The air quality index trend forecasting based on improved error correction model and data preprocessing for 17 port cities in China

• Published in: Chemosphere (Oxford) Vol. 252; p. 126474
• Main Authors: Zhu, Suling, Sun, Jianan, Liu, Yafei, Lu, Mingming, Liu, Xingrong
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.08.2020
• Subjects: Air quality index forecasting; Error correction model; Statistical learning model; Statistical learning model; Error correction model; Air quality index forecasting
• ISSN: 0045-6535; 1879-1298
• DOI: 10.1016/j.chemosphere.2020.126474

Air pollution are known to have negative impacts on human health and the ecosystem, and it also contributes to climate change. Hence, prevention and control of air pollution is an urgent need in China, and air pollution prediction can provide reliable information for this process. Therefore, it is essential to establish effective air pollution prediction with an early warning model. Currently, one widely used air pollution prediction technology is the error correction model. However, this traditional method does not use data preprocessing technology. Therefoere, this paper presents an improved hybrid model named CEEMD-SLM-ECM (Complementary Set Empirical Mode Decomposition-Statistical Learning Model-Error Correction Model), which used the CEEMD data preprocessing technology together with statistical learning models. Furthermore, selected AQI (air quality index) data of 17 port cities in the 21st Century Maritime Silk Road Economic Belt were selected to test the forecasting ability of the proposed model. Data analysis shows that the CEEMD-SLM-ECM model has much higher accuracy compared with the traditional error correction model. So, the CEEMD-SLM-ECM is a very effective predictive model that can provide accurate prediction for air quality early warning. •A new improved error correction model CEEMD-SLM-ECM is established.•Data preprocessing is helpful for improving the prediction accuracy.•CEEMD-SLM-ECM is effective for AQI forecasting.