Spatial and temporal prediction of ozone concentration in the Pearl River Delta region based on a dynamic graph convolutional network

IF 1.8 4区地球科学 Q3 GEOCHEMISTRY & GEOPHYSICS

Journal of Atmospheric and Solar-Terrestrial Physics Pub Date : 2025-05-30 DOI:10.1016/j.jastp.2025.106559

Tongshu Yang, Sheng Li, Baoqin Chen

{"title":"Spatial and temporal prediction of ozone concentration in the Pearl River Delta region based on a dynamic graph convolutional network","authors":"Tongshu Yang, Sheng Li, Baoqin Chen","doi":"10.1016/j.jastp.2025.106559","DOIUrl":null,"url":null,"abstract":"<div><div>The variation of ozone (O<sub>3</sub>) concentration is closely related to other meteorological factors such as temperature and wind speed, and there is significant dynamic uncertainty, making related research very complex and difficult. This paper will establish a time-space ozone prediction model based on dynamic graph convolution network to study the O<sub>3</sub> pollution in the Pearl River Delta (PRD) region of China. Firstly, use an isolated forest (iForest) for anomaly detection in data preprocessing. Secondly, based on data such as wind direction, wind speed, and station geographic location, establish the diffusion distance of the wind field and construct a dynamic graph sequence accordingly. Finally, a spatio-temporal dynamic graph convolutional network (STD-GCN) based on dynamic graph sequences was established for predicting O<sub>3</sub> concentration. The experimental results showed that STD-GCN outperformed long short-term memory (LSTM) and graph convolutional embedded LSTM (GC-LSTM). Specifically, by integrating wind field factors, STD-GCN exhibits better spatial interpretability.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"273 ","pages":"Article 106559"},"PeriodicalIF":1.8000,"publicationDate":"2025-05-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Atmospheric and Solar-Terrestrial Physics","FirstCategoryId":"89","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1364682625001439","RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"GEOCHEMISTRY & GEOPHYSICS","Score":null,"Total":0}

引用次数: 0

Abstract

The variation of ozone (O₃) concentration is closely related to other meteorological factors such as temperature and wind speed, and there is significant dynamic uncertainty, making related research very complex and difficult. This paper will establish a time-space ozone prediction model based on dynamic graph convolution network to study the O₃ pollution in the Pearl River Delta (PRD) region of China. Firstly, use an isolated forest (iForest) for anomaly detection in data preprocessing. Secondly, based on data such as wind direction, wind speed, and station geographic location, establish the diffusion distance of the wind field and construct a dynamic graph sequence accordingly. Finally, a spatio-temporal dynamic graph convolutional network (STD-GCN) based on dynamic graph sequences was established for predicting O₃ concentration. The experimental results showed that STD-GCN outperformed long short-term memory (LSTM) and graph convolutional embedded LSTM (GC-LSTM). Specifically, by integrating wind field factors, STD-GCN exhibits better spatial interpretability.

查看原文本刊更多论文

基于动态图卷积网络的珠江三角洲臭氧浓度时空预测

臭氧（O3）浓度的变化与气温、风速等其他气象因子密切相关，且存在显著的动态不确定性，使得相关研究非常复杂和困难。本文将建立基于动态图卷积网络的臭氧时空预测模型，对中国珠江三角洲地区的O3污染进行研究。首先，在数据预处理中使用孤立森林（ifforest）进行异常检测。其次，根据风向、风速、站点地理位置等数据，建立风场扩散距离，构建动态图序列；最后，建立了基于动态图序列的时空动态图卷积网络（STD-GCN）预测臭氧浓度。实验结果表明，STD-GCN优于长短期记忆（LSTM）和图卷积嵌入LSTM （GC-LSTM）。具体而言，通过整合风场因子，STD-GCN具有更好的空间可解释性。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Journal of Atmospheric and Solar-Terrestrial Physics 地学-地球化学与地球物理

CiteScore

4.10

自引率

5.30%

发文量

审稿时长

6 months

期刊介绍： The Journal of Atmospheric and Solar-Terrestrial Physics (JASTP) is an international journal concerned with the inter-disciplinary science of the Earth''s atmospheric and space environment, especially the highly varied and highly variable physical phenomena that occur in this natural laboratory and the processes that couple them. The journal covers the physical processes operating in the troposphere, stratosphere, mesosphere, thermosphere, ionosphere, magnetosphere, the Sun, interplanetary medium, and heliosphere. Phenomena occurring in other "spheres", solar influences on climate, and supporting laboratory measurements are also considered. The journal deals especially with the coupling between the different regions. Solar flares, coronal mass ejections, and other energetic events on the Sun create interesting and important perturbations in the near-Earth space environment. The physics of such "space weather" is central to the Journal of Atmospheric and Solar-Terrestrial Physics and the journal welcomes papers that lead in the direction of a predictive understanding of the coupled system. Regarding the upper atmosphere, the subjects of aeronomy, geomagnetism and geoelectricity, auroral phenomena, radio wave propagation, and plasma instabilities, are examples within the broad field of solar-terrestrial physics which emphasise the energy exchange between the solar wind, the magnetospheric and ionospheric plasmas, and the neutral gas. In the lower atmosphere, topics covered range from mesoscale to global scale dynamics, to atmospheric electricity, lightning and its effects, and to anthropogenic changes.