A new optimized hybrid approach combining machine learning with WRF-CHIMERE model for PM10 concentration prediction

被引:2
作者
Chelhaoui, Youssef [1 ]
El Ass, Khalid [1 ]
Lachatre, Mathieu [3 ]
Bouakline, Oumaima [2 ]
Khomsi, Kenza [2 ]
El Moussaoui, Tawfik [1 ]
Arrad, Mouad [1 ]
Eddaif, Abdelhamid [1 ]
Albergel, Armand [3 ]
机构
[1] Natl Sch Mines Rabat ENSMR, Ave Hadj Ahmed Cherkaoui,BP 753, Rabat, Morocco
[2] Gen Directorate Meteorol DGM, Casablanca, Morocco
[3] ARIA Technol, Boulogne Billancourt, France
关键词
Air quality; PM10; forecasts; Machine Learning; WRF-CHIMERE; MULTIPLE LINEAR-REGRESSION; PARTICULATE MATTER; EMISSIONS; AEROSOLS; DUST; CHEMISTRY; TRANSPORT; FORECAST; PATTERNS; IMPACTS;
D O I
10.1007/s40808-024-02086-0
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
The forecast of particulate matter PM10 concentration is crucial due to its impacts on public health and the environment. Chemical Transport Models (CTM) are used to predict air quality. However, these models are subject to bias because of the precision of inputs. This paper explores a hybrid approach combining CTM (WRF-CHIMERE) predictions with machine learning (ML) to forecast PM10 concentrations. Five ML algorithms were developed: Multiple Linear Regression (MLR), Random Forests (RF), Extreme Gradient Boosting (XGB), Support Vector Regression (SVR), and Artificial Neural Networks (ANN). This hybrid system was trained using hourly data from September to December 2020 on seven Moroccan sites, incorporating nine parameters including meteorological variables, chemical concentrations, and other spatiotemporal variables. The hybrid model was evaluated against PM10 measurements. The results reveal that CHIMERE combined with RF and with XGB presented the best accuracy of predictions of PM10, when compared to the CHIMERE model. These two hybrid models achieved high correlation coefficients of 0.756 and 0.747, and determination coefficients of 57% and 55.7%, respectively. Moreover, they reduced the mean squared error, with CHIMERE-RF decreasing from 42.39 to 21.72 mu g/m3 and CHIMERE-XGB from 42.39 to 22.05 mu g/m3. Additionally, there was improvement in the mean bias, with CHIMERE-RF changing from -24.40 to -1.324 mu g/m3 and CHIMERE-XGB from -24.40 to -0.890 mu g/m3. The significance of these results could be important for air quality monitoring during extreme dust events, as they provide crucial information and simplify the implementation of preventive measures. This would help minimize the health risks associated with PM10.
引用
收藏
页码:5687 / 5701
页数:15
相关论文
共 62 条
  • [51] Development of a Dust Source Map for WRF-Chem Model Based on MODIS NDVI
    Spyrou, Christos
    Solomos, Stavros
    Bartsotas, Nikolaos S.
    Douvis, Kostas C.
    Nickovic, Slobodan
    [J]. ATMOSPHERE, 2022, 13 (06)
  • [52] Steinwart I., 2008, SUPPORT VECTOR MACHI
  • [53] Tamas WT, 2015, THESIS U PASCAL PAOL
  • [54] APIFLAME v2.0 biomass burning emissions model: impact of refined input parameters on atmospheric concentration in Portugal in summer 2016
    Turquety, Solene
    Menut, Laurent
    Siour, Guillaume
    Mailler, S.
    Hadji-Lazaro, Juliette
    George, Maya
    Clerbaux, Cathy
    Hurtmans, Daniel
    Coheur, Pierre-Francois
    [J]. GEOSCIENTIFIC MODEL DEVELOPMENT, 2020, 13 (07) : 2981 - 3009
  • [55] Valencia A.R.Z., 2024, Nat. Environ. Pollut. Technol, V23, P711, DOI [10.46488/NEPT.2024.v23i02.009, DOI 10.46488/NEPT.2024.v23i02.009]
  • [56] Hourly Particulate Matter (PM10) Concentration Forecast in Germany Using Extreme Gradient Boosting
    Wallek, Stefan
    Langner, Marcel
    Schubert, Sebastian
    Franke, Raphael
    Sauter, Tobias
    [J]. ATMOSPHERE, 2024, 15 (05)
  • [57] Weizhen H., 2014, IOP C SERIES EARTH E, V17
  • [58] World Health Organization, 2021, WHO Global Air Quality Guidelines: Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon Monoxide, DOI DOI 10.1016/J.HEALTHPLACE.2019.102206
  • [59] Machine learning based bias correction for numerical chemical transport models
    Xu, Min
    Jin, Jianbing
    Wang, Guoqiang
    Segers, Arjo
    Deng, Tuo
    Lin, Hai Xiang
    [J]. ATMOSPHERIC ENVIRONMENT, 2021, 248
  • [60] Chemical characterization of urban PM10 in the Tropical Andes
    Zalakeviciute, Rasa
    Rybarczyk, Yves
    Granda-Albuja, Maria G.
    Diaz Suarez, Maria Valeria
    Alexandrino, Katiuska
    [J]. ATMOSPHERIC POLLUTION RESEARCH, 2020, 11 (02) : 343 - 356