Sub-synoptic polar mesoscale cyclones (or mesocyclones) are under-represented in atmospheric reanalysis data sets and are sub-grid scale processes in most models used for seasonal or climate forecasting. This lack of representation, particularly over the Nordic Seas, has a significant impact on modeled ocean circulation due to a consequent under-estimation of atmospheric forcing at the air-sea boundary. Using Rankine vortices and a statistically significant linear relationship between mesocyclone diameter and maximum wind speed, a novel parameterization is developed which allows the bogusing in of missing or under-represented vortices by exploiting a satellite-derived mesocyclone database.

From October 1993 to September 1995, more than 2500 cyclones known to be missing from reanalysis data over the Nordic Seas are parameterized into the forcing fields for a global ocean-only numerical modeling experiment. A comparison of this two-year perturbed forcing simulation to a control simulation shows enhanced surface latent and sensible heat fluxes and a dramatic increase in the cyclonic rotation of the Nordic Sea gyre – volume transports increase by four times the average interannual variability. In response to these changes, Greenland Sea Deep Water (GSDW) formation generally increases, by up to 20% in one month, indicating more active open ocean convection. However such enhancements are smaller than the considerable winter-time monthly variability in GSDW production. An accompanying increase in the volume transport of intermediate and deep water overflowing the Denmark Strait highlights an important coupling between short-lived, intense atmospheric activity and deep ocean circulation. The parameterization scheme has the potential to be adapted for use in coupled climate models.