The Model Giveth, The Model Taketh Away

In some of my previous blog posts this week, I talked about the potential for a major, extended arctic blast that could hang around for the first two weeks of January. Unfortunately for snow lovers, there has been a massive change in the models over the past couple days towards a much warmer pattern. We are used to “snow busts” here in the Pacific Northwest due to temperatures being just a hair too warm or moisture a teeny bit less than expected, but I’ve never seen such a drastic shift in the models from “bone chilling cold and heavy snow” to “utter confusion” to “snow quickly changing to rain.”

Even though the most recent models have been brutal for us snow lovers, the actual weather Wednesday was pretty interesting. The satellite picture below shows a feature we are all familiar with but are not accustomed to seeing over the ocean.

See that thin band of clouds off the Olympic Peninsula? That’s a convergence zone! It’s formed by the same process as our “Puget Sound Convergence Zone,” where flow splits around the Olympics and converges. The difference is that Puget Sound Convergence Zones require northwesterly winds, while these “Pacific Convergence Zones” require strong northeasterly flow out of the Fraser River Valley. They are rare, but they are sure an incredible sight!

Skies were clear over the majority of the Western Washington lowlands this afternoon, allowing us to see the spotty nature of snow over our area. There’s a good amount of snow from Chehalis to Longview, and if you look closely, you can see increased snow cover just to the south of Seattle where our Puget Sound Convergence Zone set up shop to ring in 2017. There’s also quite a bit of snow on the northeastern part of the Olympic Peninsula and further north near Vancouver, BC. That’s definitely one of the cooler satellite images I’ve seen around here.

With that said, let’s get back to the topic at hand – the absolute desecration of our hopes for gobs of snow in the lowlands.

Last week, models were extremely confident that we were going to have one of the strongest and longest lasting arctic outbreaks in years. We were predicted to have a very persistent “Rex Block” in the Eastern Pacific that would send the jet stream far northward into Alaska. This jet stream would then sink down over our area, opening up the floodgates for arctic air to pour into our region.

I’ve shown the models many times over the past several posts, so I’ll show something I haven’t shown yet: the temperature, precipitation, and 500 mb heights given out by the Climate Prediction Center (CPC). The temperature and precipitation maps show the probability of the temp/precip being above or below normal, while the 500 mb map shows 500 mb height lines in meters (green contours) and anomalies from climatology (dashed red/blue/purple lines).

Below is what the CPC forecast for us on December 29. The entire Western Half of the U.S. and more is deep, deep blue. I can’t ever recall seeing the CPC have this much blue over such a wide area.

The 8-14 day forecast was even more threatening. There was still an extremely high probability of colder-than normal weather over us (especially considering confidence is generally pretty low 8-14 days out), but we were expecting above normal precipitation too. Put those things together, and you get snow, and lots of it. The 500 mb pattern had even lower heights over our area, as models were hinting at sending the polar vortex, the brutally cold area of low pressure that generally hangs close to the poles, into our area. To say weather geeks like me were excited was an understatement.

But on the 30th, something extraordinary happened. The models, which had been so consistent, suddenly went haywire. Most of them warmed dramatically, while a few others stayed the same. Because models had been so consistent, the CPC only slightly modified their outlooks. But the model inconsistencies continued on New Years Eve, prompting drastic changes to the CPC’s outlook. On the afternoon of New Year’s Day, one of the forecasters at the National Weather Service office in Medford mentioned that this was the largest temperature disagreement he’d ever seen in the models, with the European model 40 degrees cooler than the GFS for parts of the region only 4 days in advance.

Compare the 6-10 day and 8-14 forecasts below to the ones just made two days prior, and notice how much warmer and wetter these are.

Why did we see such a drastic shift in the models on the 30th? This a difficult question to answer and one that is definitely worthy of some further analysis. From a data standpoint, our blast that has given us the cold temperatures this week began near the North Pole, where there weren’t many observations. On the 30th, it traveled south to Alaska/Northern Canada, where there are a lot more observations. But it wasn’t just a lack of observational data that caused such a drastic change in our forecast – it was small the details in the general pattern that ended up having huge effects.

