Reposted from the Cliff Mass Weather and Climate Blog
We are constantly exposed to extended forecasts in the media and online, with predictions extending through the next month and more.
Can you rely on such predictions? Are they really worth paying attention to?
Quite honestly, probably not–and if you do consider them, do so with the knowledge that their skill is marginal at best.
Take this month (October) for example. The official NOAA Climate Prediction Forecast for October temperatures, made on Sept. 19th, was for warmer than normal conditions over the west and MUCH above normal over the southwest U.S.
What actually happened? Nearly the entire west was much colder than normal, with the northern parts MUCH, MUCH colder than normal. A miss. In fact, a big miss.
Or the official 3-4 week forecast, made on October 4th? Warmer than normal over the west.
Such poor forecasts even a month out are not unusual. UW graduate student Nick Weber and I evaluated the skill of the main U.S. long-term forecasting model (the CFSv2) and found that skill is typically lost after roughly 2 weeks (see below and published in the peer-reviewed literature). This figure shows the forecast error (root mean square error) at 500 hPa—about 18,000 ft, a good level to view atmospheric predictability. The situation is the same over Washington, the western U.S., the continental U.S. or global. Skill is rapidly lost the second week out.
While meteorologists struggle to produce improved forecast skill past two weeks, we have gained a great deal of skill at the shorter time ranges, particularly for days 3-8.
So why is our skill improving rapidly for the shorter periods, but not the longer ones?
Because the forecasting problem is very different at the different temporal scales.
For the short periods, forecasting is an initial value problem. We start with a good description of the 3D atmosphere and our models simulate how things evolve. Because of weather satellites and other new data sources, our initial description of the atmospheric has gotten MUCH better. And our models are much better: higher resolution, much better description of key physical processes, and more. That is why a plot of the skill of skill of the 1-10 day forecasts of the European Center has improved greatly over the past decades (see below)
But small errors in the initial description of the atmosphere and deficiencies in our models inevitably lead to growing errors, and by 2 weeks such errors swamp the forecast. The forecasts are not much better than simply using the average conditions (or climatology).
There is hope for some skill beyond two weeks, by taking advantage of the forecast skill available from aspects of the environment that are changing slowly (such as sea surface temperatures, sea ice extent, snow extend, soil moisture). These aspects influence the atmosphere and potentially can torque the atmosphere one way or the other. Essentially, the forecast problem has changed from an initial values problem to a boundary-forced problem (the boundary being the surface characteristics that can influence weather).
But the skill that might be available from the boundary conditions is different—not about the conditions at a specific time, but for the average conditions over a month or season. A good example of such skill is the relationship of the warmer (El Nino) or colder (La Nina) temperatures of the tropic Pacific sea surface and weather around the world. There is some skill there, but it is relatively modest.
Unfortunately, our models still have key deficiencies (such as poor description of thunderstorms) that make it difficult for us to derive all the potential skill that should be available from the slowly changing boundary conditions. A lot of work is needed, but I am hopeful that eventually forecast skill beyond two weeks will improve.
We live in northern Virginia, not far from Washington DC. This area is apparently on the interface of two major weather cells. That interface can change location to the north or south very rapidly, and in an unpredictable manner. Eight to ten day forecasts for this region are surprisingly good in one sense, and very poor in another; the weather that they forecast will pass through in pretty much the way the weathermen say, but not exactly where they say it will. Storms are the most noticeable phenomena, and predictions of timing and amount of precipitation are generally good. It’s just that they don’t happen along the expected path, but a little north or south of it.
Weather forecasting is an art, one informed by science, but one ultimately based on judgement and experience. Meteorologists have been getting better at understanding the sources of uncertainty, and either try to mitigate them, or caveat their forecasts appropriately. But it is still an art.
I discussed this with two meteorologists from TWC back in 2002. Basically the computer “model” software that’s been used has NEVER been reliable and they have had to do stage (18 to 30 hour segments) with resets of input data between. This is why the predictions past 24 hours are almost always garbage.
It would literally be 190°F in Pittsburgh in 2 days if you let the software run because it was worse than garbage.
well even for short forecasts…what is the actual uncertainty?
short forecasts are useful..because they have been proven useful essentially empirically..
you have to make assumptions..such as the ensemble of models runs contains reality..
that s not even sure, certainly not proven.
My forecast:
More flaming wind turbines
https://www.google.com/amp/s/www.turnto23.com/news/local-news/kern-county-fire-working-to-put-out-windmill-fire-in-mojave%3f_amp=true
The result, off-topic.
Here in the Netherlands the forecasts are very detailed, suggesting high accuracy. But often the sequential forecasts for a particular day are as variable as the weather itself.
I’m curious as to whether the errors are randomly distributed around the actual value or have a bias?
