One of the most interesting aspects of the event was the intense line of thunderstorms/convection that moved through around 6:30 PM (see radar image below)
As I will describe in a future blog, this was not a failure so much of the models, but of communication of uncertainty. My profession has to stop providing the worst case or most probable weather evolution, but provide society with full probabilistic guidance. Yesterday was a good example of the failure mode when we do not. The media, such as the Seattle Times and several TV stations, were happy to hype up the storm because of all the interest in such events. Many events were unnecessarily cancelled or postponed, some on Friday or Saturday morning when there was no chance of strong winds.
The key facts are this:
1. By two to three days ago is was clear that there was no chance of a Columbus Day storm event. Prior to that, some model forecasts suggest it.
2. It was clear that a small-sized, but intense low center was going to pass through our region.
3. A small track error (say 50-100 km over a day) would radically change the forecast at any particular location.
4. Such a small error was well within forecast uncertainty as revealed by our ensemble (many forecast) systems.
5. Local meteorologists warned of the worst case situation, but failed to communicate the uncertainty of the prediction. I tried to talk about track errors, but it is clear that I needed to do much more.
More later.
That is why I like to read your blog instead of reading/watching traditional media forecasts. They love the hype and drive in viewership vs understanding the science (which of course is why we love you!).
ReplyDeleteJust remember it is the same on every issue covered by the media. So now that you know they exaggerate (or even lie) about weather, keep in mind they do the same on every other issue. Don't believe what you read/see in the media.
DeleteTo be honest, Cliff, I disagree about needing to talk more about track error here.
ReplyDeleteI write a little read blog in which I post fairly long essays. My observation had been that a high percentage of readers simply do not read more than a few hundred words of a blog post.
Writing more words is not the answer, it's just more material many readers will skip.
Possibly a separate, short, post specifically addressing the width and position of the error bars might work? But mainly I see the problem as readers refusing to read.
It's disappointing but in this hectic modern world, not surprising.
It's unfortunate that people don't read all the way, but I read every word Cliff wrote about this storm. While a high percentage may not read all the way, I do. :)
DeleteHey, I have Zero complaints. We had 10x the normal rain we normally have and the winds here were strong enough to be scary. I understood perfectly that this was a forecast; there was no miscommunication on this side... and I'd rather be warned to clean up the gutters and make sure stuff gets fixed than get caught unaware. Most of the prep was good to do whether or not the storm was coming. I am just grateful it wasn't as bad.
ReplyDeleteExactly, more probability, less hyperbole. Even the NWS was caught up in the hyperbole.
ReplyDeleteI certainly felt that you communicated that there was a good deal of uncertainty in the probability of it landing in the Puget Sound area. I kept trying to share the wisdom with my panicked Facebook friends.
ReplyDeleteCliff - on the contrary, I don't think you made any errors or judgements when it came to this storm. You offered the correct caveats at every turn.
ReplyDeleteCliff,
ReplyDeleteI wouldnt be to hard on yourself. I was reading your blog since wednesday and the de-escalation/uncertainty of the models was illustrated as well as it could have been.
If people want to complain about how insignificant this event wound up being then the next time theres essentially a typhoon careening toward their general direction i guess we just shouldnt tell them...
Keep up the excellent work.
All of that said. It is better to be prepared, in case. If you don't share the worst people today will just ignore the possibilities and therein will have difficulty. I think you gave a fair review of the possibilities with your posts. We rely on your information. I would rather know what might happen than get hit with something I was not ready for. Thank You.
ReplyDeleteThanks for all of your posts cliff, I felt well informed. I agree with your assessment of communicating uncertainty, except in the case of severe consequences. I think a low probability event with dire consequences should be given big red headlines. It's worth the risk of being called the "boy who cried wolf" when the teeth are sharp and the critter looks hungry. Yesterday, I took extra cautions to secure some things in my yard. It took some effort. I am not second guessing the time put in even though it turned out to be unnecessary. Thanks again for your work. I imagine communicating the consequences of "tracking error" in a clear and simple way is a challenge.
ReplyDeleteHi Cliff, from my perspective, you did a fine job of explaining the situation as it developed, including how a small shift in the track would change things significantly. I understood that you were explaining what was the worst case scenario, putting it in historical context against past storms, while also letting us know what was the most likely way that things would develop. I'd rather know the full range, including the unlikely worst case than be surprised.
ReplyDeleteI think anyone who didn't get that wasn't reading carefully enough. You give us all the facts you have which its why I come here and never watch the news casts! Thank you!
Instead of the Storm of the Century, it wasn't even the largest Storm of the Week.
ReplyDeleteKing 5 cancelled their normal schedule to give constant storm updates. You could tell by 11pm how disappointed they were.
Next time one of these comes around, people will ignore the warnings.
Please stop click-baiting people with hyperbolic predictions of massive destruction. It discredits your entire profession.
I just had to leave a comment this morning. I follow your blog everyday and especially so in this last week for obvious reasons. I have always appreciated the fact that you present the FACTS. Unlike TV stations that like to sensationalize. And the result is, of course, misleading information. It's like reality TV. There is not much that is real! But back to the subject of factual weather prediction. Thank you for all that you do to provide the latest information about our hard to predict weather. I was disappointed that this event was not as grand as I had hoped and grateful that you let all of us who follow your blog know that it could turn out that way. Now on to the next weather event!!!!!!!!!!
ReplyDeleteThank you, Cliff. I have been reading your blog posts since Wednesday and feel you did a very good job of describing the possibility of changes to the models. It is unfortunate that we don't also have models for the human "storm" that accompanies the meteorological one.
ReplyDeleteI for one am grateful that the storm weakened, I hate resetting clocks.
Most of these maximum gusts are at very exposed sites or at altitude (places like Destruction Island and other coastal locations and Hurricane Ridge). These are pretty common events with even modest storms, so I remain skeptical that even being 30 miles to the east, this would have been a disaster.
