Local anthropogenic impact on particulate elemental carbon concentrations at Summit, Greenland

International audienceSummit, Greenland is a remote Arctic research station allowing for field measurements at the highest point of the Greenland Ice Sheet. Due to the current reliance on diesel generators for electricity at Summit, unavoidable local emissions are a potential contamination threat to the measurement of combustion-related species in the air and snow. The effect of fossil-fuel combustion on particulate elemental carbon (EC) is assessed by a combination of ambient measurements (~1 km from the main camp), a series of snow pits, and Gaussian plume modeling. Ambient measurements indicate that the air directly downwind of the research station generators experiences particulate absorption coefficient (closely related to EC) values that are up to a factor of 200 higher than the summer 2006 non-camp-impacted ambient average. Local anthropogenic influence on snow EC content is also evident. The average EC concentration in 1-m snow pits in the "clean air" sector of Summit Camp are a factor of 1.8?2.4 higher than in snow pits located 10 km and 20 km to the north ("downwind") and south ("upwind") of the research site. Gaussian plume modeling performed using meteorological data from years 2003?2006 suggests a strong angular dependence of anthropogenic impact, with highest risk to the northwest of Summit Camp and lowest to the southeast. Along a transect to the southeast (5 degree angle bin), the modeled frequency of significant camp contribution to atmospheric EC (i.e. camp-produced EC>summer 2006 average EC) at a distance of 0.5 km, 10 km, and 20 km is 1%, 0.2%, and 0.05%, respectively. According to both the snow pit and model results, a distance exceeding 10 km towards the southeast is expected to minimize risk of contamination. These results also suggest that other remote Arctic monitoring stations powered by local fuel combustion may need to account for local air and snow contamination in field sampling design and data interpretation