Is the Arctic surface layer a source and sink of NOx in winter/spring?

Measurements of NOx (NO + NO2) and the sum of reactive nitrogen constituents, NOy, were made near the surface at Alert (82.5 degrees N), Canada during March and April 1998. In early March when solar insolation was absent or very low, NOx mixing ratios were frequently near zero. After polar sunrise when the sun was above the horizon for much or all of the day a diurnal variation in NOx and NOy was observed with amplitudes as large as 30-40 pptv. The source of active nitrogen is attributed to release from the snow surface by a process that is apparently sensitized by sunlight. If the source from the snowpack is a large scale feature of the Arctic then the diurnal trends also require a competing process for removal to the surface. From the diurnal change in the NO/NO2 ratio, mid-April mixing ratios for the sum of peroxy and halogen oxide radicals of less than or equal to 10 pptv were derived for periods when ozone mixing ratios were in the normal range of 30-50 ppbv. Mid-day ozone production and loss rates with the active nitrogen source were estimated to be similar to 1-2 ppbv/day and in near balance. NOy mixing ratios which averaged only 295 +/- 66 pptv do not support a large accumulation in the high Arctic surface layer in the winter and spring of 1998. The small abundance of NOy relative to the elevated mixing ratios of other long-lived anthropogenic constituents requires that reactive nitrogen be removed to the surface during transport to or during residence within the high Arctic.