A hybrid model for daily air quality index prediction and its performance in the face of impact effect of COVID-19 lockdown

Abstract

Accurate and dependable air quality forecasting is critical to environmental and human health. However, most methods usually aim to improve overall prediction accuracy but neglect the accuracy for unexpected incidents. In this study, a hybrid model was developed for air quality index (AQI) forecasting, and its performance during COVID-19 lockdown was analyzed. Specifically, the variational mode decomposition (VMD) was employed to decompose the original AQI sequence into some subsequences with the parameters optimized by the Whale optimization algorithm (WOA), and the residual sequence was further decomposed by the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN). On this basis, a deep learning method bidirectional long short-term memory coupled with added time filter layer and attention mechanism (TFA-BiLSTM) was employed to explore the latent dynamic characteristics of each subsequence. This WOA-VMD-CEEMDAN-TFA-BiLSTM hybrid model was used to forecast AQI values for four cities in China, and results verified that the accuracy of the hybrid model outperformed other proposed models, achieving R² values of 0.96-0.97. In addition, the improvement in MAE (34.71-49.65%) and RMSE (32.82-48.07%) were observed over single decomposition-based model. Notably, during the epidemic lockdown period, the hybrid model had significant superiority over other proposed models for AQI prediction.

Keywords: Air quality index; Attention mechanism; COVID-19; Long short-term memory; Two-stage decomposition.

Graphical abstract

Fig. 1

The overall structure of the hybrid model for AQI prediction.

Fig. 2

Error evaluation results of AQI prediction in the four cities in China.

Fig. 3

Prediction results of AQI in the four cities in China. (The shaded areas represent the studied COVID-19 lockdown periods.).

Fig. 4

Scatterplots of the (a) BP, (b) BiLSTM, (c) TFA-BiLSTM modelling results, and AQI measurements in the four cities in China.

Fig. 5

Comparison of models performance between the COVID-19 lockdown period and the whole test period. (M1: WOA-VMD-BP, M2: WOA-VMD-BiLSTM, M3: WOA-VMD-TFA-BiLSTM, M4: WOA-VMD-CEEMDAN-BP, 5: WOA-VMD-CEEMDAN -BiLSTM, M6: WOA-VMD-CEEMDAN-TFA-BiLSTM).

Fig. 6

The correlation coefficients between AQI and air pollutant factors in the four cities in China.

Fig. 7

AQI data distribution during the COVID-19 lockdown period in 2020 and the corresponding periods in previous two years (the line inside the violins represents quartile.).

Fig. 8

HLN test results of predicted AQI values during COVID-19 lockdown period for the four cities in China. ($Z_{0.01/2}$ = 2.58, $Z_{0.15/2}$ = 1.44, M1: WOA-VMD-BP, M2: WOA-VMD-LSTM, M3: WOA-VMD-TFA-BiLSTM, M4: WOA-VMD-CEEMDAN-BP, M5: WOA-VMD-CEEMDAN-BiLSTM).