Resumen

RAMOS-HERRERA, S; BAUTISTA-MARGULIS, R  y  VALDEZ-MANZANILLA, A. Statistical study of the correlation between atmospheric pollutants and meteorological variables in northern Chiapas, Mexico. Universidad y ciencia [online]. 2010, vol.26, n.1, pp.65-80. ISSN 0186-2979.

The multiple regression analysis is a statistical method used in many academic areas. In this study, this analysis was applied to the concentration data of four atmospheric pollutants (SO₂, NO₂, H₂S y PM₁₀), monitored in three stations located in northern Chiapas. The study was carried out from January 2001 to February 2005. The purpose of the study was to propose regression functions to describe the concentration as a function of time and/or meteorological variables. A stepwise multiple linear regression analysis was used in the selection of regressive variables, the most important of which included the temperature, relative humidity and wind direction. Regression functions of the annual, monthly and daily concentrations of these pollutants were obtained. A simple linear regression was obtained to explain the annual concentration of SO₂, with a determination coefficient of 0.927. The models for the monthly concentration obtained determination coefficients of 0.417 to 0.846,whereas the coefficients for the daily concentration model varied from 0.285 to 0.581. This parametric study made it possible to conclude that the meteorological variables described the annual and monthly concentrations adequately, but not the daily concentrations.

Palabras llave : Statistic correlation; linear regression; atmospheric pollutants.