Abstract
The projected decline in Arctic sea ice extent as the Earth warms in response to increased greenhouse gas concentrations will occur in conjunction with increased precipitation in the Arctic, and more of that precipitation is projected to fall as rain, especially in autumn and early winter. A recently proposed method of offsetting the decline in Arctic sea ice extent would pump seawater on the sea ice surface. Either way, we envision the liquid water first infiltrating the overlying snow layer creating slush. Winter conditions would then freeze the slush to directly thicken the ice. The net reduction in insulation would increase basal growth, adding an indirect thickening effect. Simulating the response to augmented snow layer flooding gives insights that are relevant in the future Arctic with or without the implementation of geoengineering. We use a hierarchy of models to show that flooding snow on sea ice is most effective at thickening Arctic sea ice when flooding begins early in the sea ice growth season. For the geoengineering scheme to be most effective, the pumps must be deployed almost immediately, while there is still a sufficient area of sea ice over which to flood, and must continue for decades. Sea ice loss would be best mitigated if flooding is combined with reducing greenhouse gas emissions. Furthermore, the increase in rainfall over the Arctic in the 21st century is unlikely to offset a substantial portion of the loss due to warming.