Servicios Personalizados
Revista
Articulo
Indicadores
Citado por SciELO
Accesos
Links relacionados
Similares en SciELO
Compartir
Revista Chapingo serie ciencias forestales y del ambiente
versión On-line ISSN 2007-4018versión impresa ISSN 2007-3828
Resumen
CRUZ-HUERTA, Carmina et al. Spatial and temporal modeling of air pollution in Mexico City Metropolitan Area. Rev. Chapingo ser. cienc. for. ambient [online]. 2024, vol.30, n.1, rrchscfa202302010. Epub 03-Dic-2024. ISSN 2007-4018. https://doi.org/10.5154/r.rchscfa.2023.02.010.
Introduction:
Large cities have air pollution problems due to the emission of polluting gases and particulate matter (PM).
Objectives:
To know the intra- and inter-annual variation of pollutants (NOX, CO, O3, PM10 and PM2.5) in Mexico City Metropolitan Area and to model their spatial distribution.
Materials and methods:
Data from 44 stations of the Automatic Air Monitoring Network (RAMA) were analyzed to extract information for the pollutants NOX, O3 and CO in the period 1986-2021, and PM2.5 and PM10 in the periods 2000-2021 and 2003-2021, respectively. Monthly averages per station were calculated and the temporal trend of each pollutant was evaluated using the 'Theil-Sen' operator. The spatial distribution of pollutants was also modeled and the statistical performance of four interpolation methods was compared: Neural Networks, Support Vector Machine, Random Forest and Kriging Universal.
Results and discussion:
NOX and CO concentrations were high from November to January, while O3 from April to May. The lowest concentrations of PM10 and PM2.5 took place from July to October and the highest in May. All pollutants decreased in concentration during the period analyzed, with the most noticeable changes in NOX (-1.28 ppb·yr-1), while CO had the smallest change (-0.12 ppm·yr-1). The maximum values for NOX, O3 and CO occurred in 1993 and for PM in 2003. The best model was Support Vector Machine, regardless of the pollutant analyzed.
Conclusion:
Spatio-temporal dynamics varied among air pollutants. The analysis with spatial interpolation methods is viable and favors solution strategies to pollution problems.
Palabras llave : carbon monoxide; nitrogen oxides; ozone; particulate matter; Machine Learning..