• I do apologize that it has taken me so long to get this blog out. It’s been a bit of a project for me because I’ve been looking at this from a lot of different angles, various opinions, and also looking back at situations where uncertain forecasts came into play.  We talked a little bit about forecast uncertainty in snow forecasts in the last two blogs I posted in regards to Winter Storm Maya, but this blog will be a bit more expansive in terms of information and how the local weather community has met and discussed new ways to communicate uncertainties when it comes to forecasts. Some of the TV broadcast meteorologists and the National Weather Service began implementing some of these changes during the last storm, and LBCWeather will be doing so as well too when uncertainty exists. This blog is going to be quite long, just to warn readers, but if you’d like to get a good idea of why our forecasts are sometimes very challenging, it’s well worth the read. Let’s dig in!

There have been numerous “storms” over the last few years that have been over-hyped and advertised to the public as giant storms or catastrophic storms a few days prior to the storm, but then, when the day arrives, nothing happens or the effects are minor. This is because sometimes weather just doesn’t go the way you expected it to, and there’s a whole slew of potential reasons why it didn’t. Our most challenging forecasts here in the Pacific Northwest are snow events and wind events. Sun and warm temperatures are pretty easy to predict, as is moderate temperatures and rain. But things get tricky when wind and snow get involved.

Wind Forecasts

One of the things we have that is different from other locations across the world are significant windstorms, as they impact us much differently than just wind across a flat terrain with few trees and small shrubs. Our tall evergreens amidst a blooming infrastructure can present big challenges when those trees fall and cause damage. If you live in the PacNW, you know what I’m talking about because I’m sure you’ve experienced at least one of these storms. If you’re new here and haven’t…they can be a bit interesting, and even scary for some. Our power grid transmission lines are usually nestled between tall trees and large hillsides, and our telephone poles and electricity lines often fall victim to a falling tree across the roadways. That’s why these forecasts present so many challenges; the weather community wants to be able to prepare the public for significant impacts due to wind, but there are many factors that play into how high the wind speeds will be and where the brunt of the wind will be directed.

Our storms that bring rain and wind are produced way out west in the swirling waters of the Pacific Ocean. After developing from water vapor and strengthening into a storm system, they make their way inland to landfall somewhere on the coastline and then continue inland across land. For Washington State, the areas that will see high winds are directly impacted by where it makes landfall.  This is where uncertainty can come into play.

The strength of a storm determines how high wind speeds could potentially reach. The lower the central pressure of the storm, the stronger the winds will be. The atmosphere exerts pressure on the surface of the earth, called air pressure. As a storm comes closer to a given location, surface pressure drops, as the atmosphere begins to exert less pressure on the earth. As the center of the storm passes, surface pressure will begin to rise towards higher pressure as the atmosphere begins to exert more pressure again. A rapid increase or decrease in surface pressure, or a big pressure gradient, is what almost always causes wind. So the bigger the gap in pressure from the central pressure of the storm to normal pressure, the stronger the wind speeds. The pressure of these storms is modeled by a number, typically measured in millibars (mb). Average sea level pressure is around 1013mb (101.325 kPa; 29.921 inHg; 760.00 mmHg). High pressure typically brings sunny and clear weather, while low pressure causes clouds and precipitation, or storms.

When our local meteorologists look at weather models to determine how a storm will impact an area, they look at many factors. One is the change in surface pressure to determine potential wind speeds after the front moves through. Another, the track of the storm to determine which areas are likely to see effects caused by wind. As many of you probably experienced the “bust” forecast of the Ides of October major “windstorm” this past October 15th, 2016, let’s take a closer look at that particular storm and WHY it didn’t pan out the way we expected it to. Unfortunately, this storm did get blown out of proportion by a few articles floating around the internet advertising 150mph winds and the potential for widespread damage. While the storm was still very potent and could have potentially been catastrophic, it would never have been that strong. It did have the potential to be a very impacting storm, however. That was why the media and local NWS office was so concerned and posted High Wind Warnings, advertising the potential for the severe weather event. However, it was not approached appropriately and the uncertainties in the forecast were not expressed as well as they should have been. The weather community has since met a few times and we’ve discussed better ways to communicate these uncertainties as I mentioned earlier, at the end of this blog I will explain a bit more about that.

