A two-dimensional cloud-resolving model (CRM) is employed to examine the development of the convective plumes that may form in the vicinity of arctic leads and the impacts such plumes have upon the large-scale surface heat budget. Numerous observations of varying types from the Surface Heat Budget of the Arctic Ocean (SHEBA) project were used to construct an idealized clear-sky midwinter case. In a simulation containing a 3.2 km lead under the idealized conditions, a convective plume penetrated the extremely stable boundary layer to a depth of approximately 200 m, and a near-surface ice cloud propagated at least 50 km downwind. It is notable that similar cloud features were also observed at the SHEBA site near the times when active leads were in the vicinity, though it has not been established that they are, in fact, related. Vertical mixing and lead-induced circulations approximately doubled the simulated near-surface wind speed over the lead, greatly increasing the turbulent heat fluxes compared with the initial conditions. The lead-generated plume increased downwelling longwave radiation significantly, substantially modifying the heat budget over the snow/ice surface. Results were compared with the commonly utilized large-scale mosaic parameterization. Owing to the lack of resolved leads and separate over-lead and over-snow profiles the mosaic simulation generated a shallower surface-based plume with fundamental differences in the surface heat budget. Finally, a small number of sensitivity experiments were run to investigate the relative importance of lead width, ice cover on freezing leads, and humidity of the ambient environment. Variations in these parameters had substantial impacts upon the turbulent heat fluxes above the lead and the cloud radiative forcing upon the downwind snow/ice surface. Typically, it was the balance between these two components that determined whether the large-scale net upward surface heat flux was impacted in a positive or negative sense.