The use of experimental data and their uncertainty for assessing ozone photochemistry in the Eastern Iberian Peninsula

Observation-based methods are useful tools to explore the sensitivity of ozone concentrations to precursor controls. With the aim of assessing the ozone precursor sensitivity in two locations: Paterna (suburban) and Villar del Arzobispo (rural) of the Turia river basin in the east of Spain, the photochemical indicator O-3/NOy and the Extent-of-Reaction (EOR) parameter have been calculated from field measurements. In Paterna, the O-3/NOy ratio varied from 0 to 13 with an average value of 5.1 (SD 3.2), whereas the averaged value for the EOR was 0.43 (SD 0.14). In Villar del Arzobispo, the O-3/NOy ratio changed from 5 to 30 with a mean value of 13.6 (SD 4.7) and the EOR gave an averaged value of 0.72 (SD 0.11). The results show two different patterns of ozone production as a function of the location. The suburban area shows a VOC-sensitive regime whereas the rural one shows a transition regime close to NOx-sensitive conditions. No seasonal differences in these regimes are observed along the monitoring campaigns. Finally, an analysis of the influence of the measurement quality of NOy, NOx and O-3 on the uncertainty of the O-3/NOy ratio and the EOR was performed showing that the uncertainty of O-3/NOy is not dependent on either its value or the individual values of O-3 and NOy but just on the quality of O-3 and NOy measurements. The maximum uncertainty is 26% as long as the combined uncertainties of O-3 and NOy remain below the 7.5%. The case of the EOR is different and its uncertainty depends on both the value of the EOR parameter and the individual concentration values of NOy and NOx. The uncertainty of the EOR estimation can be very high (>200%) if the combined uncertainties of both NOy and NOx are high (>7.5%), or especially, if u(NOy) and u(NOx) differ considerably from each other (>3.5%). (C) 2012 Elsevier Ltd. All rights reserved.