Ides of October Storm – Early Saturday, Oct. 15, 2016 by the Suomi NPP satellite. (Source: RealEarth/SSEC)

About 4-5 days prior to October 15th, models began to paint a very serious windstorm in Western Washington. The track of the storm was an ideally “perfect” track for high winds in the entire metro region of Puget Sound and would have had major impacts. Gusts could have topped 70 miles per hour, rivaling some of our biggest historical storms. To give you an idea, many of our routine storms are usually between 995mb and 1000mb or so, give or take a few millibars, bringing rainy conditions across the area and perhaps a slight breeze after the front passes. In the Ides of October case, models were showing the central pressure of the storm ranging between 960mb and 975mb. The strongest storm ever recorded, Typhoon Tip in 1979 that affected the Northwest Pacific, was an astounding 870 millibars and had sustained winds of 190 miles per hour. Average pressure of a typical hurricane in the Atlantic Ocean is 950 millibars. That said, you can see why there was some significant concern when models began to show this storm presenting itself with a central pressure of somewhere around 960mb when it landfalled on the Washington coast.

Over the next few days, models toyed around with the central pressure a bit, some weakening it a little, but still painted a very stormy picture for Western Washington. Each of the global models also began to show slightly different tracks that the storm could potentially take, and a slight alteration in track would make a huge difference in who saw the brunt of the wind. A “perfect” track for a severe Puget Sound metro windstorm is when the storm landfalls somewhere between southern tip of Vancouver Island and the northern Olympic peninsula. High winds are usually on the southeast quadrant of a storm system that is traveling east (on the other hand, an Atlantic hurricane traveling west would have it’s strongest winds in the southwestern quadrant. As the storm changes direction, the area in which the strongest winds are changes as well), which if a storm landfalls right on the tip of the island, the southeast quadrant winds are funneled right down the Puget Sound into the metro areas. This is where the uncertainty in this forecast comes into play. The storm system itself was a very compact and strongly wound up system, whereas many of our storms are relatively large when they are that strong. This one was a bit smaller, but still had the potential to pack a wallop. Therefore, because the storm wasn’t as big, even a small change in track would directly impact what area saw the strongest winds. A further south landfall (such as south of Neah Bay near Forks) on the Washington coastline would mean the entire coastline would see high winds as well as the North Interior (Whatcom, San Juan, Island, and sometimes Skagit Counties) since the winds would funnel down the strait and slam into that area. A further north landfall on Vancouver Island would likely affect the North Interior only and areas northward into British Columbia. The Puget Sound metro area sometimes sees routine windstorms (average wind gusts in the Wind Advisory criteria range, 45-55mph or so) when the track is off by just a little bit from that “perfect” track.

GFS model image from early morning (06z UTC) Friday the 14th, showing a strong windstorm affecting the coast, with a more southern track. Take a look at how the southeastern quadrant of the storm has a pocket of very strong winds, affecting the Washington and Oregon Coasts. This is just one model that showed a different scenario, thus making forecasting this event quite a challenge.

The day of the storm arrived and the local media and NWS were still holding onto the High Wind Warnings just in case, expecting the storm to slam right into the tip of Vancouver Island and cause widespread high winds upwards of 55mph across Western Washington. We all were aware of the “bust potential” on this storm if the track changed or if it was weaker than expected, but it wasn’t communicated to the public as well as it could have been. Weather enthusiasts and meteorologists alike watched the storm form on water vapor satellite imagery, and kept a close eye on it as it moved inland. At the last second, the storm went off that perfect track by about 50 miles, and while the coastal areas of Washington were getting battered with high winds, the Seattle Metro area remained relatively calm. One major squall line moved through the area, bringing incredibly heavy rain and a very short lived round of gusty wind, but that was about it. Winds remained moderate, between 20-40mph, just another breezy day around the sound. Why? The storm ended up further south, and the Olympic Mountains acted as a barrier for Seattle. The high winds slammed into the mountains but couldn’t make it over and into the Puget Sound area. Another really interesting factor, there has been some disagreement that the storm actually had TWO low pressure centers, both slightly weaker than modeled. We weren’t able to see that on satellite, but the weather bouy off Tatoosh Island picked up two centers as it passed overhead. However, it may have been a scatterometer error, and having a coastal radar off the coast of Oregon would have made it easier to determine that factor. Between the somewhat different track and slightly weaker low pressure centers, there was almost no impact to the Seattle metro area.

These next few images are courtesy of Wolf Read, PhD, arguably one of the most knowledgeable persons in the area when it comes to windstorms. The images were taken from his StormKing website, where he wrote an in depth breakdown of the Ideas of October Storm, which can be found here: http://www.climate.washington.edu/stormking/ . 
The breakdown linked above is an excellent wealth of information, but be warned, it is very atmospheric science term-heavy. Also, if you ever want detailed information on most any Pacific Northwest windstorm, his main web-page has a list of storms that he has written a breakdown on. Overall, an excellent resource. I highly recommend reading the first introductory page regardless of your weather knowledge as it helps gain some insight on how our various windstorms have impacted the area.

