Historic September 2020 Fires & Labor Day Windstorm

Saturday October 10th, 2020

It’s been a over a month since a series of massive (mainly forest) fires broke out across western Oregon. It was an event not seen in Oregon for at least 100 years, possibly farther back into the late 1800s. For historical purposes I want to make sure all the information/graphs/charts I’ve gathered will be in one place. This is nothing groundbreaking or “new”, but a summary of the event for future reference.

First, my heart goes out to all of you that lost family members, friends, & neighbors in the fires. As far as I’m aware, we have not seen a loss of life this great in Oregon from forest/range fires since the Bandon Fire of 1936. As of early October, the official death count is 10.

The loss of homes is staggering and beyond anything experienced in Oregon for at least the past 80 years. I remember when the Canyon Creek Fire in 2015 burned 43 homes just south of John Day; I couldn’t believe that many homes could be lost in one fire. But in September 2020, Oregon lost almost 100 times that number! The latest estimate is slightly over 4,000 homes destroyed; over half of those were in the Phoenix/Talent area just south of Medford during the Almeda Fire.

How much of Oregon burned? Just under 1 million acres west of and over the Cascades…most of that in a week starting Labor Day (Sept. 7th).

A summary of the fires (click for a better view)


A historic easterly wind event (storm) showed up at the worst possible time; early September before fall rains had arrived. The same setup anytime between late September all the way through June likely wouldn’t have produced any sort of significant fires.

We’ve always known that September CAN produce big fires west of the Cascades in NW Oregon and SW Washington, but it had been a long time since we’ve seen a huge fire during this period. For years we’ve looked back to the 1933 Tillamook Burn as an example, although this one started in August. This was the view from (I assume) the west metro area; fields look like Hillsboro or North Plains area.

Of course just three years ago, Labor Day 2017, a dry east wind helped burn 50,000 acres. That was the Eagle Creek Fire which spread along 35 miles west, then east through the Columbia River Gorge.

This year we did not see any late summer or early fall rain by Labor Day. Take a look at Summer 2020 ERC (Energy Release Component) of some weather stations on the west side of the Cascades. It reached a record high just after Labor Day (black line). Red is the highest on record for any date in the warm season.

Next image is a measure of “1000 hour fuel” for those same stations. That’s how dry 3-8″ dead branches/logs are on any one day. Notice most of the summer (which was very quiet for wildland firefighters) fuels were a bit slower to dry out than normal (gray line). Until the last week of August…the red line is lowest fuel moisture on record for any date during the summer. Just a few days with light rain showers at the end of August would have made a significant difference. This weather setup that arrived Labor Day came at the worst possible time; fuels over and west of the Cascades were ready to burn and burn quickly. Fire conditions were extreme.

A hot upper-level ridge was centered over and just offshore the Pacific Northwest over Labor Day weekend. During the last few days of August it appeared this ridge would sit directly overhead, possibly pushing Western Oregon up to around 100 during the holiday weekend. But as the weekend approached, models started forecasting a fast moving upper-level trough diving down the east side of the ridge. That would be an early fall pattern with cool air diving into the Rockies and eastward. As the weekend arrived it became obvious this would be an unusually cool shot of air; Denver was looking forward to snow. And models showed a very strong push of northeasterly flow for Labor Day all across the Pacific Northwest. This was pretty much a warm season version of an “arctic front”. Here’s the ECMWF forecast for Sunday afternoon (the 6th). Hot ridge over and west of us. We had already hit 95 on Thursday in Portland, then onshore flow cooled us Saturday into the 70s. On this day we were back up to around 90 in the metro area.

The forecast for Labor Day evening shows the trough dipping down to the east; hot ridge over and west of us is ready to move back in overhead.

Down around 5,000′, the 850mb anomaly chart for Monday afternoon shows an extreme setup for us. Over NE Oregon, temperatures are running about 18 degrees BELOW normal as the “cold” air plunges south. At the same time temperatures over southwest Oregon are forecast to be 23 degrees ABOVE normal. An incredible contrast!

This leads to a tremendous pressure gradient between the cool high pressure to our east and lower pressure to the southwest. Models were forecasting up to 23 millibars from Eastern Washington down to SW Oregon. Take a look at the “cross-section” of wind speed forecast over Portland. I’ve highlighted the important part. This was from the Sunday morning WRF-GFS model run (UW-Seattle). Note the abrupt change forecast around 5pm Labor Day; strong and deep easterly flow suddenly arrives from 7,000′ all the way down to sea level.

