CLIMATOLOGIES OF NOX AND NOY - A COMPARISON OF DATA AND MODELS

Climatologies of tropospheric NOx(NO + NO2) and NOy (total reactive ni trogen: NOx + NO3 + 2 x N2O5 + HNO2 + HNO3 + HNO4 + ClONO2 + PAN (pero xyacetylnitrate) + other organic nitrates) have been compiled from dat a previously published and, in most cases, publicly archived. Emphasis has been on non-urban measurements, including rural and remote ground sites, as well as aircraft data. Although the distribution of data is sparse, a compilation in this manner can begin to provide an understa nding of the spatial and temporal distributions of these reactive nitr ogen species. The cleanest measurements in the boundary layer are in A laska, northern Canada and the eastern Pacific, with median NO mixing ratios below 10 pptv, NOx below 50 pptv, and NOy below 300 pptv. The h ighest NO values (greater than 1 ppbv) were found in eastern North Ame rica and Europe, with correspondingly high NOy (similar to 5 ppbv). A significantly narrower range of concentrations is seen in the free tro posphere, particularly at 3-6 km, with NO typically about 10 pptv in t he boreal summer. NO increases with altitude to similar to 100 pptv at 9-12 km, whereas NOy does not show a trend with altitude, but varies between 100 and 1000 pptv. Decreasing mixing ratios eastward of the As ian and North American continents are seen in all three species at all altitudes. Model-generated climatologies of NOx and NOy from six chem ical transport models are also presented and are compared with observa tions in the boundary layer and the middle troposphere for summer and winter. These comparisons test our understanding of the chemical and t ransport processes responsible for these species distributions. Althou gh the model results show differences between them, and disagreement w ith observations, none are systematically different for all seasons an d altitudes. Some of the differences between the observations and mode l results may likely be attributed to the specific meteorological cond itions at the time that measurements were made differing from the mode l meteorology, which is either climatological how from GCMs or actual meteorology for an arbitrary year. Differences in emission inventories , and convection and washout schemes in the models will also affect th e calculated NOx and NOy distributions. (C) 1997 Elsevier Science Ltd.