This image above shows the tracks of 6 classic path windstorms that affected the Pacific Northwest. A classic path storm is one that is “moving inside 130ºW south of 45ºN, then recurving and tracking north-northeast up the coast.” (reference from the above linked stormking website, courtesy Wolf Read).

This image above shows the path of the Ides of October storm from it’s developing stage in light blue, it’s peak intensity stage in yellow, and it’s mature stage in dark blue. As you can see, the second dot of the peak intensity shows that if the storm had in fact continued in a straight line in the direction it was headed, it would have slammed right into southern Vancouver Island, thus creating a significant wind event for the Seattle metro area. Instead, it veered slightly more northward instead of northeast, and tracked over almost Mid Vancouver Island, a classic path windstorm. (Credit: Wolf Read – Storm King website linked above)

This graphic shows the peak gusts from the Ides of October Storm. As you can see, the Washington Coast had a whopping 78mph gust, but yet the metro area only reached a strongest gust of 43mph, not even Wind Advisory criteria winds. (Credit: Wolf Read – Storm King website linked above)

After the storm, many people called out the National Weather Service and TV broadcast meteorologists, saying “Why did we prepare for a catastrophic storm, and then nothing happened? Meteorologists are never right and I’m never taking forecasts seriously again.” This was a big concern for the weather community, as it is incredibly important that the general public takes forecasts seriously, because in some cases, it can be a life or death situation. Those kind of situations are rare here, but they do potentially exist, and it became a huge concern to the weather community to figure out how to communicate uncertainty better so that the general public was able to better understand the different outcomes a storm could possibly have dependent on what happens. Thus, we had those meetings I mentioned and came up with some strategies to combat this concern.

Now that we understand a bit better some of the challenges associated with windstorm forecasts, let’s move onto snow forecasts.

Snow Forecasts

If you’ve been following my weather page for a while, you’ve probably heard me say that trying to forecast snow in Seattle is like trying to nail jello to a tree. I dare you, try it. It’s pretty much impossible. OK, so it’s not entirely impossible to forecast snow here, but it’s pretty darn close. I talked about this a bit in my second to last blog. Here’s the link if you missed it, it’s a really good example of an incredibly difficult snow forecast here in the PNW: http://lbcweather.weathertogether.net/2017/02/04/superbowl-sunday-snow/

Marginal temperatures, elevation differences, terrain, all of it has an effect on snow forecasts here. Since I linked to my other blog, I’m not going to break down the storm like I did above with the ides storm, but I will explain generally how snow forecasts are challenging for local meteorologists by using that storm as an example. Our global forecasting models are excellent tools to help create a local forecast, usually. However, when it comes to snow in the PNW, it becomes a bit more challenging. These models are a lower resolution model, which means they look higher up in the atmosphere, usually about 10 kilometers up. This means they don’t really have the best grasp on temperatures closer to sea level. With our wide range of elevations across the region, the models can’t really view the temperature variations very easily. As I said in my blog linked above, “The models however sometimes have a hard time with marginal temperatures like what we will likely see. Sometimes they overdo the cold, or overdo the warmth, and struggle with taking into account additional factors like elevation and how that relates to the temperatures. We also get what is called mountain bleed, where sometimes the models overdo snow accumulation totals because of how close the Puget Sound lowlands is to the mountains. It assumes the mountain temperatures “bleed” down into the lowlands. Ultimately, the big thing to know is that if we don’t get the right combo of cold and moisture, that obviously means no snow, or very little snow.” I also mentioned in the previous blog that the best model tools for forecasting snow in our area are the higher resolution models that look much closer to the surface. The difficulty with those models? Most only become available 24-12 hours out from when the storm is slated to arrive. This presents another challenge when it comes to getting severe weather alerts out to the public. In the case of Winter Storm Maya (February 5-6th, 2017), the NWS office did an excellent job of conveying uncertainty by sharing various outcomes on their social media accounts, as well as issuing a Special Weather Statement a day in advance of the storm, outlining the possibilities. When the higher resolution models came in the next morning with much more consensus that we would see an impacting snow event, there became higher confidence in the forecast, and a Winter Storm Warning was issued. However, even so, the NWS did mention that if for some reason temperatures remained just a tiny bit too warm, we would only see cold rain. The uncertainty of this system was handled much better than previous potential snow events.