ReplyDeleteAs late as 8 pm at night, the local stations had a crawl across the bottom of the screen announcing major winds were working there way up from the south. I kept going to the NWS "Weather and Hazards Data Viewer" as well as looking at the data from individual stations and I was not seeing any evidence of this. So, I chalk all of this up to reporting negligence by the media and maybe even the NWS.
I agree that having a probability attached to a catastrophic event would help. For non-catastrophes, having a range in the forecast accommodates ordinary uncertainty.
ReplyDeleteHere's one...
https://medium.com/@hul10/mega-storm-packing-150-mph-winds-and-50-foot-waves-set-to-pummel-west-coast-78ee7a56ca6c#.jkxa02qxm
This is an interesting problem. Years ago a tornado hit Atlanta and caused a lot of destruction. The possibility of the tornado was mentioned by weather forecasters but it was given a low possibility of happening and so wasn't talked about much. It happened anyway, against the odds. And people were calling for the heads of the weather folk who didn't provide "proper" warning.
ReplyDeleteAs a result the area weather forecasters went out of their way to make sure that even if something only had a very small chance of happening they hyped it up to the hilt. I guess it's a kind of game theory situation where you don't want to miss out talking about the possibility of serious damage for fear that people will hold you responsible so instead you take the easy way out. Cancelling a few events because we don't understand the odds and the overall situation is better than the 1% chance that things will go really badly and kill everyone. I guess.
And then I seem to recall a similar situation with an earthquake in Italy a number of years ago? Weren't criminal charges brought against the scientists who didn't properly predict the eruption? Pretty sure they were eventually found not-guilty but it's still a scary possibility.
I was thinking this type of event would be a great way to teach about statistics. I wonder what the daily reports would have said, for a specific location, the predicted intensity at the same percentage levels. So for Saturday's event maybe on Thursday it would be 65% chance of wind gusts up to 70mph, 80% chance of wind gusts up to 50mph; 90% chance of wind gusts up to 30mph. On Friday, 65% chance of wind gusts up to 50mph, 80% chance of gusts up to 30moh, 90% chance of gusts to 20moh
ReplyDeleteMy thought is that if we keep the same % every day, and report the prediction at those %, and the whatever happens happens, folks would build up a sense of how often a 65% probability actually occurs.
Somebody must be researching how people interpret weather probabilities.. would appreciate a link to such work in a future post. Thanks for all your work and explanations!!
NWS Seattle consistently warned about possible effects of "wobble" in the storm track. Starting 48 hours in advance, they advised "focus on preparations, not on the storm track." You (meteorologists) SHOULD warn of the "worst-case" scenario, if it is possible, which it was in this case. The public needs to get used to the idea that you can prepare for something that doesn't happen. Then, you say "Thank you," and move on.
ReplyDeleteThanks, Cliff. I always appreciate the updates and analysis. I have learned from you to always read between the lines and take every forecast with a grain of salt. A storm was predicted. A storm materialized.
ReplyDeleteThe US WB blew this whole storm because they relied (as usual) totally on their assorted computer models (9), and discarded the right one - the Canadian one, that showed a more westerly track with much lower wind velocities. The winds here (Hansville) were basically very mild all day with a high gust of 36 mph at 7:16 PM - exactly HALF of their prediction (which they jacked up from 60 to 70 at 4 PM.)
ReplyDeleteAlso, as is often the case, the Aviation forecast was in major disagreement with the Public forecasts and had much lower wind predictions throughout the event. As did the Canadians. Meanwhile, Cliff Mass and his enviro buddies were out somewhere in left field, predicting the Storm of the Century, repeat of the Columbus Day storm (which he now denies - but it's documented on the web) due to this being a left over Typhoon, yadda yadda. By 2-3 PM, I had revised my own prediction and was dead nuts on. The weather radar clearly showed the center further west but the jerks at Sand Point, as ever, are incredibly stubborn.
There is a continual and major disconnect between the Aviation forecast, especially with regard to wind speeds , which I use daily for my flying and the Public forecasts. I have discussed this by phone and in person with the NOAA types repeatedly and get no more than a shrug.
There are two major messages here - 1) Despite a mind-boggling increase in technology over the years (radar, satellites, ocean buoys, computer modeling), the quality of forecasts continually declines, and 2) we have created a generation of meteorologists who are totally dependent on computers and their models with barely a shred of judgement or experience or intuition left in their forecasting quiver.
The scary part of all this is that these atmospheric models, which demonstrate daily their inability to accurately forecast the weather, not just within 24-48 hours, but within 3 hours (!), are basically the same monsters the global warming crowd assure us are accurate for climate predictions going forward 50-100 years. Incredible.
Bob - You raise many good and valid points. It's unfortunate that you cloud them with your right wing vitriol.
DeleteAnd your stretch at the end to tie this crappy forecast into proof that global warming is bunk makes you just look foolish.
It's always wise to hope for the best but prepare for the worst. But you are right that there should be better ways to communicate uncertainty.
ReplyDeleteCliff - I thought you did a good job of representing the uncertainty in this forecast. Sat morning you were upfront that there was "uncertainty on exactly the strength at landfall." Regarding local TV reporters I cant comment as I dont watch them since I know they are not reliable or credible sources of info.
ReplyDeleteCliff, I think you did a decent job communicating the uncertainty compared to EVERY other news source. Because of you, I was much less concerned that this was going to be a historic event, at least for the Sound area. Thanks.
ReplyDeleteThank you for your insights all the way along (and in general.)
ReplyDeleteIncidentally, I downloaded the uWx app the other day and noticed that it seemed to offer much more conservative estimates of wind speed during this storm than media did. Any insight on that? It's a neat app. I didn't even realize my phone had a barometer.
"...this was not a failure so much of the models, but of communication of uncertainty. My profession has to stop providing the worst case or most probable weather evolution, but provide society with full probabilistic guidance."
