Ocean-Atmosphere Environments of Antarctic-Region Cold-Air Mesocyclones: Evaluation of Reanalyses for Contrasting Adjacent 10-Day Periods (“Macro-Weather”) in Winter.

Abstract:

Mesoscale cyclones in cold-air outbreaks (mesocyclones) feature in the weather and climate of the Antarctic (e.g., Ross Sea) and sub-antarctic (Drake Passage). They adversely impact field operations, and influence snowfall, the ice-sheet mass balance, and sea-air energy fluxes. Although individual mesocyclones are poorly represented on reanalyses, these datasets robustly depict the upper-ocean and troposphere environments in which multiple mesocyclones typically form. A spatial metric of mesocyclone activity—the Meso-Cyclogenesis Potential (MCP)—used ERA-40 anomaly fields of: sea surface temperature (SST) minus marine air temperature (MAT), near-surface winds, 500 hPa air temperature, and the sea-ice edge location. MCP maps composited by teleconnection phases for 1979-2001, broadly correspond to short-period satellite “climatologies” of mesocyclones. Here, we assess 3 reanalysis datasets (CFSR, ERA-I and MERRA) for their reliably to depict MCP patterns on weekly to sub-monthly periods marked by strong regional shifts in mesocyclone activity (frequencies, track densities) occurring during a La Niña winter: June 21-30, 1999 (SE Indian Ocean) and September 1-10, 1999 (Ross Sea sector). All reanalyses depict the marked variations in upper ocean and atmosphere variables between adjacent 10-day periods. Slight differences may owe to model resolution or internal components (land surface, coupled ocean models), and/or how the observations are assimilated. For June 21-30, positive SST-MAT, southerly winds, proximity to the ice edge, and negative T500, accompany increased meso-cyclogenesis. However, for September 1-10, surface forcing does not explain frequent comma cloud “polar lows” north-east of the Ross Sea. Inclusion of the upper-level diffluence (e.g., from Z300 field) in the MCP metric, better depicts the observed mesocyclone activity. MCP patterns on these “macro-weather” time scales appear relatively insensitive to the choice of reanalysis.