During Winter Storm Maya, Little Bear Creek Weather measured 4 inches of wet snow by the end of the event. Many areas to the south, where the bulk of the moisture was, saw close to a foot of snow. Photo Credit: LBCWeather/JB Hawkins Photography.

Obviously temperature is a huge issue for snow. Typically, when we get cold enough for snow here in Western Washington, we dry out and lack the moisture needed for snow. This is because our cold snaps usually occur from a ridge of high pressure caused by an arctic intrusion of significantly cold air. These ridges in a sense protect the area from frontal systems that contain moisture, whether it be a cold front or a warm front. Sometimes we can see what’s called an “overrunning” snow event at the end of one of those arctic intrusions, when a warm front pushes into the area and overruns the cold air. But, before the cold air gets scoured out completely, the rain that is falling with the front starts as snow. These cold temperatures occasionally hold on and we see a few inches of accumulation before the precipitation changes back to rain as the warmer air aloft wins out over the colder air at the surface. Very rarely, we see a transition from snow to rain occur with a quick round of freezing rain before changing to regular rain. This happens when our surface temperatures are at or below freezing. I also wrote a blog on this a few weeks ago, feel free to scroll back in the archives and check it out if you’re interested.

In Western Washington, it’s not often that we actually have cold air already in place before a cold front comes in with colder air and precipitation falls as snow. We are either just too warm, or we are cold enough but lack the moisture. With Winter Storm Maya this weekend, we were relatively cold already, and a cold front was due in with decently heavy precipitation. The question became, would it push us cold enough for the precipitation to actually fall as snow? Most likely, no. However, the Fraser River Valley outflow, where cold air pours through the valley from up in Canada, was just about to kick in, potentially spreading significantly colder air into the Puget Sound. But would that air come down in time and far enough to keep us cold enough to support snowfall? Meteorologists were concerned that it wasn’t quite going to be enough of a push to really turn us over to snowfall. But, as the storm came ashore, heavy precip rates helped the atmosphere continue to cool through a process called evaporative cooling (I talked about that as well in my previous blog), and the Fraser Outflow winds kicked up quite strong, shoving cold air into the region, causing it to snow the whole duration of the storm. Had we started out one or two degrees warmer, and the outflow not kicked in, we would have remained a cold rain.

Thus, snow forecasts in Western Washington can be incredibly challenging for a variety of reasons.

What is the Local Weather Community Doing Now to Better Communicate Forecast Uncertainty?

Local TV Broadcast meteorologists, private sector meteorologists, NWS meteorologists, and Weather Ready Nation Amabassadors (such as LBCWeather) gathered together for a few various meetings this winter in regards to determining ways to better communicate forecast uncertainty to our “weather consumers”, most likely, YOU. We brought many issues to the table and discussed the challenges to this, and in what ways could we create better forecasts when uncertainty exists. One thing we’ve recently started to implement in forecasts is “possibilities.” Instead of just saying, “windstorm expected this weekend, will bring high winds to the region, prepare now,” we would now say…”there is the potential for a significant windstorm in the area this weekend. There is still considerable uncertainty, however if the track of the storm follows the path a few of the models have shown, we are looking at high winds. We urge you to prepare now, just in case.” A few of the local meteorologists, such as Scott Sistek from KOMO 4 News, have been doing breakdowns, such as a most likely scenario, possible scenario, or simply a “range of possibilities,” using terms like best odds, lesser odds, least odds. We have also tried to implement a better breakdown of what will happen during a storm, timing, effects, etc.

On Little Bear Creek Weather, I have tried very hard to communicate forecast uncertainty and I hope that my efforts have been noticeable. I do my very best not to hype storms, I want my followers to be informed of a situation, but not blasted with false info or overhype. I am also going to begin implementing a range of possibilities aspect to my forecasts, and I hope they will be a valuable tool for you as weather information consumers. With that in mind, I encourage any comments and feedback regarding this issue. Feel free to comment below or send me a message on the Facebook page. My goal for Little Bear Creek Weather is to be an effective and an as accurate as humanly possible resource for the hyper local area of North King and South Snohomish County. All I ask is that you remember that weather isn’t an exact science. As meteorologists and weather nerds, we all do our best with the tools we are given. I am very grateful that I live in an area with so many amazing meteorologists and such an incredible weather community. I have learned so much, and I am able to continue learning thanks to their help and guidance. Eventually, I do hope to pursue a degree in atmospheric sciences so that I can further my knowledge base.

Again, feel free to ask any questions if you have them, but I do hope that this particular blog entry has been insightful when it comes to forecast uncertainty and why our storms occasionally “bust” here in the Pacific Northwest.