Abstrakt
The aim of our study was to identify the factors substantially affecting day-to-day variability in O3 concentrations in Czech mountain forests and to describe their influence in detailed, quantitative way. We examined the effects of meteorology and ambient NOx recorded in regular long-term continuous monitoring at five mountain forest sites representing different regions, covering both polluted and relatively unpolluted areas over the time period of 1992-2018.
To investigate the association between ambient O3 concentrations on one hand, and precursor NOx concentrations, and meteorology on the other hand, we used a generalized additive model, GAM, with semiparametric (penalized-spline-based) components to capture properly the possible departures from linearity that is not captured by traditional linear regression approaches. Our results revealed that the O3 concentrations showed significant associations with all selected explanatory variables, i.e., air temperature, global solar radiation (GLRD), relative humidity, and NOx.
Apparently, both meteorology and air pollution are highly important for day-to-day O3 concentrations, and this finding is consistent for all five rural sites, representing middle-elevated forested mountain areas in Central Europe. In addition to individual variables, we were able to detect interactions between three pairs of explanatory variables, namely temperature*GLRD, temperature*relative humidity, and GLRD*relative humidity.
Moreover, we confirmed non-linear O3 behavior toward all individual explanatory variables.