NOy partitioning from measurements of nitrogen and hydrogen radicals in the upper troposphere

Recent studies using NO, NOy, OH and HO2 (HOx) observations have postulated acetone and convection of peroxides as significant sources of HOx in the u pper troposphere (UT). This work focuses on the effect these additional HOx sources have on the modeled NOy partitioning and comparisons of the modele d NOx/NOy ratio to observations. The measured NOx/NOy ratio is usually much higher than predicted regardless of the presence of acetone in the model. The exception occurs for air parcels having low NOy and O-3 values. For the se air parcels the measured NOx/NOy ratio is much Power than the calculated ratio unless acetone is included in the model. In all cases acetone increa ses the fraction of NOy that is peroxy acetyl nitrate (PAN) from typical va lues of much less than 0.1 to values as high as 0.35. Including acetone als o reduces the scatter in a comparison between modeled and observed NOx/NOy ratios.