Automatic weather station measurements in Marguerite Bay (West Antarctic Peninsula) during 2001–2002

Two Automatic Weather Stations (AWSs) were deployed on small low islands (Kirkwood and Dismal) in the center of Marguerite Bay on the West Antarctic Peninsula (WAP)in late May 2001 as part of the Southern Ocean GLOBEC program. Provided by the University of Wisconsin Antarctic Meteorological Research Center (AMRC), these stations transmitted vector-averaged wind speed and direction, air temperature, relative humidity and barometric pressure data back to AMRC via ARGOS, where the raw data were decoded and sent to the Woods Hole Oceanographic Institution for analysis. Both AWSs produced high-quality data through September 2002, with only a few short periods of low winds thought due to anemometer icing. The AWS data are self-consistent, and compare favorably with data collected at the British research station Rothera located on the southeastern side of Adelaide Island which forms the northern boundary of Marguerite Bay, except for wind speed and direction which are influenced strongly by the terrain around Rothera. Wind stress was estimated at both AWSs using the Large and Pond (1976) neutral drag formula and the sensor height above sea level. The AWS monthly-averaged wind stresses are primarily southward, with an apparent increase in magnitude during austral fall and winter to values in the range 0.05-0.1 N/m2. The AWS wind stress is highly variable on synoptic time scales, with the monthly-mean rms magnitude generally exceeding the monthly-mean wind stress. The monthly-mean AWS and Rothera air temperatures track closely, and show that May and June 2002 were much colder (roughly 5 and 8 degrees C respectively) than the previous year and the long-term (1977-2002) seasonal cycle at Rothera, which contributed to the early onset of sea ice in Marguerite Bay and adjacent WAP shelf in 2002. These two months were the coldest June and second coldest May on record at Rothera since 1977. Comparisons of the Marguerite Bay AWS data with the Antarctic Mesoscale Prediction System Polar MM5 model results will be presented.