ReplyDeleteI'm sure I speak for at least 99% of your readers when I say 'thank you' for the work you do to keep us apprised of developing weather systems, but I think the statement in quotes could be applied directly to the climate change issue. I know the media and those of the activist mentality are the ones most responsible for the hyping of catastrophic global warming, but in my view far too few scientists are willing to talk about the uncertainty inherent in climate science, Judith Curry being a notable exception. And what about the uncertainty of the social cost of carbon? Greenpeace co-founder Patrick Moore among others maintains that it's probably negative due to the net greening effect of higher CO2, enhanced crop production of C3 plants such as soybeans, rice & wheat, and the general benefit to life of warmth vs. cold. We can debate the specifics but those within the scientific community who comment on these things ought to at least discuss without prejudice the full range of possibilities.
Thank you for your insights all the way along (and in general.)
ReplyDeleteIncidentally, I downloaded the uWx app the other day and noticed that it seemed to offer much more conservative estimates of wind speed during this storm than media did. Any insight on that? It's a neat app. I didn't even realize my phone had a barometer.
Man... I have a lot of respect for you but take up an issue about integrity. 955am Saturday you said, " So what the key take aways? This is going to be bad along the coast and I expect lots of power outages there. Same for NW Washington, with the San Juans, northern Whidbey, and the Bellingham area to have gusts to 60-70 mph and plenty of power outages. Much less over the south Sound."
ReplyDelete444pm you said it's not going to be bad.
My issue is the media cited you mostly, yet you blame them for the "hype" which just Friday you told everyone this is s bad storm (on mobile, can't go back to cite it but it's there).
The point is, it's easy to say the storm is not as strong while or just an hour before hits, essentially washing your hands of the errors, and then blame the media for over thinking it. And had they not, and the storm did develop as models and you and NWS had said, people would blame the media for being irresponsible. Again, this is not an exact science, you know this, no blame should be given. Rather, a post about the science of why it didn't happen would have been more beneficial to the public and helping them to understand. At least that's what I expected to read when I open up your post today. Thank you cliff Mass for the posts and all the information you have given over the past but blaming the media is irrational.
thank you for all the knowledge and wisdom that you provide. I am grateful that things turned out to be so gentle. We are blessed. Keep up the good work.
ReplyDeleteAll great comments! There is a simple, yet difficult solution to yesterday's forecast problem. More frequent updates. The NWS maps are useful, but often crude (like yesterday's map using county lines for wind advisories). The forecast updates do not come fast enough for changing situations. How many times have we seen knowledgeable comments get ahead of forecasts on KOMO, NWS, or this blog?
ReplyDeleteThis blog is awesome, but you can't be expected to provide that level of updating either. The best weather forecasting I ever experienced was in the Washington, D.C. area. They had great maps showing probabilities better than anything I have ever seen out here. That's because they had a team of dedicated weather junkies, the Capital Weather Gang, providing regular updates. More frequently during notable weather events. If they only had had a forecast ever 4 to 8 hours during a fast-changing event they would be about as accurate as what we saw yesterday around here. But during weather events you could rely on frequent updates from somebody on the team. If a model was going bust for serious wind or snow they had updates with an updated map clearly showing who should still watch out. They were fantastic at cutting through hype too, which meant when a serious storm was on the way they were taken more seriously than the local TV coverage. They started off independent, but then were so effective that the Washington Post picked them up.
Better probability graphics and more frequent updates, that's the difference for 21st century weather forecasting. The NWS does not provide that and KOMO tweets are not a substitute. Until we have a team like that out here situations like yesterday will continue to happen because the conditions get ahead of the forecasts quickly. Do not get me wrong, hats off to a great blog and great site, but improved forecasting for fast-changing events needs faster and more frequent updates.
https://www.washingtonpost.com/news/capital-weather-gang/
I still appreciated reading your blog on the progression of the storm! I purposely did not watch any news coverage on it to escape the sensationalism. All in all, I ended up with a couple of flashlights and spare batteries, which I should've had anyway. Thank you for your coverage.
ReplyDeleteI got spooked because someone forwarded your blog post to me.
ReplyDeleteI tend to filter the news as sensational.
Scientists talking is what makes me think about stuff.
Cliff, I would respectfully say that you have no control over how a probabilistic forecast would be repeated in the press. On one hand, emphasizing the worst case scenario leads to layperson complacency when that scenario does not play out. On the other hand, you will save lives when it does. The best thing about these events is that you briefly cause people to think about preparedness, which is not a bad thing indeed.
ReplyDeleteI understand what you say about presenting probabilities in forecasts, and I salute that, but you are working against human nature. People all love a big event, and the media is extremely good at feeding that love (or horror as the case may be). Worst case is much more interesting than best case, whether in face-to-face discussions, social media or news media. Everyone wants to say things that other people find interesting. Apocalypse is more exciting than party cloudy.
ReplyDeleteYou can examine (almost) everything in society and see the same process. This is not a new thing, at least since humans have been around (as a species, we are all survivalists at heart), but electronic media provides a lot more fuel than was present in the past.
For my money, a best case/worst "meter" with a needle moving back and forth within the range of probabilities (updated with each forecast) would be much more interesting to me. The range of the meter can change as well as the 'highest probability needle' can change as the event nears.
Hurricane forecasts do this fairly well (albeit with different visuals), and people have learned to work with this in areas like Florida. They sometimes have to prepare for the worst case (given that damage potentials can be catastrophic), boarding up windows and evacuating, but in our region our responses can be more calibrated.
But I was truly taken back when relatives on the East coast, who had watched the national news, called to ask if we were evacuating. Talk about crazy.
Hey, at least people got their storm kits ready for November. That's a good thing.
ReplyDeleteThanks, cliff. I aporeciated all the info you gave. Your site and wunderground.com have become my two weather outlets. You help me understand what im looking at and looking for and wunderground gives me updated maps on whats happening. Between the two i can usually keep a decent idea of whats actually happening.
ReplyDeleteIt does seem like though that ALL the models were suggesting a stronger low and a different track than what developed -- I'm guessing that doesn't happen often. Yet, this is a discipline based on probabilities, not certainties.
ReplyDeleteDon't be too hard on yourself about uncertainty. Society doesn't handle nuance or uncertainty well. Indeed, climate change deniers use uncertainty to provoke inaction. Social science shows that people process uncertainty in a way that leads to denial or inaction. Do we need to communicate scientific uncertainty better? Sure. But communication is a two say process. Until people receiving and processing information learn better ways to use uncertain science, we're still going to have problems making good decisions.
