Abstract. Short-lived climate forcers have been proven important both for the climate
and human health. In particular, black carbon (BC) is an important climate
forcer both as an aerosol and when deposited on snow and ice surface because
of its strong light absorption. This paper presents measurements of elemental
carbon (EC; a measurement-based definition of BC) in snow collected from
western Siberia and northwestern European Russia during 2014, 2015 and 2016.
The Russian Arctic is of great interest to the scientific community due to
the large uncertainty of emission sources there. We have determined the major
contributing sources of BC in snow in western Siberia and northwestern
European Russia using a Lagrangian atmospheric transport model. For the first
time, we use a recently developed feature that calculates deposition in
backward (so-called retroplume) simulations allowing estimation of the
specific locations of sources that contribute to the deposited mass. EC concentrations in snow from western Siberia and northwestern European
Russia were highly variable depending on the sampling location. Modelled BC
and measured EC were moderately correlated (R=0.53–0.83) and a systematic
region-specific model underestimation was found. The model underestimated
observations by 42 % (RMSE = 49 ng g−1) in 2014, 48 % (RMSE = 37 ng g−1)
in 2015 and 27 % (RMSE = 43 ng g−1) in 2016. For EC
sampled in northwestern European Russia the underestimation by the model was
smaller (fractional bias, FB > −100 %). In this region, the
major sources were transportation activities and domestic combustion in
Finland. When sampling shifted to western Siberia, the model underestimation
was more significant (FB < −100 %). There, the sources included
emissions from gas flaring as a major contributor to snow BC. The accuracy
of the model calculations was also evaluated using two independent datasets
of BC measurements in snow covering the entire Arctic. The model
underestimated BC concentrations in snow especially for samples collected in
springtime.