The temperature traces over the past four weeks for Seattle and Pasco are enlightening...if not chilling (red and blue lines are average highs and lows, respectively). In Seattle, we have have been cooler than normal for the past two weeks, and the last few days were particularly frigid, with our high temperatures remaining below our normal lows.
Over Seattle right now the temperatures aloft are warming rapidly (red lines, this figure show changes above Sea Tac with time--increasing to the left on the x axis and height on the y axis)
What is going to happen now? The meteorologically enlightened prefer to use ensemble forecasts and a probabilistic viewpoint, so here are min and max temperatures forecast by the European Center model for Seattle. The range of possibilities is shown by the blue brackets. The ensembles are emphatic about a warm up (you can bet on that!), with highs reaching to near 48F in a few days. A slow cooling follows, but there is increasing uncertainty past a week. Enjoy the warmth...it will feel nearly tropical after what we had.
The Tri-Cities will have to wait a few days, as shown by the predicted temperatures at Hanford. But they will enjoy some thermal relief as well.
Do you predict a KING TIDE Monday, 12/19/2016 around 9-10am?
ReplyDeleteWeather forecasting is ridiculous. At least 2 times in the past month there have been forecasts of big storms, lots of snow, batten down the hatches, here it comes, blah, blah, blah. Then, hours later, suddenly the forecast changes. No big storm, a little snow, nothing to worry about. Here we go again. The forecast I follow, Weather Underground, has lowered the big snow on Monday, Tuesday to maybe 5". For the week, 8"or so. Just hours ago it was going to be 18" How can it be so far off??? Do you weather people error on the side of caution to cover your tails?
ReplyDeleteDid somebody call the wammmmmmmbulance?
DeleteCliff spoke on the blog recently about the difficulty in forecasting snow events in the Puget Sound region. As a rare ocurrence, the amount of snow depends on many factors which interact in near realtime,
ReplyDeletewhich creates uncertainty in their forecast models.
Mark-- The barometer is too high for a KING TIDE.
ReplyDelete@Tandrew, you might be misunderstanding probabilistic forecasting. That is, stating there is a ##% chance of [insert weather event here], given what we know X days in advance. Just because there is a decent chance of snow, doesn't guarantee snow. And the closer the event gets in time, the higher the certainty, hence the last-minute swings in probability and depth of snow fall that you mentioned.
ReplyDeleteThink about how the Trump/Clinton probabilities swung wildly on election night. Trump always had a significant probability of winning, although it was only 15 to 35%, depending on the source. When the one-in-three chance became reality, everyone blamed forecasters.
With up to an inch if rain forecast in the east Puget lowlands over the next 24 hrs, what about flooding concerns? I still have 6" of snow at ~500' on the North Fork of the Skykomish, so this fast warm-up is a bit alarming!
ReplyDeleteThere was a dusting in Olympia and more than a dusting outside of Oly this morning.
ReplyDelete@Tandrew
ReplyDeleteI've asked for a comment on that too. It seems weather forecasts generally start strong, then as the event gets closer the forecast continues to trend downward.
I can't think of a single storm that was both predicted, but also matched it's prediction in strength at the time of the event.
I can, however, cite many examples of storm events that were very strong and never predicted, or that were predicted to be strong, but ended up being downplayed right at the time of the event when nothing really happened.
Cliff,
Is it common for meteorologists to hype a storm prior to the event? Why do we see so few events actually happen as predicted, and also so few events that are predicted and have increasingly severe predictions leading up to the event?
Thank you.
One of the first things taught in atmospheric science is that the atmosphere is chaotic. Highly unpredictable. So anything outside 2-3 days is merely a craps shoot, where a large front or low can severely change its course/strength during that time. There isn't hype involved,at least with scientists, that is the news people. They always (at least cliff and NWS) state unpredictability and uncertainties they have in forecasts. Great example of uncertainty, the snow forecast for snoqualmie pass over the past four days was incredible, went from forecasting 20+" Sunday-Tuesday to 0" and now it seems snoqualmie will get a few inches. A single degree can change a lot, imagine forecasting snoqualmie.. I wouldn't want to!
DeleteOne of the first things taught in atmospheric science is that the atmosphere is chaotic. Highly unpredictable. So anything outside 2-3 days is merely a craps shoot, where a large front or low can severely change its course/strength during that time. There isn't hype involved,at least with scientists, that is the news people. They always (at least cliff and NWS) state unpredictability and uncertainties they have in forecasts. Great example of uncertainty, the snow forecast for snoqualmie pass over the past four days was incredible, went from forecasting 20+" Sunday-Tuesday to 0" and now it seems snoqualmie will get a few inches. A single degree can change a lot, imagine forecasting snoqualmie.. I wouldn't want to!
DeleteWhenever I hear complaints from people like Tandrew, I think of the 1940 Armistice day blizzard in my home state of Minnesota. My mother recounted the tale every November as she and my dad survived it while on a hunting trip.