ReplyDeleteWhile everyone at work was freaking out and calling out I was calmly reading your blog and continuing my life. The news outlets are a failure, it's easy to see with weather because they are quickly proved wrong yet people still trust their "hype" about all the other events (election, etc). Thank you for being a source of information and not hype or opinion.
ReplyDeleteI agree with your point, but...seems like you did plenty of the hyping yourself (yes, up until two days ago). I totally understand how hard this kind of thing must be to predict, and you did definitely point out that there was still uncertainty. But these are the bits people paid attention to:
ReplyDeleteCliff Mass, Tuesday, 10/11/16: http://cliffmass.blogspot.com/2016/10/warming-major-storms-threaten-pacific.html
"Now in most years, the events along would be impressive, representing one of the strongest events of a typical winter season. But folks, this is just a warm up for the real action on Saturday.
A true monster storm, potentially as strong as the most powerful storm in NW history (the Columbus Day Storm of 1962) will be approaching our area on Saturday."
Cliff Mass, Thursday 10/13/16, http://cliffmass.blogspot.com/2016/10/storm-update_13.html
"Bottom line: the Saturday storm is the big windstorm threat and has the potential to be one of the strongest windstorms we have seen in a few years. There is still uncertainty in the forecast but it is far less than earlier."
Cliff Mass, Friday, October 14, http://cliffmass.blogspot.com/2016/10/major-windstorm-tornadoes-and-much-more.html
"I have looked at the latest forecast model output and they are all on pretty much the same page, which increases forecast confidence substantially. The bottom line is that we have a dangerous storm, comparable to the 2006 Chanukah Eve storm or the 1993 Inauguration Day Storm, one that is following nearly a perfect track to produce strong winds over the Puget Sound region. And the coast is guaranteed to be hit hard."
Dr. Mass, You call out to the media is good, but do not forget your part in this. Your prior posts all week hyped up the "historic" nature of this, "record-breaking" of that just like they (the media) did. You did provide more context, of course, but you are not blameless here.
ReplyDeleteThanks Cliff, I always turn to you as the "voice of reason" for Seattle area weather. You had me a little worried early in the week, but then your calm descriptions, had me ready, but not worried. Thanks
ReplyDeleteIn one word (about media coverage): engagement.
ReplyDeleteThat's how they make money, by keeping people engaged in their transmission. See how the two things that matter most are the ones that are sliced throughout the news edition: traffic and weather. They never give the full report at once. The slice it in layers, like a cake. Then, they serve it bottom up, finishing with the icing and sometimes with a cherry on top of all.
Is it bad? Yes, it is. Is it their fault? Unfortunately, no... That's how they make money and how they get paid by announcers to relay news to us, for free. Do I like that model? Certainly not. But it seems that it is the only model that has been proven viable until now and probably for many years to come.
What can we do about this, again in one word: education. People need to be educated to become critical thinkers. That will help them across a variety of areas, so they don't get fooled by anyone in control of the information and how it's relayed. That's true freedom, the only one worth fighting for. The freedom to make your own decisions, based on the best available data, recognizing its errors, and following a reasoning model that will issue the same results when presented to someone else with the same set of data. Once they recognize their public doesn't see them as a trustworthy source of information anymore, they will change up a notch.
It's hard to get there, likely impossible. But that should not prevent us from trying.
Please keep up the good work, educating us on this amazing thing that is weather!
Thank you!
that storm never ever looked like it would do the worst case as October 12, 1962 and you are absolutely correct when you say all forecasts were heavy on the front end and the hype never let up. Of course we were prepared for being one of those places that catches it just right/wrong but the media left the story line basically unchanged. Boooo.
ReplyDeleteThe errors are small enough to me, and they are for an extratropical cyclone. They don't know that errors are usually much larger for tropical cyclones and those systems are much smaller.
ReplyDeleteThis fairytale storm was hyped by everyone. Even the national media got hold of the story and publicized it. No wonder the local media and the NWS has no credibility. What a joke!
ReplyDeleteI've lived here for decades, and storm predictions usually don't pan out. I had a strong suspicion this one wouldn't, either. The mass media communicates with the lowest common denominator. Our topography is just too complex to forecast anything more than a few hours out.
ReplyDeleteOn the one hand, it seems like some standardization of the probability of any scenario might be in order. Alas, as that might result in a new mysterious rating table.
ReplyDeleteOn the other, it isn't necessarily a bad thing as a society to slow down some and enjoy a day at home while being prepared for a storm. Brings us all together against a common reality.
This comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteWell said. I too remember Columbus day, Feb. 79 and other major storms. Cliff and his blog are valuable. Us weather geeks do know the risk of these storms and know the uncertainty as well.
DeleteGreat job Cliff!
We sure appreciate having more rather than less information, and I'm so happy you and others work to provide it.
ReplyDeleteThe conclusions drawn are another story, by Thursday I thought people were overreacting in terms of canceling events during the day, but one never knows how it looks from the event staff's position...
The NWS was quite clear from day one that (a) there was a lot of uncertainty, and (b) even a small shift in track would have substantial consequences as to the storm's intensity in Puget Sound. And, on top of this, the storm did not "bomb out" as predicted.
ReplyDeleteThe problem is not the NWS or even the TV meteorologists. Every TV meteorologist I heard was saying things along the lines of what the NWS was saying.
The problem IS the wall-to-wall "storm coverage" that occurs outside of the weather forecast segment. News crews driving around to look at storm drains before the big storm, etc. In the general coverage, very little effort is put into cautioning folks about the uncertainty. People focus on what they hear during the 26 minutes of general coverage, not the 4 minute weather forecast segment.
I will say this, though. Better over-prepared than under-prepared. A 100 mi rightward shift in track and 10mb lower pressure, and everyone in Seattle would be without power right now.
