Spatio‐Temporal Hourly and Daily Ozone Forecasting in China Using a Hybrid Machine Learning Model: Autoencoder and Generative Adversarial Networks. (28th February 2022)

• Record Type: Journal Article
• Authors: Cheng, Meiling; Fang, Fangxin; Navon, Ionel M.; Zheng, Jie; Tang, Xiao; Zhu, Jiang; Pain, Christopher
• Is Part Of: Journal of advances in modeling earth systems. Volume 14:Number 3(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-02-28
• Subjects: ozone forecasting -- spatio‐temporal -- generative adversarial network -- variational autoencoder -- deep learning; Geological modeling -- Periodicals; Climatology -- Periodicals; Geochemical modeling -- Periodicals; 551.5011
• DOI: 10.1029/2021MS002806
• Languages: English
• ISSNs: 1942-2466

Abstract

Abstract: Efficient and accurate real‐time forecasting of national spatial ozone distribution is critical to the provision of effective early warning. Traditional numerical air quality models require a high computational cost associated with running large‐scale numerical simulations. In this work, we introduce a hybrid model (VAE‐GAN) combining a generative adversarial network (GAN) with a variational autoencoder (VAE) to learn the dynamic ozone distributions in spatial and temporal spaces. The VAE‐GAN model can not only decipher the complex nonlinear relationship between the inputs (the past states/ozone and meteorological factors) and outputs (ozone), but also provide ozone forecasts for a long lead‐time beyond the training period. The performance of VAE‐GAN is demonstrated in hourly and daily spatio‐temporal ozone forecasts over China. The training datasets from 2013 to 2017 and validation datasets from 2018 to 2019 are the collection of data from the air quality reanalysis datasets. With the use of VAE, large dataset sizes are decreased by three orders of magnitude, enabling hourly and daily forecasts to be computed in seconds. Results show that the VAE‐GAN achieves a reasonable accuracy in the prediction of both the spatial and temporal evolution patterns of hourly and daily ozone fields, as compared to the Nested Air Quality Prediction Modeling System (commonly used in China), the reanalysis data and observations during the validation period. Thus, the VAE‐GAN is a cost‐effective tool for large data‐driven predictions, which can potentially reinforce air pollution prediction efforts in providing risk assessment and management in a timely manner. Plain Language Summary: This work presents a hybrid machine learning model for hourly and daily ozone forecasting in spatial and temporal spaces. Our goal is to use the machine learning model for exploring the complex nonlinear relationship between the meteorological factors and ozone concentration, and to perform long lead‐time ozone forecasts accurately and efficiently. The reanalysis ozone datasets from 2013 to 2019 over China are used for processing different training and prediction scenarios. Our results show that the proposed machine learning model can predict the spatio‐temporal evolution patterns of the hourly and daily ozone concentration accurately and much more efficiently in comparison to the Nested Air Quality Prediction Modeling System. Such a data‐driven model is promising in different applications, for example, providing early warnings of high ozone concentrations in densely populated areas. Key Points: A hybrid deep learning model is proposed for efficient and accurate real‐time ozone forecasting with a high spatial resolution The VAE‐GAN enables hourly and daily forecast to be achieved in seconds while the same level of accuracy is reached compared to Nested Air Quality Prediction Modelling System The performance of the VAE‐GAN model is evaluated in hourly and daily ozone forecasts over China from the year 2018–2019