Our project extends previous work in the analysis of western Arctic ice conditions to: (1) understand the climatic conditions associated with the recent ice reductions; (2) improve our ability to predict the potential for extreme ice conditions; and (3) assess the likelihood that ice conditions in recent years might represent a shift to a new sea-ice regime in the western Arctic in response to broader-scale climatic change.  This work is being carried out using a combination of newly-available data sets and sea ice modeling.  This combination permits, for the first time, a detailed observational and modeling analysis of the role of dynamical and thermodynamical processes in producing extreme variations in ice cover.

Activities during Year 1 have involved assembly of data sets, calculation of ice conditions from ice charts and satellite data, and statistical analyses of relationships between ice conditions and atmospheric pressure patterns, winds, and air temperature.  A principal task was the updating and extension of "sea ice severity" indices that had been developed by previous researchers.  In particular, we consider the index set comprising the Barnett Severity Index (BSI) developed by Barnett (1980) (Figure 1).  A second area of investigation underway is the estimation of local and regional correlations between atmospheric circulation and melt onset in the North Slope area.
 
 

Figure 1.  Barnett (1980) Severity Index.  High severity numbers correspond to "light" summer ice conditions along the North Slope area.  Note the general trend toward
 lighter ice conditions, and the frequency of relatively light ice summers in recent years.  (The notations "Favorable Ice Conditions" and "Unfavorable Ice Conditions" pertain to navigation and access to open water along the coast, and not necessarily to other activities or habitat that benefit from extensive and persistent ice cover).Higher values indicate lighter ice years.
 

Click here to access Year-by-year values of 6 ice severity indices forming the Beaufort Sea Ice Index. Portions of these data were obtained from the National Ice Center.
Indices in this table include:

• Distance from Point Barrow northward to the ice edge on September 15
• Distance from Point Barrow northward to the 4/8th ice concentration line on September 15
• Number of days the entire sea route to Prudhoe Bay is ice free
• Number of days the entire sea route to Prudhoe Bay is less than or equal to 4/8th ice concentration
• Time period from the initial date the entire sea route to Prudhoe Bay is less than or equal to 4/8th ice concentration to October 1

Our results to date from these analyses include:
 

• The temporal trend in the BSI from 1953-1999 shows an increase of 1.87 units per year (implying an increase in light ice conditions), but this is not statistically significant (p-value = 0.29).
• Light (heavy) ice years are preceded in spring by abnormally small (large) MYI fractions in the Beaufort Sea (Figure 2), low (high) pressure in Siberia, high (low) pressure in the Canadian Arctic Archipelago (Figure 3), stronger (weaker) off-shore winds, decreased (increased) ice movement into the Beaufort Sea, and increased (decreased) ice flow out of the Beaufort Sea.
• The relative lack (abundance) of MYI in the Beaufort Sea and the increased (decreased) advection of ice out of the Beaufort Sea presumably lead to increased (decreased) ice volume in the Beaufort Sea, which in turn sets up conditions for light (heavy) ice summers several months later.
• Light (heavy) ice years are also associated with high (low) pressure from Alaska to Greenland, high (low) pressure in the Canadian Arctic Archipelago, stronger (weaker) off-shore winds, decreased (increased) ice movement into the Beaufort Sea, and increased (decreased) ice flow out of the Beaufort Sea in summer.
• Composites of February, March, and April (FMA) 2m air temperatures (Figure 3) imply that spring conditions leading light (heavy) ice summers are significantly warmer (colder) in the Bering Sea, along an area from the Alaskan North Slope to the western Yukon, and part of Russia.
• The lowest BSI value occurs in 1975, when the ice never left the Barrow coastline, while the period from 1964-1967 contains the longest string of years with below normal BSI values.
• The longest run of above average BSI values (five years) occurs from 1995-1999.  Abnormally large BSI values also occur in 1958, 1968, 1993, and 1998.

Figure 2. FMA composites of multiyear ice concentration during a) light ice years, b) heavy ice years, and c) the difference in light minus heavy ice years.
 

Figure 3. FMA composites of sea level pressure during a) light ice years, b) heavy ice years, and c) the difference in light minus heavy ice years.
 

REFERENCES:

Barnett, D.G., 1980: A long-range ice forecasting method for the north coast of Alaska. Sea Ice Processes and  Models. R. Pritchard, Ed. Univ. of Washington Press, Seattle, pp. 402-409.