November 23, 2020

An Important Finding about the September Labor Day Wildfires

One reason why research is so much fun is that once in a while you learn something important that is unexpected.  A new and highly significant finding. 

I had such a "wow" experience recently regarding the September wildfires that caused so much damage and smoke on the western slopes of the Cascades.

Currently, I have a research grant form the National Science Foundation, and smaller grants with the USDA Forest Service and WA DNR, to work on wildfire/weather issues. Recently, our group has completed three papers on the intersection of weather/climate with wildfires:  on the Wine Country Fires of 2017, the Camp Fire of 2019, and Diablo downslope winds of northern CA.

I was working on a paper on the meteorology of the great western Oregon/Washington wildfires, when the big Labor Day Oregon wildfire siege occurred.  So naturally my group turned to understand the event.

Why did this major event happen?  How did it evolve? Was something unusual going on?  How well was it forecast?  These and other questions were on the table.

We knew the strong easterly (from the east) winds during the event were critical for initiating and/or spreading the fires.  In fact, my initial work suggested that ALL major fires on the western slopes of the Oregon and Washington Cascades are associated with powerful easterly winds.

I asked research meteorologist David Ovens to take a look at the upper air weather observing sites in the region, locations where balloon-lifted weather stations (radiosondes) are launched twice a day to give us winds and other weather variables aloft.    

Of particular interest was the Salem, Oregon radiosonde data at the first standard elevation above the surface (925 hPa--around 800 meters above the surface).   This elevation is very relevant to winds observed over the nearby western slopes of the Oregon Cascades.

The record at Salem goes back 64 years to 1956, long enough to tell us a great deal about how unusual the situation was this September.  It did not take Dave long to send me a figure with the requested information and I had my wow moment.

Below is the figure.  Let me explain it.  

I asked him to only plot the 925 hPa (again about 800 m about sea level) winds and to limit the analysis to August and September, since those are the months of historical big fires on the western slopes of the Cascades.  Each observation during the 64 years during those months was plotted, with the associated wind direction indicated by the x-axis and the wind speed on the y-axis.   

You will notice two major peaks in strong winds during those late summer months:  (1) northerly to north-northeasterly and (2) south to southwesterly.    The northerly wind peak occurs when high pressure builds over the eastern Pacific and the southwesterly powerful winds occur when a strong trough or low-pressure system approaches the coast.    The southwesterly winds are the strongest (up to 51 knots!), but they are associated with clouds and rain, so little fire danger from them.

I asked Dave to identify the observations taken during the Oregon fire storm period with red dots--- and that is when the wow moment came.
Look at the red dot for September 8th at 1200 UTC--5 AM (2020-09-08 12z)---just when the fires were accelerating over western Oregon (indicated by red arrow).   

Just extraordinary.  The winds at that time were THE STRONGEST EVER OBSERVED  at the site  during those months for any wind direction from the north, east, southeast, or northwest.  

 The stronger the winds the greater the potential for rapid fire growth, and the greater the potential for fire ignition by failing electrical infrastructure and other causes.

Importantly, these were the strongest winds by far from the east and thus downslope on the western slopes of the Oregon Cascades.   Downslope winds from the east are inevitably very dry and the air progressively became drier the longer they blow from that direction.

So why were the easterly winds so strong?   

Our research identified the reason:   unusually cold air and accompanying high pressure moving to the east of the Cascades.  

The pressure analysis 5 AM Sept 8th is shown below, with a measure of the difference from normal shown by colors. High pressure was centered over Idaho and extended into eastern Oregon.   The colors indicate the pressure was VERY unusual--up to 4-5 standard deviations from the mean for that date.  To give you an idea of how unusual, if the deviation was 4.5, this would indicte an event that would occur once in 147,000 times.

The air was so cold that it brought record cold and snow to the Front Range of Colorado and environs.

So the whole situation is ironic and interesting:  record cold to the east brought record fires to the west.
My latest podcast on Thanksgiving week weather and major NW storms:

Click the play button to listen or use your favorite streaming service
Or stream my podcast from your favorite services:


  1. Fascinating Cliff! Thanks for your insight into this incredibly unusual event. I, as a life long Forester (UW 88) totally understand the dynamics between East wind events and fires. I thoroughly enjoy your BLOG. My son is enjoying your class this quarter.


  2. You have stated in the past that these cold outbreaks are less likely under global warming (on average). But at some point I would like you to address the impact of a negative Arctic Oscillation condition, due to extensive loss of Arctic summer sea ice. This is associated with a weakened polar vortex that allows cold air to move south across the U.S. and northern Europe. Were the conditions on Sept 8 influenced by a negative AO condition? If not, why not? thanks.

  3. What caught my eye in that event, it looks like the regional synoptic pattern for the Oregon winds and fire was very similar to the infamous Santa Ana winds of Southern California. The overall Santa Ana wind setup pattern was shifted well north, which I had never seen before for that time of year.

  4. Cliff you have found the smoke's smoking gun.

  5. Sorry if this is a double post the blogger is being buggy but I’m going to reinstate us do you feel that the energizing the power stations and lines would reduce the threat of the powerlines touching combustible materials during Highwind threats?

  6. Why wouldn’t power companies be required to deenergize at the notification or a threshold of Highwinds to protect structures and Forest

    1. This was an unprecedented fire weather event, so I doubt that such a procedure is even in the "playbook" for the PNW. Most of our wind events occur during the wet season,when there is little concern for fire. But widespread power outages create their own fire problems. When people lose power they look for creative "alternatives" in cooking, heating and lighting. A few years ago in my fire service career I responded to three house fires in ONE DAY caused by power outages.

    2. In this instance, Portland General Electric did depower their lines prior to the event.

      Unfortunately, there were already wildfires in place, mostly ignited by lightning. These then blew up.

  7. People outside of OR have no idea how catastrophic these fies were, and how close they came to taking out a significant part of Portland proper. The fire crews were told to stand down and abandon their posts when it became untenable for them. I live less than twenty minutes from where the raging fires paused at the last minute. Others were not so fortunate.

    1. Yes, we had some unprecedented fire weather in Western WA but it paled in comparison to Western OR. If the Santa Ana-type event had continued for another day the results would have been apocalyptic.

    2. Yeah, Salem was also in the crosshairs, along with cities on the coast like Lincoln City and Newport. It would've made the Tillamook fire of legend look like child's play.

  8. Good work, Cliff.

    As your dot-plot so clearly shows, the strongest winds and the most predominate winds in Western Oregon are from the south and west. Thus most of the power lines are located on the east and north sides of the roads, that is, on the opposite side of the road from which the winds typically blow.

    With the power lines predominately on the north and east sides of the roads, north and east winds are much more likely to blow the trees and limbs into the power lines.


  9. Very interesting. One quibble - the COLORS ON THE MAP.

    I know, they're pressure - but any time there's red on a map near a fire, heads explode at Dunning & Kruger LLP. I did an inversion online and high pressure looks great in dark blue ....

  10. So now armed along with the peculiarities of our snow pattern we have good information on our Western Slope fire patterns. Great information to have. And while all I could find on that then unexpected smoky Friday was a MERV 12 filter it did an OK job by running it with a box fan 24/7 low speed, but now have the 16 in our storage area.


Please make sure your comments are civil. Name calling and personal attacks are not appropriate.

Is Mid-June Getting Warmer or Colder?

 As I will demonstrate below, this past week has been unusually cool around the region. But that leads to another question.... is mid-June g...