Models were having trouble dealing with the interaction between two areas of low pressure: one “cutoff” low to the south of the giant ridge of high pressure that was sending the jet stream to Santa’s Workshop, and the other upper-level trough that ushered in our cool air. Earlier models kept these two separate, meaning that the cutoff low was allowed to stay over the Pacific and help anchor the high in the prime position for an arctic outbreak, while the upper-level trough was allowed to fully push through and deliver a full-blown arctic blast to the Pacific Northwest.

Here’s what the models were showing when we were looking at seriously cold air for a seriously long time. This is the 132 hour forecast initialized last Tuesday morning (12/29) and valid 4 pm this Wednesday (1/3). You can see a huge ridge in the Pacific with the cutoff low near 160W. This is well separated from the upper-level trough that brought us our arctic blast, which is centered over Northern California. This pattern (one ridge with two lows on either side) is known as an “omega block” (because it looks like the greek letter omega, Ω) and is a pretty darn stable pattern. All charts below were retrieved from Lewi Cowan’s “Tropical Tidbits” site, which is an extremely good site for looking at weather models, all sorts of tropical cyclone data, and much more.

With this pattern, the ridge moved slightly north and east as the upper-level trough passed through.

But once it got settled, it strengthened into an even more formidable beast a couple days later and sent the Pacific Northwest into an icebox. Additionally, a cutoff low formed to the south of the ridge, marking the return of the very stable “Rex Block” pattern.

Finally, this pattern held while a polar vortex took an express ride straight from Siberia into the Pacific Northwest. Though the cutoff low weakened, we still had NW flow and a VERY cold air mass. With this frigid air briefly dipping over the Pacific and directing moisture and storms into our region, we had the potential to pick up feet of snow. This would truly be snowmaggedon.

Over the past two days, models have more-or-less latched on to a much warmer pattern. Let’s look at tonight’s run and compare it to last Thursday’s.

Now, here’s tonight’s model run at 4 pm. Notice how much closer the two lows are, and how the high pressure has retreated even further north. Even though you can see the individual ingredients that made up our omega block, it is gone and we are instead seeing strong zonal flow off the Pacific into California and Southern Oregon.

After the trough passes through, the ridge retrogrades further west than it did in previous models. Additionally, the upper-level trough with frigid air dropping down from the north shown in previous models is at least 24 hours faster.

Because the ridge is further west, the “super cold” air goes over the water, missing us completely. It goes so far west that it actually gives is warmer, southwesterly flow instead of the cold northwesterly flow we so desire for snow.

By Sunday morning, the trough has completely absorbed the cut-off low and we are seeing warm southerly flow that will quickly change any snow to rain instead of the epic snows we feared earlier.

There are a couple of ensemble members within the models that still hint at bringing this massive trough a little further east, which would cool us off more and open our chances for snow. However, these models are outliers, and the numbers of them are decreasing by the day. This is really looking like a case of the “Big One That Got Away.” The models have come into good agreement with how the cut off low and trough currently over us will interact, and this gives me confidence that significant snow is not in the cards.

Our best chance for snow is Saturday afternoon, when we will still have cool temperatures at the surface but warmer Pacific moisture will be arriving from the southwest. This will be one of our classic “rain-to-snow” events, with perhaps a slushy inch or two in Seattle before we changeover to rain. Portland will have to content with cold Gorge flow and will likely see freezing rain. I’ll have a more extended post on that tomorrow morning/afternoon, though not nearly as extensive as this one.

Thanks for reading! Winter’s definitely not over, and I have a feeling this won’t be the last time snow is in the forecast for us. Even if January 2017 doesn’t end up being the snowiest month on record, I still had a great time tracking the evolution of the fascinating pattern, and I hope you enjoyed reading my thoughts. See your tomorrow!

Charlie