I’ve never seen 70 kts forecast by this model, even with strong easterly events in the winter. It was obvious this was going to be a historic wind event (or storm). And in this setup EVERYONE west of the Cascades gets the wind, it’s not like a typical Gorge easterly wind episode. When I saw this Sunday I knew we were headed for something big, especially in the foothills. That’s due to the wind rolling down the west slopes of the Cascades. I figured gusts down at the surface in the valleys would be around 35-45 mph, but they went higher than that in a few spots.

The SPC (Storm Prediction Center) nailed the extremely critical fire weather forecast. Due to high winds and very low relative humidity; remember this started as cool/dry Canadian air.

By midday Labor Day things were looking ominous to our northeast. Numerous fires were started by downed power lines across Eastern Washington (orange areas) and blowing dust (brown streaks) could be seen spreading south toward Oregon in this satellite image.

Then right on schedule, around 4-6pm, strong wind suddenly arrived with a bang. When we get these “downslope” easterly windstorms that sometimes happens; a very quick start. Often the central part of the metro area gets the strongest wind; from just east of the Interstate Bridge east over PDX to the I-205 bridge. Peak gusts over the next 24 hours show that was the case again this time. 52 mph gust at PDX was the strongest easterly wind gust on record in the month of September.

There aren’t many observations in the Cascades, but wind was extreme in those locations. Note the 66mph gust at Horse Creek RAWS, I rarely see real strong gusts there. This was right between the Riverside and Beachie Creek fires. Check out the coastal gusts! I’ve never seen such strong easterly gusts there at ANY time of the year.

LocationPeak GustNotes
Timberline Ski Area (7,700′)106 mph12 hrs w/peak gusts 90+
Hoodoo Ski Area (5,700′)98 mph
Mt. Hebo (Coast Range)92 mph
Horse Creek RAWS (3,400′)66 mph10 mi. NE Silver Falls SP
Lincoln City (tower in city)62 mphMany hrs. missing obs.
Tillamook 52 mphWind blew hard all night
Astoria 51 mphODOT
Newport Airport45 mph

The strong wind coming from an unusual direction plus fully leafed out trees led to thousands of power outages. PGE was up to 80,000 customers out at one point. But they did cut power to the Hwy 26 corridor from Sandy to Government Camp around 6pm Labor Day; a very good move since no significant fires started in that area.

Dewpoints, a measure of moisture in the air, dropped to mid-winter values. PDX went as low as 20, Tillamook to 18! By Tuesday afternoon relative humidity was in the high single digits and teens as the easterly wind blew; we were in southern California fire conditions Tuesday & Wednesday, the 8th and 9th

Of course we were all hoping we could get through 2-3 days of extreme fire weather with no new fire starts. But Monday night GOES fire detection showed massive growth of the Beachie Creek and Lionshead fires. Later we learned numerous downed power lines in the Santiam Canyon started fires that merged into Beachie Creek Fire. Along with the Big Hollow fire in Skamania County they show up as black areas. Also the Holiday Farm fire started and blew up along the Mckenzie River; presumably also due to downed lines. This is just before sunrise Tuesday morning. Notice the Riverside Fire SE of Estacada was still very small at this point.

Sunrise revealed a startling sight from space; massive plumes spreading westward off those fires; along with streamers of dust blowing of the volcanic peaks due to 100 mph gusts up there. I saw this and knew it was going to be bad since we had two more days of dry easterly wind.

The easterly wind blew all day Tuesday, most of Wednesday, and then retreated back into the Gorge early Thursday. Satellite image midday Wednesday

Wednesday was the day fire plumes spread a bit more to the north; during the afternoon thick smoke moved over the metro area. This reminded me of pics from Yakima the day Mt. St. Helens erupted in 1980. Keep in mind there are no clouds here…only smoke. 86 degrees, a dry east wind, and a thick layer of orange/gray smoke moving in.

Thursday was the “transition day” from low dewpoints and gusty wind back into normal (weak) onshore flow and higher humidity. Just in time because Thursday was the day Molalla was evacuated; fire officials say another day of easterly wind may have burned Estacada, Colton, and Molalla. You can see the smoke spreading eastward across the region that day

Fire intensity diminished rapidly Friday and beyond. The vast majority of acreage burned was Monday night through Thursday.

How rare was this easterly wind event/storm?

