Storm climatology of the southern Beaufort Sea

Building on the expertise from the Beaufort Weather Office, an objective method of identifying storm periods in the southern Beaufort Sea area based on surface wind speed criteria was developed. Algorithms that used hourly observations from Tuktoyaktuk and Sachs Harbour were trained to identify storm periods over the southern Beaufort Sea. The Master List produced by Eid and Cardone (1992) was used as a training set to tune the algorithms. A check with independent observations taken by a drilling platform verified the soundness of the approach. The National Centers for Environmental Prediction (NCEP) reanalysis dataset was used to characterize storm type. In doing so, upper air data were introduced to the classification process. The important variables were 50-kPa wind speed and direction, and the 85-kPa temperature. The former variable is a reflection of the steering current of the storm systems while the latter is related to airmass characteristics within the storm system. The algorithms were applied to the 1970 to 1995 time period for the months of June to November inclusive. On the average, there were 14 storms per storm season, with a standard deviation of 5. The years 1976 to 1982 were the most stormy with an average of 19 storms per storm season. There was no discernable trend in the storm frequency over the 25-year period. By month, October had the highest storm frequency, July the lowest. The average 50-kPa wind direction during storm periods was used to classify the storms as Arctic, Pacific or Irregular. The three storm types were further subdivided based on the average 50-kPa wind speed and the 85-kPa temperature change. Overall, 58% of the storms were Arctic, 27% Pacific and 15% Irregular. There was an indication that Pacific storms have become less frequent in recent years. This decrease was the result of an absence of Pacific storms associated with relatively small net 85-kPa temperature changes. The data were then stratified by El Nino episodes. There were more storms during El Nino years because of an increase in Arctic storms. However, despite this overall increase, there was a decrease in the percentage of Arctic storms associated with strong cold air outbreaks. This storm information was used to aid in the interpretation of two studies related to conditions in 1994. The first was the Beaufort and Arctic Storms Experiment and the second the 1994/95 water year of the Mackenzie River basin.