CR 30:83-97 (2006)  -  doi:10.3354/cr030083

ABSTRACT: The coastal geography of Barrow, Alaska, makes the city vulnerable to weather events that cause flooding and erosion. This study uses the self-organizing map (SOM) algorithm, an unsupervised learning process that codifies large, multivariate datasets onto a 2-dimensional array, or map, to study large-scale circulation patterns associated with temperature and high wind extremes at Barrow. The analysis first uses the SOM algorithm to produce an automated 55 yr synoptic climatology of daily sea level pressure patterns for the western Arctic for August to November, when the area is potentially ice free. The results are in agreement with previous Arctic climatologies, showing the Aleutian Low to be dominant in southern Alaska, and high pressure prevalent over the Beaufort and Chukchi Seas. The SOM algorithm is then used to study circulation patterns associated with temperature and high wind extremes at Barrow. These results show that high winds are associated with patterns containing a strong pressure gradient between the Aleutian Low and the Beaufort and Chukchi Seas and also with patterns that contain a low pressure system to the north of Barrow. High (low) air temperature extreme anomalies are associated with patterns that produce strong, southerly (northerly) air flow at Barrow. This study demonstrates the utility of using SOMs to investigate the relationship between local weather conditions and large-scale patterns. This approach can be applied to future global climate model (GCM) simulations to investigate the impact of changes in large-scale circulation patterns to local extreme events.

KEY WORDS: Synoptic climatology · Self-organizing maps · SOM · Large-scale circulation ·Extreme events