Quite rare. Definitely a historic event. Professor Cliff Mass did some research and found this was the strongest 2 day easterly wind event during fire season since at least 1950. That’s July-September over NW Oregon. A similar strong event occurred September 16-18th, 1971. That one was quite a bit cooler with high temps near/under 80 degrees. Relative humidity values were not as extreme. Portland had seen 3″ of rain from mid-August to mid-September. Forests were damp; totally different situation 50 years ago.

Were these fires historic?

You bet! I did some research and it’s a bit disturbing. In the past 100 years, we’ve seen 11 “mega-fires” over and west of the Cascades. 3 in the 1930s, one in the 1940s, then 52 years with none! 2 more in 2002/2017. So 5 OF 11 MEGA-FIRES OVER/WEST OF THE CASCADES IN THE PAST 100 YEARS BURNED IN SEPTEMBER 2020

1933 Tillamook Burn 240,000

1936 Bandon Fire 287,000

1945 3rd Tillamook Burn 173,000

1939 2nd Tillamook Burn 217,000

2002 Biscuit 500,000

2017 Chetco Bar 191,125

2020 Riverside 138,000

2020 Beachie Creek 193,000

2020 Lionshead 199,000

2020 Holiday Farm 173,000

2020 Archie Creek 133,000

Why did it happen? Climate Change? Mismanaged Forests? Utility Companies?

It’s a loaded question; my personal opinion is that it was a combination of factors. I’m sure many studies are on the way…

  1. The number one factor is what I’ve already outlined. The most dangerous fire weather pattern showed up at exactly the wrong time. And it happened to be an “extra-extreme” version of that pattern (wind/humidity). Everything lined up just right this year in a very bad way. Even if our climate wasn’t warming, at least some of these fires would have exploded out of control.
  2. No, they weren’t “climate fires”, that’s silly. But our warming climate likely played a part. But how much? 10-20% responsible? 50%? No one knows right now. Summers (especially July-September) have been warming/drying for several decades. I’m not a forester, but it’s safe to assume trees are more stressed & drier than they used to be.
  3. Fires have been suppressed for many decades, but we know fires were a natural part of Cascade forests in the past. Pioneers that arrived in the late 1800s probably experienced late summer fire smoke almost every year. Forests west of the Cascades were/are going to burn at some point; either this way or with controlled burns.
  4. What part did downed powerlines play? I’m sure we’ll be hearing more about this in the future.


As the fires slowed down (Friday the 11th), we were left with light wind and thick smoke throughout the region. It looked like this that first day (Thursday the 10th)

Downtown Portland in the smoke

For most of the next week the Portland metro area experienced the worst air quality seen in modern times. It lasted about 8 days here, but 10-11 days from Salem south since that part of the valley was immediately downwind of the first fire blowups. AQI numbers were regularly in the 300-500 range for most of us. I saw a few 600-700 numbers as well; just unreal. Many of us stayed indoors for most of that week. The numbers from September 8th-14th in Portland. One more reason to refer to this event as historic.

We had a terrible time getting out of the smoke because the weather pattern was just stagnant; nothing unusual for early fall. No strong onshore OR offshore flow to blow the smoke one way or the other. It took a weather system moving through the 17th-18th to finally clear us out.

A few bits of info that still blow my mind; I’ll never forget these

  1. Easterly wind pushed the Echo Mountain Complex from Otis just about to the beaches. Almost 300 homes burned just miles from the 55 degree ocean. Parts of Lincoln City, on the cool coastline, was under Level 3 evacuation orders.
  2. Labor Day morning, just hours before those 50-100 mph gusts hit the higher elevations of Mt. Hood, a small fire popped up within Mt. Hood Meadows ski resort. If it wouldn’t have been doused within those first few hours by multiple water drops, I think we could have seen another mega-fire come racing around the south side of the mountain. Probably right through Government Camp and on down to Rhododendron/Welches. Incredibly lucky! Video here: https://youtu.be/In53N_6ixzg
  3. The explosive fire growth the first two days was incredible. The Riverside Fire grew from nothing to 110,000 acres in just 1.5 days!
  4. A dewpoint around 20, with a temperature around 90 in Portland…amazing.

That’s about it. One last image showing the burn scars seen from space in early October once skies cleared and fires were pretty much out. You can clearly see the five big fire footprints on the landscape.

Chief Meteorologist Mark Nelsen

Comments are closed.

%d bloggers like this: