer and Communications (ICCC), Chengdu, China.","DOI":"10.1109\/ICCC62609.2024.10942043"},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"121883","DOI":"10.1016\/j.jenvman.2024.121883","article-title":"Transformer-based ozone multivariate prediction considering interpretable and priori knowledge: A case study of Beijing, China","volume":"366","author":"Mu","year":"2024","journal-title":"J. Environ. Manag."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"142541","DOI":"10.1016\/j.jclepro.2024.142541","article-title":"Learning spatiotemporal dependencies using adaptive hierarchical graph convolutional neural network for air quality prediction","volume":"459","author":"Hu","year":"2024","journal-title":"J. Clean. Prod."},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Xu, J., Wang, S., Ying, N., Xiao, X., Zhang, J., Jin, Z., Cheng, Y., and Zhang, G. (2023). Dynamic graph neural network with adaptive edge attributes for air quality predictions. arXiv.","DOI":"10.1016\/j.heliyon.2023.e17746"},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"4408","DOI":"10.1038\/s41598-024-55060-2","article-title":"An adaptive adjacency matrix-based graph convolutional recurrent network for air quality prediction","volume":"14","author":"Chen","year":"2024","journal-title":"Sci. Rep."},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"4349","DOI":"10.5194\/essd-13-4349-2021","article-title":"ERA5-Land: A state-of-the-art global reanalysis dataset for land applications","volume":"13","author":"Dutra","year":"2021","journal-title":"Earth Syst. Sci. Data"},{"key":"ref_40","unstructured":"Cover, T.M., and Thomas, J.A. (2006). Elements of Information Theory, John Wiley & Sons. [2nd ed.]."},{"key":"ref_41","doi-asserted-by":"crossref","unstructured":"Goovaerts, P. (1997). Geostatistics for Natural Resources Evaluation, Oxford University Press.","DOI":"10.1093\/oso\/9780195115383.001.0001"},{"key":"ref_42","unstructured":"Veli\u010dkovi\u0107, P., Cucurull, G., Casanova, A., Romero, A., Li\u00f2, P., and Bengio, Y. (May, January 30). Graph attention networks. Proceedings of the International Conference on Learning Representations (ICLR), Vancouver, BC, Canada."},{"key":"ref_43","unstructured":"Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, \u0141., and Polosukhin, I. (2017, January 4\u20139). Attention is all you need. Proceedings of the 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA, Long Beach, CA, USA."},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"311","DOI":"10.1007\/s12145-025-01798-w","article-title":"Stabilized Long Short Term Memory (SLSTM) model: A new variant of the LSTM model for predicting ozone concentration data","volume":"18","author":"Kafi","year":"2025","journal-title":"Earth Sci. Inform."},{"key":"ref_45","doi-asserted-by":"crossref","unstructured":"Cherukat, S., Niharika, G., Raghavendran, S., Bajpai, I.K., and Lalmohan, K.S. (2025). Comparative analysis of LSTM, GRU, and random forest regression for air quality prediction. NIELIT\u2019s International Conference on Communication, Electronics and Digital Technologies, Springer.","DOI":"10.1007\/978-981-96-9932-2_22"},{"key":"ref_46","first-page":"48","article-title":"The spatio-temporal prediction of ozone in Zhuhai based on graph convolutional memory network","volume":"63","author":"Sun","year":"2024","journal-title":"Acta Sci. Nat. Univ. SunYatseni"},{"key":"ref_47","doi-asserted-by":"crossref","unstructured":"Niresi, K.F., Zhao, M., Bissig, H., Baumann, H., and Fink, O. (November, January 29). Spatial-temporal graph attention fuser for calibration in IoT air pollution monitoring systems. Proceedings of the 2023 IEEE SENSORS, Vienna, Austria.","DOI":"10.1109\/SENSORS56945.2023.10325090"},{"key":"ref_48","doi-asserted-by":"crossref","first-page":"e43","DOI":"10.1017\/eds.2023.37","article-title":"Short-term forecasting of ozone air pollution across Europe with transformers","volume":"2","author":"Hickman","year":"2023","journal-title":"Environ. Data Sci."},{"key":"ref_49","doi-asserted-by":"crossref","first-page":"106559","DOI":"10.1016\/j.jastp.2025.106559","article-title":"Spatial and temporal prediction of ozone concentration in the Pearl River Delta region based on a dynamic graph convolutional network","volume":"273","author":"Yang","year":"2025","journal-title":"J. Atmos.-Sol.-Terr. Phys."}],"container-title":["ISPRS International Journal of Geo-Information"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2220-9964\/15\/3\/101\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,5]],"date-time":"2026-03-05T11:03:16Z","timestamp":1772708596000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2220-9964\/15\/3\/101"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,28]]},"references-count":49,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,3]]}},"alternative-id":["ijgi15030101"],"URL":"https:\/\/doi.org\/10.3390\/ijgi15030101","relation":{},"ISSN":["2220-9964"],"issn-type":[{"value":"2220-9964","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,28]]}}}