(cover photo credit: Grant McFarland)

I am not a professional meteorologist, and have never taken an atmospheric science course.  There are times when I wonder what it would be like to have taken that path (sadly, Oregon Ducks have no such option).  Obviously I have immense respect for the many quality mets that the Pacific Northwest has nurtured, both young and old. It’s also been a pleasure to see such a vibrant amateur community, one that takes weather science seriously.

That said, there are prices to pay in such a profession.  For example, grueling overtime during winter weather events, means TV weather anchors rarely get the opportunity to enjoy such events in peace.  And Portland winter events are notoriously difficult to predict, due to uncertainty as to where storms track and what form (snow, pellets, freezing rain, cold rain, mild rain) the precipitation falls.

The snow and ice storms of February 11-14, 2021 were notorious for their high impact on the Portland metro area and NW Oregon.  Nearly 300,000 homes and businesses lost power due to the ice.  The worst damage was in Clackamas and Marion counties, where some communities had greater than 80% power loss.

What I find most remarkable about this storm, is how accurate the overall forecast was, at least pattern-wise.  The low pressure center responsible for the worst snow and ice was on Friday night; and it was only slightly north of where preceding forecasts had hinted.  Heaviest ice accumulations were a little further east than expected, possibly due to an expanded “Forest Grove Effect” on the west side of the Valley (more on that below).  But the models nailed the basic pattern, and nailed it 3-4 days in advance.  This was a welcome surprise for weather professionals and geeks alike, who tend to be used to worse model performance during winter events.

The Forecast

Back in the first week of February, long-range models like the GFS (United States) and ECMWF (Europe), hinted at cold air coming into the Pacific Northwest during the 2nd week of the month. Models disagreed on details at first:  do we get a massive arctic blast over the entire region, or just a marginally cold east wind?  Will the lows hit Portland and NW Oregon directly…or will they dive south and give Roseburg and Medford snow instead? This part of “model riding” is perfectly normal, and happens with nearly every cold outbreak.

By Monday the 8th, things were settling down to a basic theme:  multiple waves of moisture would move directly into our area at exactly the same time the cold air arrived from the east.  That spelled out a major snow/ice storm in and around Portland, and one that would continue intermittently for four days.

Whenever cold air slides underneath mild air, the potential for mixed winter precipitation presents itself. Moisture that melts in a warm layer and then refreezes at or near the surface, will fall as ice rather than snow or rain:

Here are some original forecast maps for snow and ice, brought to my attention courtesy of Mike Newman:

So that was the idea:  heavy snow in Portland, and heavy ice immediately to the south after light snowfall there.  The reality?  The epicenter of everything was shifted about 30 miles to the northeast.  East Portland and Vancouver and Gresham still got lots of snow, in most cases 8-12″ total.  But snow totals dropped, and ice totals rose, almost as soon as you crossed to the south and west side of town.

The Event 

This nationwide cold blast was associated with a collapse of the Polar Vortex (PV), that normally dominates the Arctic latitudes in the winter.

The first significant wave of precipitation moved in Thursday afternoon, at exactly the same time that cold east wind was kicking up.  Parts of East Portland and Gresham had some light freezing rain Thursday morning, at least on stairs and railings and car tops.  At least in my neighborhood, the Thursday night precip was all snow.  About 2-3″ fell by Friday morning.  Then at some point Friday afternoon the cold air overhead was pushed back a little bit at the upper levels.  That allowed for a warm layer to form up above 2,000 feet, turning the snow into pellets (sleet) and then freezing rain. The drive home from work that day was a doozy for me- I needed quite the push to get out of the parking lot!

Friday night was also when the worst ice was falling south of town.  Over an inch fell on the east side of the Valley, devastating places like Oregon City and Molalla and Silverton.  Further west, there was enough cold air piled up against the Coast Range, that ice totals were lower and snow totals higher.  We also had a ‘Forest Grove Effect ‘ come into play during the late January snow tease; yet that time the models did a much better job picking up on the localized pattern. (Wonder why…?)

To the north of the Columbia River, heavy snow was the big story. All of Western Washington got buried, in some cases by a foot or more.  And to Portland’s immediate east, a Blizzard Warning was declared in the Columbia River Gorge. That doesn’t happen very often; probably no more than once or twice per decade.

Things were quite calm Saturday, with less east wind and only light snow in the afternoon. But hundreds of thousands of Oregonians were now without power, in some cases lasting nearly a week.

The final round of frozen precipitation struck Sunday evening. By now the cold air was eroding overhead and everything fell as liquid rain before freezing on the ground.  This is my Valentine’s Day gift for 2021…how romantic!

The final wave of ice struck Portland and the Gorge, harder than before. My neighborhood probably saw close to 3/4″ accumulation before it was all over.  That’s a lot, and explains why multiple friends in my neck of the woods had outages.  Gratefully, my apartment was spared that fate.

The east side of the Portland metro finally rose above freezing Monday morning, beginning a messy thaw that continued for several days after.  Not until today (Friday the 19th), did bare ground re-emerge in most places…and the large piles of icy snow will likely remain for a few days yet.

The Numbers

It was the second snowiest February in PDX history. The only bigger year was 1949, when 13.2″ fell at the airport.   That said, the Parkrose total of 10.1″ was not representative of the city as a whole; southern and western totals were a lot lower.  Historically, February is not a very snowy month here. Over the 81 years since 1940, it averaged 1.0″ per year, compared to 1.3″ for December and 2.8″ for January.  Only 8 Februarys have seen 5.0″ or more snow at PDX/Parkrose.  However….4 of those 8 snowy Februarys, occurred in the last 8 years. The monthly average for 2011-2020 alone was 2.2 inches, more than double the average for the previous 3 decades.  And now it’s happening again for February 2021.

