Abstract:With the rapid development of the economy, urban pollution has become a hot issue of human concern, and people are put forward higher requirements for the accuracy of pollutant concentration simulations. Based on the the WRF-Chem model simulation results and Beijing environmental monitoring data, this study constructed the XGBoost algorithm through the process of data cleaning, feature selection and super parameter optimization, and the optimized simulation of PM2.5 and O3 concentrations in Beijing was then carried out based on the constructed algorithm. The results showed that the XGBoost algorithm can better capture the spatial and temporal variation patterns of pollutant concentrations, and has a greater improvement on the simulation results of the WRF-Chem model, and that the XGBoost algorithm shows better optimisation results in urban areas compared to suburban areas. In addition, the overall analysis of the features of PM2.5 and O3 concentrations based on the SHAP value theory showed that the time series and periodic features, aerosol ion concentration of Sodium (NAAJ) and Nitrate concentration (NO3AJ) were the important features affecting the prediction of PM2.5 concentration by the XGBoost algorithm, and the most important factor affecting O3 concentration is temperature. PM2.5 and O3 concentrations were divided into three levels, and several samples were selected for single sample analysis. The analysis showed that at low pollutant concentration, most of the features made negative contributions to the concentration prediction, while at high concentration, most of the features made positive contributions. The contribution values of different features varied greatly and were unevenly distributed. The results of prediction were basically composed of a few features with large feature contributions, and the feature contributions of the same feature to different concentration prediction results were also different. Moreover, the XGBoost algorithm was used to optimize the concentration of pollutant at each grid point in Beijing, and a set of pollutant concentration data set with spatial resolution of 6 km and time resolution of 1 h covering the whole Beijing was established, and the optimized spatial distribution of pollutant concentration was closer to the spatial distribution of observed concentration than WRF-Chem simulation. At last, compared with SVR, LR, DTR and RF algorithms, XGBoost algorithm was better than other statistical algorithms in optimising PM2.5 and O3 concentrations. The results of this study provided a new idea for an in-depth analysis of optimization principle of algorithm model for air pollution and a quantitative study of the influencing factors.

Hybrid machine learning model for hourly ozone concentrations prediction and exposure risk assessment

Development of a High-Performance Machine Learning Model to Predict Ground Ozone Pollution in Typical Cities of China.

Application of the XGBoost Machine Learning Method in PM2.5 Prediction: A Case Study of Shanghai

Machine Learning‐Driven Spatiotemporal Analysis of Ozone Exposure and Health Risks in China

Ozone Concentration Estimation and Meteorological Impact Quantification in the Beijing‐Tianjin‐Hebei Region Based on Machine Learning Models

Spatiotemporal distributions of surface ozone levels in China from 2005 to 2017: A machine learning approach

A hybrid XGBoost-SMOTE model for optimization of operational air quality numerical model forecasts

[Atmospheric Ozone Concentration Prediction in Nanjing Based on LightGBM]

Prediction of Ozone Hourly Concentrations Based on Machine Learning Technology

Time series-based PM2.5 concentration prediction in Jing-Jin-Ji area using machine learning algorithm models

A novel framework for daily forecasting of ozone mass concentrations based on cycle reservoir with regular jumps neural networks

Prediction of atmospheric pollutants in urban environment based on coupled deep learning model and sensitivity analysis

Estimating ground-level high-resolution ozone concentration across China using a stacked machine-learning method

A machine learning model for estimating daily maximum 8-hour average ozone concentrations using OMI and MODIS products

Spatiotemporal Variations of Air Pollutants and Ozone Prediction Using Machine Learning Algorithms in the Beijing-Tianjin-Hebei Region from 2014 to 2021.

Application of Xgboost Algorithm in the Optimization of Pollutant Concentration

Machine learning and deep learning modeling and simulation for predicting PM2.5 concentrations

Development of high-resolution spatio-temporal models for ambient air pollution in a metropolitan area of China from 2013 to 2019

Prediction of daily PM 2.5 and ozone based on high-density weather stations in China: Nonlinear effects of meteorology, human and ecosystem health risks

Modeling spatial variation of gaseous air pollutants and particulate matters in a Metropolitan area using mobile monitoring data

Air Pollutant Concentration Forecast Based on Support Vector Regression and Quantum-Behaved Particle Swarm Optimization