Spatio‐Temporal Hourly and Daily Ozone Forecasting in China Using a Hybrid Machine Learning Model: Autoencoder and Generative Adversarial Networks

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.

Parsimonious estimation of hourly surface ozone concentration across China during 2015–2020

“…Correspondingly, all predictor variables (including satellite retrievals, climate, land use, population distribution, and GDP data) were aggregated or resampled to the targeted grid resolution of 0.1° × 0.1° using the nearest neighbor interpolation or the bilinear interpolation approach. To avoid high collinearity among predictor variables, we conducted variance inflation factor (VIF) tests to all the predictor variables and only those with a VIF value less than 8.0 were retained 30 (Supplementary Figure S3 ).

Fig.

…”

“…Cheng et al . 30 used a hybrid deep learning model to explore the complex nonlinear relationships between meteorological factors and ozone concentrations and applied it to hourly and daily forecasts of ozone concentrations in China.…”

Spatial and temporal characteristics analysis and prediction model of PM2.5 concentration based on SpatioTemporal-Informer model

“…The data are then normalized [33], which is treating each feature equally by removing the variability of values and units between feature parameters and normalizing the initial feature data to increase prediction accuracy. The normalization expression is as follows:…”

Revealing the Covariation of Atmospheric O2 and Pollutants in an Industrial Metropolis by Explainable Machine Learning

“…Machine learning (ML) models have been demonstrated to be a powerful tool for reconstructing, simulating, and predicting atmospheric pollution, including PM 2.5 , − O 3 , , NO x , , etc., outperforming finely designed chemical transport models . The use of ML models provides greater flexibility and efficiency when utilizing real-world data and is especially adept at revealing complex and hidden nonlinear correlations , that might not be easily identified using traditional physical models, providing new insights into the underlying mechanisms of the studied phenomena .…”