Here are some official totals for the region:

Parkrose NWS Office:  10.1″  

Troutdale: 14.0″

Gresham: 11.0″

Scappoose: 9.0″

Vernonia:  8.0″

Sellwood: 5.4″

Oregon City: 1.5″

Newberg: 3.5″

Cascade Locks: 23.0″ (Sun. 9am)

Hood River: 14.0″ (by Sunday morning)

 

Camas:  7.0″   

Vancouver Mall:  11.0″   

Salmon Creek: 14.0″ 

Battle Ground:  12.0″

I’m not going to post Puget Sound snow totals. But you can look at the totals for just the Fri-Sat storm here

Ice totals for Portland and NW Oregon:

NWS Office:  0.75″   

Downtown Portland:  0.7″   

Tigard: 1.1″   

Oregon City:  1.25 to 1.5″   

Woodburn:  1.2″   

Salem:  1.0 to 1.25″ 

Silverton:  1″

Aumsville:  0.5″

 

It’s easy to see why Clackamas and Marion counties had the most outages.  Even half an inch of ice can cause quite a few power disruptions.  But 1.25″?  Very little of the grid survives.

Of course, Portland and the Valley got so much ice, precisely because we were on the edge of a YUUUGE and historically cold arctic airmass; one driven southward by a fragment of the disintegrating Polar Vortex.  (Read all about it!)  The coldest air, as usual, plunged into the central U.S.  What was not usual was how far south the cold air dove.  Snow fell all the way down to the Lower Alamo near Brownsville, Texas!  Houston reported a HIGH temperature of 25° on Monday the 15th, a full 41 degrees below climate normals!  And Dallas-Fort Worth saw -2°F on the morning of the 16th.  Subzero temps near Dallas are extremely rare.  I don’t have a sexy-looking map of Texas temps.  But here is the forecast map for nighttime lows, issued by NWS Fort Worth the day before:

The Texas power grid was hit much worse than Oregon’s, though in Texas’s case the temperatures were a bigger culprit than the ice.  Also, the Texas grid is highly isolated from the rest of the country, adding to difficulties in extreme weather conditions.

A Few Additional Thoughts 

The nature of this ice and snow storm brings up a few tangential topics, which are worth addressing:

1.  Are models getting better at predicting winter storms?  Northwest weather geeks are accustomed to seeing “false hope”  on winter weather maps.  Quite often what looks like a promising snowstorm or cold pattern, will bust just 2 or 3 days before the expected event. Even when a cold pattern does come to pass, events rarely play out according to the “original” plan a few days in advance.  The big variable is placement of the incoming lows needed to produce frozen precip.  Normally there’s an expected error of 100 miles or more.  But this time, everything was within 50 miles of the forecast position.  Forecast models are, of course, improving over time. But I don’t know if there were any recent upgrades that might have made a difference.

2.  Why do models clue into the “Forest Grove Effect” in some instances but not others?  They did a great job predicting the late January snowfall in places like Hillsboro and Cornelius and Banks.  But they showed heavy ice on the west side this time, and snow to the east.  By the same logic, the models should have called for more snow on the east side in January.  But in both cases the FG effect won the day, even if models tried to have it different ways.

3.  Will human climate change cause more amplified warm and cool spells from one week (or month) to the next?  This topic has come up repeatedly. If the polar jet stream becomes less progressive and more wavy, it usually means more extreme temperature anomalies in the middle latitudes.  Cold spells could actually become slightly colder, even as warm spells grow much hotter.

4.  Is the recent trend of late-winter snow and cold, part of our region’s “new normal”?  This would take us into uncharted territory. Historically our peak of winter seems to trade off between late December and early January.  Then February sees more days in the 50s, and often feels just a bit like early spring on the milder days.  But now we’ve seen four major February storms in less than a decade, and quite a few minor March events in the past 15 years too.  Maybe it’s just a temporary anomaly, or maybe it’s related to the jet stream amplification. A big climate test will occur the next few years.  Perhaps we see February and March flip back to warm, bringing a record ‘false spring’ event.  We’ve hit 65-70 degrees F before under sunny skies in the last week of February; maybe it’s time to play that game again, or even set some new records?

5.  Why did the snow take so long to melt this time?  I’m pretty sure the ice had something to do with it. Having lived in The Dalles much of my life, I’m very familiar with ice on top of snow…or snow that partially melts and refreezes.  A regular snowstorm probably would have melted away in 3 days of 45+ degree temps, not 5-6 days like this time.  I expect some of the bigger snow piles will hang around for several more days.

 

Looking Ahead 

Is this it for low-elevation snow?  Can we declare the 2020-21 winter season over yet?

I’m quite confident we won’t see another BIG snow or ice storm in Portland this season. There’s no sign of cold east wind returning between now and the end of February.  And March snow events are almost always light, and melt away quickly.  We can occasionally see modified arctic air and cold east wind in the first couple weeks of March (it happened in both of the past two years.)  But that same east wind gives us less snow in March than it would in midwinter.  Daytime temps are warmer too; usually in the low 40s even with modified arctic air blasting through the Gorge.

But we do have a chance of at least flirting with lowland snow again next week. We will have chilly North Pacific troughs on Tuesday and Wednesday, and again by week’s end. Generally you need 850mb temps of -7C or colder to get lowland snow with onshore flow. We may come pretty close to that benchmark, though it remains to see if precipitation coincides with the coldest air:

In any case, a cool and showery finish to winter is good for Cascade snow. The slopes have nearly made up their deficit, and should go above normal in the next two weeks!

Have a peaceful rest of February everyone! -Karl