ReplyDeleteThe preceding days were very warm and sunny, weather forecasts did not predict a blizzard. Duck hunters, like my dad, were dressed lightly. They didn't bring snow gear. The storm hit suddenly with 60 mph winds, the temperature dropped 50 degrees in about an hour, Minneapolis recorded 16 inches of snow, snow drifts to 20 feet and 5 foot waves on the Mississippi river. 145 people, many of them duck hunters, died in the blizzard, thousands more suffered frost bite.
A week ago, weather forecasters warned Minnesotans of a coming blizzard and -25F temperatures with wind chills to -45F. My home town received 5 of the predicted 8 inches of snow. Temperatures fell to -24F with wind chills to -40F. Mighty good weather forecasting! It saved lives!
Forecasting snow in the lowlands of western Washington is very tricky. There is a big ocean to the west and mountain ranges surround the lowlands. Fortunately for us, most big storms fail to materialize. But better warned than not warned of the potential for snow and wind.
If forecasters wrote there was a 20% chance of a storm, how many people would shrug it off and go hiking/boating? If the storm materialized how many of those people would complain that they were not warned!
Being a forecast meteorologist is a thankless job.
Being a climatologist is worse. These people look at the science and see AGW as very real. As do the people in Barrow, Alaska who are experiencing climate change first hand. But politicians from oil/coal/red states call them liars, hoaxes and bunk. It's not easy.
@Tandrew
ReplyDeleteYou should take a look at some of Cliff's previous entries on snow forecasting. It is very difficult to pinpoint all the required aspects for snow fall. He posted a solid entry from March 2012 titled "Why is it so hard to forecast lowland snow?" His first key point is the following
"You've got to get the amount of precipitation right--really right-- or you have a very bad snow forecast. On most days, the forecasts hardly talk about the amount of precipitation, just the probability of precipitation. Most people don't care whether .1 inch or .4 inch of precipitation falls...they would hardly notice it. But a difference like that would be HUGE for a snow forecast. Around here there is typically a 10 to 1 ratio of snow versus precipitation in the form of liquid water. Thus, .1 inch of precipitation would be .1 inch of rain or, if cold enough, 1 inch of snow. (Precipitation is always reported in terms of liquid water--so snow has to melted before measuring it). So the difference between .1 and .4 inches of precipitation, which you would be oblivious to of if it were rain, would be the difference between 1 and 4 inches of snow. That you would notice."
Another thing to keep in mind you aren't shown the uncertainty in the forecast when you are looking at a basic forecast on Wunderground, your local news station or most other places. Cliff wrote multiple entries about this too. Look for Feb 2009 entry titled "Uncertainty"
Changing the subject....
ReplyDeleteHow about the high wind/gale force winds warning at the Admiralty Inlet for today, the 20th.
Already gusting to 40mph.
Let's start with this: What is the purpose of a weather forecast? What decisions do you and others make based off of forecasts? I imagine most consumers of basic forecasts are using this information to determine what to wear or bring. In more extreme cases, they may be making decisions about what to do or where to go -- should I rain/wind proof stuff before I leave for the night, should I go hiking north or south, should I bring my snow chains. On balance, would you rather have prepared for the 20% worst case scenario, or been caught unprepared? Most people would rather bring the snow chains if there's a chance of a bad snow event, or hike the north end instead of the south end, or leave for the airport an hour earlier.
ReplyDeleteIf you're in a situation where this decision has significant financial impacts, then you need to upgrade your forecast from consumer-grade to professional-grade. The consumer "there is a potential for 8" of snow in the region this weekend" is not what a business should be using to decide if they reschedule $20,000 of shipments, or preemptively close their restaurant for the weekend.
One legitimate problem with these probabilistic forecasts is that so few forecasters with large audiences employ them. Consider the forecast for tonight (Tuesday evening): Weather Underground and NWS itself issued detailed, short-term forecasts that predicted significant rainfall during the overnight hours; neither the public forecasts nor the NWS forecast discussion indicated any issues with confidence in these predictions.
ReplyDeleteThis kind of variability - which involves short-term forecasts predicting unexceptional rainfall events - is itself quite noteworthy. Yet NWS, at any rate, did what it loves to do in such cases: Revised the forecast and carried on without so much as an acknowledgment, much less offering detailed discussion of what looks to me like a spectacular bust.
We hear quite a bit nowadays about the highly reliable nature of short-term forecasts. This may simply have been a normal outcome when talking about a short-tern, probabilistic forecast with, say, a 99% confidence level - such a forecast will bust once in a blue moon, and that's just the nature of the beast. Unfortunately, the forecasters themselves allow this to morph into a situation where a typical layperson sees something else entirely: a spectacular short-term forecasting "failure" combined with deafening silence from the people they hold responsible for the botched job.
The really annoying aspect, however, at least to me, is that there really might be something genuinely out of the ordinary happening to cause a bust like this. A proactive and transparent approach to probabilistic forecasting - an approach applied consistently to ALL forecasts, not just the "interesting" ones - would address such questions before they stir up the whambulance chasers and tin-foil hat crowd. But what we get instead seems to be exactly the opposite of those things, with predictable results.