Refractory black carbon at the Whistler Peak High Elevation Research Site – Measurements and simulations

Measurements of black carbon at remote and high altitude locations provide an important constraint for models. Here we present six months of refractory black carbon (rBC) data collected in July–August of 2009, June–July of 2010, and April–May of 2012 using a single particle soot photometer (SP2) at the remote Whistler High Elevation Research Site in the Coast Mountains of British Columbia (50.06°N, 122.96°W, 2182 m a.m.s.l). In order to reduce regional boundary layer influences, only measurements collected during the night (2000-0800 PST) were considered. Times impacted by local biomass burning were removed from the data set, as were periods of in-cloud sampling. Back trajectories and back trajectory cluster analysis were used to classify the sampled air masses as Southern Pacific, Northern Pacific, Western Pacific/Asian, or Northern Canadian in origin. The largest rBC mass median diameter (182 nm) was seen for air masses in the Southern Pacific cluster, and the smallest (156 nm) was seen for air masses in the Western Pacific/Asian cluster. Considering all the clusters, the median mass concentration of rBC was 25.0 ± 7.6 ng/m3-STP. The Northern Pacific, Southern Pacific, Western Pacific/Asian, and Northern Canada clusters had median mass concentrations of 25.0 ± 7.6, 21.3 ± 6.9, 25.0 ± 7.9, and 40.6 ± 12.9 ng/m3-STP, respectively. We compared these measurements with simulations from the global chemical transport model GEOS-Chem. The default GEOS-Chem simulations overestimated the median rBC mass concentrations for the different clusters by a factor of 1.2–2.2. The largest difference was observed for the Northern Pacific cluster (factor of 2.2) and the smallest difference was observed for the Northern Canada cluster (factor of 1.2). A sensitivity simulation that excluded Vancouver emissions still overestimated the median rBC mass concentrations for the different clusters by a factor of 1.1–2.0. After implementation of a revised wet scavenging scheme, the simulations overestimated the median rBC mass concentrations for the different clusters by a factor of 1.0–2.0.