Seasonal Forecasting Using an Analog Technique

Wednesday, July 19, 2017 at 11:00 AM AKDT
Brian Brettschneider, University of Alaska Fairbanks

Many places have a local saying that reads, “if you don’t like the weather, wait 15 minutes.” This idiom is not as applicable to Alaska, where strong seasonality is an ever present fact of life. In December, January, and February, all of Mainland Alaska has snow on the ground and experiences sub-freezing temperatures. In June, July, and August, long days mean warm temperatures, clouds, rain, and mosquitoes. Within those seasons, large variation exists from one year to the next. Winter 2016-17 is much colder and snowier than either of the previous two winters. Those differences are meaningful in the cold season for activities that involve travel on frozen rivers and shorefast ice, following game tracks in snow, meat and fish storage, and more. In summer, year-to-year differences in climate affect fish runs, berry production, river runoff, and fire activity.

An IARC project with John Walsh and Rick Thoman, Brettschneider uses an analog technique to forecast seasonal conditions based on previous months’ conditions. If conditions over a 1, 2, 3, or more month time period are similar to some other year, then highlighting that match year is useful for determining what the current year’s conditions will resemble 1, 2, 3, or more months in the future. For example, a forecast for June temperatures based on the three years with the best global 500 mb geopotential height match to October-December 2016 in the northern hemisphere indicate near normal temperatures for most of the state with slightly below normal temperatures for the eastern interior. This type of pattern-match analog forecasting provides long lead-time indications for seasonal conditions.

Check out the tool HERE!