HIGH-LATITUDE SPRINGTIME PHOTOCHEMISTRY .1. NOX, PAN AND OZONE RELATIONSHIPS

Springtime measurements of NOx, ozone, PAN, J(NO2), and other compound s were made near Ny-Alesund, Svalbard (78 degrees 54'N, 11 degrees 53' E), in 1994 and Poker Flat, Alaska (65 degrees 08'N, 147 degrees 29'W) , in 1995. At Svalbard median mixing ratios for PAN and NOx of 237 and 23.7 pptv, respectively, were observed. The median mixing ratios at P oker Flat for PAN and NOx were 79.5 and 85.9 pptv, respectively. These data are used to estimate thermal PAN decomposition using several dif ferent approaches. At Svalbard PAN decomposition was very small, while at Poker Flat up to 30 pptv/h PAN decomposed. At both sites the NOx/P AN ratio increased with temperature between-10 and 20 degrees C implyi ng that PAN decomposition is an important NOx source. In-situ ozone pr oduction was calculated from the measured NO, NO2, O-3, J(NO2), and te mperature data, using the steady state assumption Median ozone product ion was 605 pptv/h at Poker Flat, and one order of magnitude smaller a t Svalbard during the daytime. Only at Poker Flat could a direct influ ence on the diurnal ozone cycle be observed from in-situ production. T hese results imply that PAN decomposition is a major source of NOx in the high latitude troposphere, and that this contributes to the observ ed spring maximum in surface ozone.