Thank you for that Cliff. There's likely a stat/decision science term for this phenomena....my impression in this storm and some other big ones is you have 2 elements going on. First the direction/path of a storm, and then equally important its size/likely impact. It feels like in both the TV silly forecasts but also even in the NOAA stuff the severity drifts into the path prediction and then the ending forecast sort of "creeps" worse. I saw this in Miami where the predicted wind speeds in town were really low, but because of the severity the Warnings were elevated to a higher level.
ReplyDeleteI saw this here also; one of the NOAA discussions said the intensity (a few days prior) didn't meet the speed threshold but because it was the first storm they'd elevate the warnings anyways.
I suspect that also then gives the TV goofballs license to run wild. My "favorite" was (I think it was Nick Allard) in the middle of a reasonable sounding forecast, then show the tracks from 3 models...2 of which were way out to sea. He said almost exactly, "but because this one is closer to shore" we'll use that. Yipes.
I think you point here to a general problem in Atmospheric Science. In this case it is a storm intensity prediction but it could also be climate change predictions. We rely on complex mathematical models or ensembles of models that output a range of possible outcomes in the future -- there is the worst case, the most likely case and the best case. Continually pointing to the worst case makes one seem alarmist and leads to the "little boy crying wolf" problem.
ReplyDeleteOn the point of some sources hyping: true of all media all the time because it sells; true of humanity on a tragic scale. What if after 9/11 we had a president declare that it was not an act of war, but an act of terrorism which is impossible to prevent 100% of the time and unlikely to be repeated on that scale in the future?
ReplyDeleteAn inherent conundrum: will people get what they want or what they need? If the latter, who will decide what they need, and who will adjudicate such adjudicators?
Hi Cliff - I would disagree with your "I must do more" comment. I think, in the end, this was a PEBCAK problem. You - and the NWS - talked about the uncertainties, how different the models were earlier in the week, and how a relatively small change in the storm's track (exactly what we saw) would affect the local area forecasts substantially. Some people may speed-read those detailed discussions, but that is not your problem - I've learned that some people simply will not pay attention to details.
ReplyDelete(Long boring explanation for my cynicism ahead)
I do web work; much of the time I'm developing intranet-type tools for a UW department. Back when Microsoft stopped letting Internet Explorer (IE) languish and started trying to bring it up to par with other web browsers, we had to make a decision that we would stop supporting anything older than IE 7 on the intranet (no big deal, since we controlled what browsers were installed - except apparently users could throw IE7 into "compatibility" with IE6). This change was talked about in staff meetings and warned about in emails, yet people would complain "the intranet isn't working right". We would explain one on one that they couldn't use compatibility mode, but they'd still do it and then complain. We started adding warning text to the pages for "compatibility" users (which included an FAQ link)... still we'd hear the same complaint. Finally I got a bit ridiculous and put the warnings in bright red bold text in a big red box at the top of the page for those users... I still remember going to this one woman's office after that to deal with "broken intranet". I looked at her screen, saw the bright red box, and asked her about it. Her answer was "oh I just ignore stuff like that".
So does your advice also apply to global warming models?
ReplyDeleteI'd much rather be prepared and aware of the potential worst case scenario from experts. I can always pick and choose my sources of information, which is the beauty of a free society. The consumer has to be discerning. I'm grateful for your knowledge and look forward to more intelligent weather discussion! I'm just glad it was a great outcome.
ReplyDeleteCliff, no apologizes necessary. We live in MX and deal with TS and Hurricanes several months a year. Here the Watch and Warning system for severe weather is well understood and updated by NHC and the MX weather service with the advance of the storm. Along with cone of uncertainty for the track, folks are well prepared and no one is disappointed if the storm passes by.
ReplyDeleteThe Weather Guy, Alamos, SO. see my blog on FB
North Sound weather was more severe during Friday first round of the storm.
ReplyDeleteI find it ridiculous about the amount of people that are complaining to Dr. Mass. First off he was regularly updating the for case. Yes it changed, like western washington normally does. And yes the media made a big deal about everything. But..... He writes this blog for free with his own time and expert forcast. Yep it might have not been 100% correct but normally western WA forcasts are not... So quiet down before he decides not to post any more.
ReplyDeleteThe reason the predictions were off, and the media bought and then sold the sensationalism to their viewers, readers and listeners is the continued drumbeat of CAGW. Rather than a dry, cautious and scientific approach to climate and weather most meteorologists, scientists and almost all politicians, bureaucrats and journalists have taken a cavalier and sensational approach to every detail of climate and weather. And this has been going on for decades.
ReplyDeleteSo it's no surprise that even relatively reasonable scientists like Cliff err towards the side of the sensational now and again. If it bleeds it leads, even if it doesn't bleed.
Cliff -
ReplyDeleteLong time reader here but disappointed in all of this.
You had as much to do with the hype as the "media" you blame above. Do you not see that?
You can preach to the choir (see comments above) as much as you want but the reality is your credibility, and the credibility of your profession, took a big hit here. If this truly is a science blog then I look forward to what you learned about the science and the modeling and why they failed to the degree they did. If it is a political blog (see all of the well placed psa's for I-732 on every post you made on this storm of the century) then count me out.
It's your blog, you can do what you want. But the optics on this are not good for someone who I have always trusted to be way above that.
Well said, sunsnow 12, and I wholeheartedly agree.
DeleteI'm glad the storm didn't pan out in a worse case scenario. Now, since all my gutters are clean, trees trimmed, outdoor furniture secured, generator fueled up, chainsaw prepped, food and water stored, fresh batteries for all my flashlights, plenty of firewood for heat and propane for cooking, I can rest easy the rest of this winter!
ReplyDeleteGreat work, saved my weekend including my 30 yr high school reunion. It's not for everyone, but Cliff if you ever want to product-ize your approach, let me know :)
ReplyDeleteDr. Mass,
ReplyDeleteYou did a terrific job explaining and teaching about the uncertainties involved. We are so fortunate that you take the time to write your blog and explain complex ideas in terms that laypeople can understand.
Thank you, thank you. You provide a great service to the community.
