Development of a deep neural network for predicting 6-hour average PM_2.5 concentrations up to two subsequent days using various training data

This study aims to develop a deep neural network (DNN) model as an artificial neural network (ANN) for the prediction of 6-hour average fine particulate matter (PM2.5) concentrations for a three-day period—the day of prediction (D+0), one day after prediction (D+1) and two days after prediction (D+2)—using observation data and forecast data obtained via numerical models. The performance of the DNN model was comparatively evaluated against that of the currently operational Community Multiscale Air Quality (CMAQ) modelling system for air quality forecasting in South Korea. In addition, the effect on predictive performance of the DNN model on using different training data was analyzed. For the D+0 forecast, the DNN model performance was superior to that of the CMAQ model, and there was no significant dependence on the training data. For the D+1 and D+2 forecasts, the DNN model that used the observation and forecast data (DNN-ALL) outperformed the CMAQ model. The root-mean-squared error (RMSE) of DNN-ALL was lower than that of the CMAQ model by 2.2 μgm−3, and 3.0 μgm−3 for the D+1 and D+2 forecasts, respectively, because the overprediction of higher concentrations was curtailed. An IOA increase of 0.46 for D+1 prediction and 0.59 for the D+2 prediction was observed in case of the DNN-ALL model compared to the IOA of the DNN model that used only observation data (DNN-OBS). In additionally, An RMSE decrease of 7.2 μgm−3 for the D+1 prediction and 6.3 μgm−3 for the D+2 prediction was observed in case of the DNN-ALL model, compared to the RMSE of DNN-OBS, indicating that the inclusion of forecast data in the training data greatly affected the DNN model performance. Considering the prediction of the 6-hour average PM2.5 concentration, the 8.8 μgm−3 RMSE of the DNN-ALL model was 2.7 μgm−3 lower than that of the CMAQ model, indicating the superior prediction performance of the former. These results suggest that the DNN model could be utilized as a better-performing air quality forecasting model than the CMAQ, and that observation data plays an important role in determining the prediction performance of the DNN model for D+0 forecasting, while prediction data does the same for D+1 and D+2 forecasting. The use of the proposed DNN model as a forecasting model may result in a reduction in the economic losses caused by pollution-mitigation policies and aid better protection of public health.

Clustering diurnal cycles of day-to-day temperature change to understand their impacts on air quality forecasting in mountain-basin areas

Air pollution is substantially modulated by meteorological conditions, and especially their diurnal variations may play a key role in air quality evolution. However, the behaviors of temperature diurnal cycles along with the associated atmospheric condition and their effects on air quality in China remain poorly understood. Here, for the first time, we examine the diurnal cycles of day-to-day temperature change and reveal their impacts on winter air quality forecasting in mountain-basin areas. Three different diurnal cycles of the preceding day-to-day temperature change are identified and exhibit notably distinct effects on the day-to-day changes in atmospheric-dispersion conditions and air quality. The diurnal cycle with increasing temperature obviously enhances the atmospheric stability in the lower troposphere and suppresses the development of the planetary boundary layer, thus deteriorating the air quality on the following day. By contrast, the diurnal cycle with decreasing temperature in the morning is accompanied by a worse dispersion condition with more stable atmosphere stratification and weaker surface wind speed, thereby substantially worsening the air quality. Conversely, the diurnal cycle with decreasing temperature in the afternoon seems to improve air quality on the following day by enhancing the atmospheric-dispersion conditions on the following day. The findings reported here are critical to improve the understanding of air pollution in mountain-basin areas and exhibit promising potential for air quality forecasting.

Air Quality Forecasting with Hybrid LSTM and Extended Stationary Wavelet Transform

Artificial intelligence (AI) technology-enhanced air quality forecasting is one of the most promising directions in the field of smart environment development. Despite recent advances in this area, two difficulties remain unsolved. First, multiple factors influence forecasting results, such as weather conditions, fuel usage and traffic conditions. These factors are usually unavailable in air quality sensor data. Second, traditional predicting models typically use the most recent training data, which neglects the historical data. In this study, we propose a hybrid deep learning model that embraces the merits of the stationary wavelet transform (SWT) and the nested long short term memory networks (NLSTM) to improve the prediction quality in the problem of hour-ahead air quality forecasting. The proposed method decomposes the original PM2.5 data into several more stationary sub-signals with different resolutions using an extended SWT algorithm. A framework that leverages several NLSTM recurrent neural networks is constructed to output forecasting results for different sub-signals, respectively. The final forecasting result is obtained by combining all sub-signal forecasting results using the inverse wavelet transform. Experiments on real-world data show that, accuracy-wise, our proposed method outperforms most of the existing prediction models in the literature. And the resulting forecasting curves of the proposed method are much closer to the real values without any lags, comparing with existing prediction models.