Seasonal variations of acetone in the upper troposphere–lower stratosphere of the northern midlatitudes as observed by ACE-FTS

Abstract

This study reports on the climatological acetone distribution and seasonal variations in the upper troposphere and lower stratosphere of the northern midlatitudes, derived from observations by the Atmospheric Chemistry Experiment Fourier Transform Spectrometer (ACE-FTS) onboard SCISAT. The acetone profiles retrieved from 5 to ∼20km cover the period from January 2004 to September 2010. The 1σ statistical fitting errors are typically ∼5–20% within the upper troposphere (UT), increasing in the lower stratosphere (LS) with decreasing acetone. The systematic errors range between 15% and 20%. The largest UT acetone mixing ratios (∼1200ppt on average in July over Siberia) are observed in summer in the northern mid- and high latitudes. Mixing ratios are larger over continental regions than over the ocean. Comparisons with airborne measurements available in the literature point toward a possible underestimation in acetone retrieved from ACE-FTS. The largest differences occur primarily in winter and for the background values. This underestimation is attributed to the complexity of the spectral region used for the retrieval. The annual cycle of acetone for the 30–70°N midlatitude band shows a maximum during summer, reflecting the annual cycle of the primary terrestrial biogenic source of acetone. By comparison with ACE-FTS, the LMDz-INCA global climate-chemistry model systematically overestimates acetone mixing ratios lower than 400ppt. This overestimation is thus generalized for the lower stratosphere, the Tropics and beyond 70°N for the upper troposphere. In contrast, in the upper troposphere of the 30–70°N region, where the acetone levels are the highest (>450ppt on average), the model-observation differences are in the range of the observation uncertainty. However, in this region, the model fails to capture the annual cycle of acetone, culminating in July. A seasonal cycle can only be obtained by considering high biogenic emissions but this cycle is shifted toward autumn, likely indicating an underestimation of the chemical destruction in the northern high latitudes.