Cliff thank you for doing the right thing and I love how excited you get with weather. Comparing you and how you kept the forecast honest and the local news freaking everyone out, you for things right. Thank you.
ReplyDeleteIn the aftermath of the big storm that did not materialize as expected, I would like to say a few things about weather prediction in the Pacific Northwest and about some comments I have heard or read today in which folks state that it was all nothing but hype and that they won’t believe future predictions.
ReplyDeleteI have lived in the Northwest almost all my life and have lived through the famous 1962 Columbus Day Storm as well as a number of other notable storms.
Columbus Day in Portland in 1962 was a school holiday and I was out driving around with a friend. By the time I dropped her off at her house the storm had begun and my drive home was terrifying with telephone poles waving like twigs and debris howling down the street corridors. My brother was downtown at Portland State College and my family was afraid he would not get home OK. Thankfully, he did and we spent the next four days in without power amid huge destruction in the Northwest.
Years later, on the eve of Valentine’s Day 1979, I was a young mother with a pre-one-year-old whose husband left Fisherman's Terminal in Seattle the day prior for another few-months-stint in Alaska on a 100-foot king crab boat. We lived in a rented house at Maxwelton and when the storm hit, the smoke from the heat stove began pouring back down the chimney and into the house. This is the storm that took out the Hood Canal Bridge.
To prevent us being asphyxiated by the smoke, I had to carry the burning logs outside in the dark in wind gusts that were recorded at nearly 80 mph, tossing them on the lawn with sparks flying at me and back toward the house as well as wind-driven debris. I could barely stand up in the wind and it was very scary as I was alone with a small baby to protect. I had no idea if my husband’s boat had found shelter or had sunk. In the days after, trees and power poles littered the roads of Whidbey and there was no power for about a week. Fortunately for me, Harold and Sally Robinett offered me and my baby shelter at their much-better-prepared home and I will always be grateful. My then-husband and his shipmates found shelter in the lee of an island and suffered no damages, I learned about 10 days later.
Both of these storms struck with little or no warning.
Since that time, the science of weather predicting has advanced with satellites, computer modeling and a better scientific understanding of weather functions. Predicting weather in the Pacific Northwest is notoriously difficult for a number or reasons.
On the leading edge of weather science is Cliff Mass, an atmospheric scientist at the UW. He has a regular spot on KNKX, has written a book on weather in the PNW and has a blog online that explains weather in our region to the average reader. I respect his work and opinion.
So, when Cliff issued a warning about Saturday’s storm saying it had the potential to approach the danger of the ’62 storm, I listened.
His blog today gives good information about what occurred — or didn’t. I don’t watch TV so have no idea what the news readers on local stations said. Maybe they hyped it with no background info. But, I, for one, am very grateful to Cliff Mass and other weather scientists for reasoned information and warnings. I would rather be prepared than surprised. And, now most of us are prepared for this winter.
We are all fortunate for better weather forecasting, some warning about volcanoes and maybe in the future about earthquakes.
Probably just annoy you now to comment further about storms realized or otherwise. But i found the whole set of details leading up to what was short but weirdly vigorous event to be highly educational. And hey, i got fresh batteries in all the emergency devices - and that can only be a good thing. Thank you for keeping us in the loop! It would be so easy to play it safe and close to the vest.
ReplyDeleteJust give us the best guess you can make and adjust it as things get closer .... this year the weather guessers have been a bit hyperbolic and blowing the weather calls
ReplyDeleteLet's be clear. For a large number of people, including myself, this was a primary source of information. There were many scary statements typed in BOLD that warned people that hundreds of thousands would be out of power and to be prepared and safe.
ReplyDeleteI am OK with a failed forecast. Let's just not put it all on the Seattle Times or dumb readers. Readers READ the blog and responded accordingly to what was posted here.
David Riley above made the most profound statement: "The problem IS the wall-to-wall "storm coverage" that occurs outside of the weather forecast segment. News crews driving around to look at storm drains before the big storm, etc. In the general coverage, very little effort is put into cautioning folks about the uncertainty. People focus on what they hear during the 26 minutes of general coverage, not the 4 minute weather forecast segment"
ReplyDeleteI forecast each night on at least 4 newscasts and always assume people will pay very close attention to only what I say. I bet you are right, the other 8 minutes in the "A" block most likely drown out my 1 minute or so...just in that segment alone. Never though about it that way. That said, my producers are very good and asking me if "it's okay to say xxxx" or "can I say xxxx". I don't think that communication is happening at some other tv stations...Mark Nelsen, KPTV, Portland. Keep up the good fight Cliff!
What I think was missing from all the forecasts on this coast was the probability percentage. Having lived through many hurricanes on the east coast, I understood the chance of predicting where the storm would hit but I seemed to be explaining that to a lot of people here! Nowhere did I see a probability map leading up to the Saturday storm.
ReplyDeleteI work as an economist for the State of Washington and have been beating my head against the same wall my whole career...How do you communicate uncertainty SO effectively, that it dominates urges toward apophenia and drama.
ReplyDeleteIf you ever come up with a way to convey "reduced uncertainty" as a model's principle product, please let me know.
Cliff, all you guys can do is prepare us for the possible, this is better than being unprepared.
ReplyDeleteCliff, you ARE the media. Your blog was the most widely-circulated bit of weather journalism last week. When you use language like this:
ReplyDeleteOct 12th: "Now in most years, the events along would be impressive, representing one of the strongest events of a typical winter season. But folks, this is just a warm up for the real action on Saturday. A true monster storm, potentially as strong as the most powerful storm in NW history (the Columbus Day Storm of 1962) will be approaching our area on Saturday."
And even on Friday, October 14th: "I have looked at the latest forecast model output and they are all on pretty much the same page, which increases forecast confidence substantially. The bottom line is that we have a dangerous storm, comparable to the 2006 Chanukah Eve storm or the 1993 Inauguration Day Storm, one that is following nearly a perfect track to produce strong winds over the Puget Sound region. And the coast is guaranteed to be hit hard."
you scare the pants off people. You are the most listened-to meteorologist in the area. When you put out alarmist verbage like that, the OTHER media and disseminators of public information think you're onto something and follow your lead. The newsrooms reacted (and ignored their in-house meteorologist that were much more conservative), the Times reacted, and in my opinion, even the NWS reacted and made a bigger deal out of this than they would have had you not sounded the sirens so fervently.
