A numerical study of the development mechanisms of polar lows

Two well-documented polar lows, one occurring in the Atlantic Ocean and the other in the Pacific Ocean, are modeled numerically. These numerical results are compared with our analytic results in order to determine how the effects of nonlinearity and more complete physics may modify our tentative conclusions. The numerical model used in this study is The Pennsylvania State University/National Center for Atmospheric Research (PSU/NCAR), three-dimensional (3-D) primitive equation mesoscale model with a high-resolution planetary boundary-layer parameterization and moist physics. A prediction equation for ground temperature takes into account the existence of variable surface parameters such as albedo, emissivity, thermal inertia. roughness and moisture availability.

The Atlantic polar low developed in a shallow but intense baroclinic zone in the Denmark Strait region during the first 24 h of the forecast period. Sensitivity studies with the 3-D model revealed that the baroclinity was sufficient to allow realistic development while the polar low was in the vicinity of this baroclinic zone. However, both convective and non-convective latent heating and the surface fluxes of sensible and latent heat were necessary for the simulation of observed development after this polar low departed from this baroclinic zone and entered the Norwegian Sea. The fact that convective and non-convective latent heating were both essential for simulating the observed development of the Atlantic polar low is consistent with previous analytic-model results. Like the Atlantic polar low, the Pacific polar low developed in an intense baroclinic zone, but this baroclinic zone was deep. Sensitivity studies revealed that baroclinity in conjunction with only non-convective latent heating, overdeveloped the polar low, unlike the analytic results. It was only through the use of the convective parameterization and the convective heating profile that correct development of the simulated Pacific polar low resulted. Surface fluxes played little role in this development. Thus, moist baroclinity and CISK (conditional instability of the second kind) were both important to the observed development of both the Atlantic and Pacific polar lows.