Black carbon is a key source of uncertainty in regional climate predictions through aerosol-radiation interactions, cloud modifications and enhanced snow melt, and the arctic is particularly sensitive to these effects. Understanding the influence of continental emissions on arctic aerosols is crucial in earth system science, and this influence can be expected to evolve with changes to the atmospheric circulation in response to climate change. This paper uses a machine learning approach to study the factors controlling observations of black carbon in the arctic and quantitatively links these to meteorological processes and trends. This phenomenological assessment will facilitate predictions in the long range transport of black carbon transport under various climate change scenarios.