We get former typhoons in some shape or form around here nearly every fall.
The models are a great tool, and certainly give us a much better idea of what will happen than when we had virtually nothing to go on 30+ years ago. But as we always see with wind and snow in particular, they're not gospel, and you never *really* know what's going to happen until the storm comes on board.
I remember you threw the media under the bus last year when they followed your lead on an approaching windstorm that fizzled, too. So disappointing.
You did a perfect job of communicating the track errors, and your day of report on the storm spelled out very clearly that the storm had been downgraded and would not be doom and gloom like the NWS original report. That NWS original report, while spelling out the 3 possible scenarios, painted a grim forecast (the 1 in 3 chance) and that is of course what folks seized onto .
ReplyDeleteYou communicated effectively and should not feel in any way you failed to deliver the right message. Think like lots have said, people don't read the entire statement and want the quick bit of info.
This is EXACTLY why you're trustworthy Mr. Mass. Your honesty and integrity show through in times like this. Thank you.
ReplyDeleteEden said...
ReplyDelete"What I think was missing from all the forecasts on this coast was the probability percentage. Having lived through many hurricanes on the east coast, I understood the chance of predicting where the storm would hit but I seemed to be explaining that to a lot of people here! Nowhere did I see a probability map leading up to the Saturday storm.
October 16, 2016 at 11:02 PM"
This is an interesting point, as are a number of others raised here, including the mention of "Risk communication" that Cliff thinks that professionals, such as himself, need to pursue more effectively. Considering the unlikely potential of the consumer public suddenly becoming "risk say" any time soon, communication skill becomes ever more important, if we wish to communicate risk rather than sensationalism.
As Cliff has also pointed out numerous times, this is also key with communicating the risk of AGW, rather than sensationalism.
This guy, Gerd Gigerenzer of MIT, has done quite a bit of work in this regard. Here is his TED talk which is the "Readers Digest condensed version".
https://www.youtube.com/watch?v=g4op2WNc1e4
I learned a lot. I updated my supplies. I don't think you miscommunicated anything. The fact that you were able to foresee that storm at all is amazing. Just looking at the satellite images, it would have been impossible for a layperson to spot it days in advance.
ReplyDeleteThanks for all that you do!
Do you believe the uncertainties and variables of climate models are adequately communicated, understood and publicized?
ReplyDeleteIf you boil this all down, the message was "a potentially serious storm is coming. Take precautions." That was perfectly delivered by Cliff as well as the local news media.
ReplyDeleteThe problem is that people expect precision regarding timing and effects, no matter where they live in our diverse region. You want to bet that the (thankfully few) people who live on the west coast of Vancouver Island where the storm came ashore aren't complaining about "hype" today?
Everyone has a personal responsibility to make decisions about how to deal with uncertainty, but I do love the ones who scream: "I'll never believe a forecast again!" Here's hoping the future storm you choose to ignore doesn't drop a tree on you.
For those who complain that this illustrates that we should ignore models regarding global warming because they likely aren't exactly right, I would only say: "A a serious storm is coming. Take precautions." The difference is that with an ordinary storm, we have zero potential to change its behavior. With AGW, we have some limited potential to strengthen or weaken the storm. To those who ignore that potential, I hope a "tree" doesn't fall on your grandchildren.
The local media aren't as concerned about being right as they are about pulling in viewers, so whenever something like this (and snow!) is a possibility, they compete with each other like crazy to prove they are THE ONES with the best coverage.
ReplyDeleteIt's insane, it's embarrassing and it's irresponsible. But it has been this way since TV began and it will be this way until TV dies.
I appreciate your approach, Cliff. I only wish you posted more often. You are really the only source I trust.
As the models and predicted strength wobbled, while NWS hedged a bit, both my spouse and I commented that what might be useful for them (and for the local news) is to adopt the "Cone of Error" strategy used by the hurricane center. If you are in the cone of error, you should prepare, and could get some wind etc, but because it is called the cone of error, there is implied uncertainty in the track, and they tend to emphasize how much a small wobble in a big storm can change who gets affected. It seems pretty effective at getting people prepared, without a lot of accusations of "crying wolf" if the storm doesn't hit you as badly as predicted.
ReplyDeleteIt turned out pretty wimpy where I live (N. Bothell). Seemed less that Thursday-Friday's storm.
ReplyDeleteOh quit excoriating yourself. You did a very honest and meticulous job. If folks were misled then they need to work on their understandings of statistical prediction and risk analysis. I would much rather sit in my brightly-lit living room surrounded by flashlights and batteries and radios than spend a few days in the dark and cold and wet. Thank you for lending us your efforts and knowledge - I always learn something.
ReplyDeleteIt seems the TV media, and to some extent the forecasters, are trying to let themselves off the hook by saying 20 miles to the east and it would have been a disaster. I am very skeptical. I have not looked at detailed reports on Vancouver Island, but it appears that the worst winds at very exposed locations were far less than what was being predicted. I think everyone recognizes that path prediction carries a fair amount of uncertainty. But I think it is worthwhile to also focus on predictions about the strength of the storm seemed way off as well. Not being an expert, I don't know the reason for the error. Perhaps, the prediction assumed too much additional intensification.
ReplyDeleteI agree with SecretLittleSongbook. This event was not handled correctly. It amazes me of the excuses being tossed around. To go from a potentially MAJOR event to a non-event is all the proof needed to suggest a HUGE lesson learned here.
ReplyDeleteYes Cliff, you are the media. All the rose petals thrown your way does not excuse the results.
Having said all that I do enjoy your blog. Always always keep it professional.
On the subject of alarmists, let's focus on the NWS. I hope you are reading this NWS employees since every time I have emailed in the past you NEVER deliver a response.
Subject: Thunderstorm.
