Climatology and forecast modeling of ambient carbon monoxide in Phoenix, Arizona

摘要: We perform a climatology of factors influencing ambient carbon monoxide (CO), in which we examine the relationships between meteorology, traffic patterns, and CO at seasonal, weekly, diurnal time scales Phoenix, Arizona. From this analysis identify range potentially important variables for statistical modeling. Using stepwise multivariate regression, create suite models hourly 8-h designed daily operational forecasting purposes. The resulting include interaction terms related to anticipated nocturnal atmospheric stability as well antecedent climatological behavior. are evaluated using error statistics skill measures. most successful approach employs two-stage modeling strategy an initial prediction is made that may, depending on forecast value, be followed by second improves upon first. best provide accurate forecasts CO, with explained variances approaching 0.9 errors under 1 ppm.