Sunday, August 11, 2013

Summer Smoke

The Fairbanks International Airport has observed smoke the last several days from the Mississippi Fire near Delta Junction. So where does the summer of 2013 rank in terms of the number of smoky days in Fairbanks? Through August 10th, 9 days have recorded smoke during the hourly observations. Since 1997, smoke was observed an average of 10.6 days during the months of June, July, and August. It appears that Fairbanks is on track for an average number of smoky days.


  1. Smoke from forest fires in Fairbanks is a function of fire proximity and the direction of the prevailing winds. We've had years when the fires burned close to town (in the neighboring Tanana Flats to the south for example), and years when the winds have been favorable to bring it from a distance, like recent condx with SE surface flow.

    I'd be curious to know how a 360 degree wind chart favors direction for August. Seems like it tends to SW with normal August rains, but that may be incorrect.

    As we move into later August and September I suppose future moisture input and winds will determine the fate of any late season burning.


  2. Nice analysis. Is the overall Alaska fire activity also near average? I wonder if it has been SO dry that fires have not been started by lightning, or if other factors might have suppressed fire activity relative to other hot summers like 2004.

    1. Overall, this season has had 1.1 million acres burned. Last year had about 0.25 million acres burned. This season is more typical. I suspect that early season wetness and number of lightning strikes is highly correlated the the fire acreage (I'm sure this has been looked into extensively). My guess is that the number of smoke days is highly correlated to the total acreage burned. This assumes that the spatial distribution of fires is fairly random.

  3. Gary, I looked through the data and it is difficult to characterize the variability of the wind between smoke events and non smoke events. When I looked at the numbers, smoky days had slightly lower wind speeds (about 1/2 mph) but the distribution of directions was not all that different. Since lower wind speeds mean that it takes longer for a smoke plume to move through, it's probably disproportionately weighted toward smoke days versus non-smoke days. When I can dig up the numbers, I would guess that there is a strong linear relationship over long periods of time with the acreage burned within 100 miles of Fairbanks and the number of smoke days. However, for any given event, randomness plays a huge role. I know, not a great answer.

  4. The long term (since the 1930s) median number of days with smoke reducing visibility to six miles or during August in Fairbanks is zero. It's only with the fire "regime change" in the late 1980s that smoke has become at all common in August. Prior to 1988, 22% of years at a million plus acres burn. Since then about 42% of years have a million acres or more burn. A good example of the "new normal".