Being a former Seattlite and having transferred to the midwest, everytime a bolt of lightning is present a thunderSTORM is reported. There is a phenomenon where there is lightning in drizzle or rain. Never is recorded a Thunder rain or Thunder drizzle.
NWS, ask yourself or the public what they think of when the word STORM is used. I expect damage or tree branches or other debris on the ground after a storm.
At a minimum, something different than prior to this so called storm. The large majority of the time a storm never existed.
I am well aware of the variability in meteorology. In science and engineering we are to be exact. That is our profession.
NWS, you have work to do. Cliff, I have more faith in your creating change than those at NWS who consistently irritate their viewers with suggestions and NEVER deliver.
I disagree that the possibility for error was not communicated by the NWS. In fact I found their communications to be quite excellent and spot on. I followed this storm quite closely and read all of the Forecast Discussions because I was planning to be outdoors that day. They made it quite clear that slight variations in the track could mean a big difference in the outcome for Seattle. They were also spot on about when to place the warnings. They started at 3pm, allowing for the uncertainty, but stated the expected big blows would start at 8pm. I was out on the water when the 50 knot gusts hit at 6pm. The duration was only about 1/2 hour, then it died back to a more typical 30 knots. This is typically what happens when a front crosses. The NWS cannot predict the exact timing of this crossing, or of tracks, so their forecasts should be treated as a potentiality, not a certainty.
ReplyDeleteAs for poor communication, you only have to look at prior posts on this blog to see that Mass also failed in this regard, having hyped the storm like the rest of the newscasters. And now I find the criticism of miscommunication to be further hype from him.
Um, I work in media, and everyone in the news business reads this blog — closely. I suspect the folks at Sand Point got more than a few media queries that referenced this blog, seeking confirmation that the end was nigh.
ReplyDeleteCliff, your early blog posts were pretty sensational in tone, and journalists heeded that. Yes, you walked it back some Saturday morning, but it was less than clear even then, to me anyway, that this was going to be nowhere near as big a deal as earlier predicted. By Saturday at 4:22 p.m., you were pretty clear about the lessened danger. But that was quite late in the game.
At my house on south vashon, I recorded 5.47 inches of rain (thurs - Mon afternoon) and a peak wind gust of 36mph. The wind blew the tops off several alder trees but I didn't lose power! Friday morning's wind and rain flooded the garage floor. The storm's peak wind gusts were 20 to 25 mph less than forecast.
ReplyDeleteIt was windy and it rained a lot. It was stormy.
My favorite comment above came from 'sewhidbey'. Young people today don't understand how much weather forecasting has improved over the last 75 years. When my Dad was a teen, a tornado that today would be an F-3 or F-4 struck our home town without warning. Scores were killed and injured. In today's era of weather forecasting and media hyping those deaths and injuries would be halved or even less.
When I was a kid, a small F1 tornado following a moderate thunderstorm blew through our neighborhood. There were no tornado watches or warnings. I saw the funnel cloud when it was almost overhead and we ran to the basement. It sounded like a freight train roaring down the driveway. When we emerged, our house was surrounded by downed trees and broken branches. Our neighbor was sitting on his front porch enjoying the cool air when it hit. He was lucky, a strong tornado would have killed him.
This storm had the potential to be a whopper, lucky us, it stayed off-shore instead of riding up the Puget Sound. We should be happy not blaming forecasters and the media.
If your religious and believe in the power of prayer, maybe an act of God kept the storm at bay.
To all those who criticize the weather forecasts and media, ignore the media and weather forecasts, don't change your plans. See how that works out for you then decide.
Although this storm was not as severe as first forecast, I still wouldn't want to be caught sailing, hiking, camping or biking in it! Who knows, maybe the storm warnings saved a life or two.
Yeah I heard one in three chance and figured that was worth getting the batteries and radio together. I heard weakening in the days and time coming up. We had wind. We briefly lost power. I figured by the day of the storm that it would be at worst as bad as Thursday's storm. We went to bed with emergency kits together but feeling fairly comfortable that Bellingham wasn't going to be totally slammed. I'm so glad that the worst case didn't pan out, but I felt like most of the forecasting I saw was pretty straightforward for me.
ReplyDeleteAnd seriously it happens almost every year. Either a major weather event happens and everyone is upset because we weren't prepared adequately or it doesn't and everyone's upset they actually prepared. I'd rather be prepared and spared, personally.
This comment has been removed by the author.
ReplyDeleteEmergency Managers do not listen to the media and or bloggers. They listen to the NWS. Bloggers and the media do not have the responsibility of cause and effect for actiions we take. That being said, without that heavy responsibility, it's easer for them to take more liberty without consequences. I remember a particular fire weather forecaster from the NWS who's urgent message didn't make it out to the fire line. It was a contributing factor (1 of many) that lead to deaths. He is haunted by this lost forecast. The responsibility of lives and not reputation is what sets the media and bloggers apart from the folks in government buildings.
ReplyDeleteBack in the day the media was a direct conduit from official government agencies. People now watch news forcasters who take some liberty and are siloed instead of reading a true forecast discussion.
The Euro model had it right the whole time.
ReplyDeleteGFS wrong as usual
The storm was described as a remnant of a Typhoon. But, I wish there was a explanation of for how insignificant the Typhoon became in crossing the Pacific and then how it then gathered so much intensity approaching the West Coast. The satellite view of the storm out in the Pacific was hardly more than cloud cover and the pressure was at least 1000 mb.
ReplyDeleteIt would appear that about 10% of your 28 million page hits (from the beginning of this blog) came in the last week or so in relation to this storm. That in and of itself, is a very interesting tidbit in my mind...
ReplyDeleteGreat communication, nothing you needed to improve. But there is a serious need for wider public understanding of statistics and modeling. This is a math education problem, with great opportunities to integrate these skills across K12 science, social studies, health and technology courses.
ReplyDeleteI wish there was a explanation of for how insignificant the Typhoon became in crossing the Pacific and then how it then gathered so much intensity approaching the West Coast. Visit us Investment Fraud Securities Lawyer
ReplyDelete