I tracked the forecast for London from the BBC for a while, and there was a significant warm bias to the errors. Indeed, the longer range forecast tended to become cooler the close in they got.
In my opinion, the NWS uses the wrong methodology so they get erroneous long-term forecasts. My friend Joe D’Aleo and his colleagues at WeatherBell use a different methodology, and in my experience they get much more reliable long term forecasts. Joe’s long-term Winter forecast will appear in November. I believe I understand their general methodology, but it’s not up to me to give away their secrets.
Regards, Allan
This is the second paper I’ve written with veteran meteorologist Joseph D’Aleo:
THE REAL CLIMATE CRISIS IS NOT GLOBAL WARMING, IT IS COOLING, AND IT MAY HAVE ALREADY STARTED
By Allan MacRae and Joseph D’Aleo, October 26, 2019
https://wattsupwiththat.com/2019/10/27/the-real-climate-crisis-is-not-global-warming-it-is-cooling-and-it-may-have-already-started/
Our first paper was published in 2015 on Excess Winter Mortality:
COLD WEATHER KILLS 20 TIMES AS MANY PEOPLE AS HOT WEATHER
By Joseph D’Aleo and Allan MacRae, September 4, 2015
https://friendsofsciencecalgary.files.wordpress.com/2015/09/cold-weather-kills-macrae-daleo-4sept2015-final.pdf
Joe is one of the best meteorologists on the planet – this story illustrates how very competent he and the team at Weatherbell are, based on their strong predictive track record.
The U.S. National Weather Service (NWS) forecast a warm winter for 2014-15, and Joe called me in October 2014 to say he was concerned that the NWS forecast was seriously incorrect, and that the next winter would be particularly cold and snowy, especially in the populous Northeast. This was the second consecutive year that the NWS has made a very poor (excessively warm) Winter forecast.
Joe and I had been working together on a paper on Excess Winter Mortality, and we agreed that this incorrect “warm winter” NWS forecast was dangerous, especially if the country and its people were unprepared.
I proposed an approach, and we sent a presentation for my friend at the US Energy Information Administration (EIA). At the EIA’s request, Joe prepared his own monthly Winter Forecast by region for the EIA, who re-ran their winter energy demand calculations. Using Joe’s forecast, the EIA projected 11% more winter energy required for the USA than the “warm” NWS forecast had projected.
After that brutally cold and snowy winter, the actual energy used was 10% more than the EIA had projected using the warm NWS forecast, and just 1% less than Joe’s forecast projection. That is a huge amount of energy for the entire USA. I’m not sure if we saved any lives, but we definitely did a good deed.
Regards, Allan MacRae
Calgary
The ability to predict is probably the best objective measure of scientific and technical competence. Note that every scary global warming prediction made by the climate alarmists has failed to materialize. Nobody should believe them.
To heck with 10-day forecasts or even seasonal forecasts – here is a successful 17-YEAR forecast.
The last of my three climate-and-energy predictions made in 2002 has now come true. There it is – the perfect Trifecta – my work here is done.
Best regards, Allan MacRae
___________________________________________________
In 2002 co-authors Dr Sallie Baliunas, Astrophysicist, Harvard-Smithsonian, Dr Tim Patterson, Paleoclimatologist, Carleton, Ottawa and Allan MacRae wrote:
http://www.friendsofscience.org/assets/documents/KyotoAPEGA2002REV1.pdf
1. “Climate science does not support the theory of catastrophic human-made global warming – the alleged warming crisis does not exist.”
2. “The ultimate agenda of pro-Kyoto advocates is to eliminate fossil fuels, but this would result in a catastrophic shortfall in global energy supply – the wasteful, inefficient energy solutions proposed by Kyoto advocates simply cannot replace fossil fuels.”
Allan MacRae published on September 1, 2002, based on a conversation with Dr. Tim Patterson:
https://wattsupwiththat.com/2009/01/10/polar-sea-ice-changes-are-having-a-net-cooling-effect-on-the-climate/#comment-63579
3. “If [as we believe] solar activity is the main driver of surface temperature rather than CO2, we should begin the next cooling period by 2020 to 2030.”
Allan MacRae modified his global cooling prediction in 2013:
https://wattsupwiththat.com/2013/12/02/study-predicts-the-sun-is-headed-for-a-dalton-like-solar-minimum-around-2050/#comment-1147149
3a. “I suggest global cooling starts by 2020 or sooner. Bundle up.”
_______________________________________________________
THE REAL CLIMATE CRISIS IS NOT GLOBAL WARMING, IT IS COOLING, AND IT MAY HAVE ALREADY STARTED
By Allan M.R. MacRae and Joseph D’Aleo, October 26, 2019
https://thsresearch.files.wordpress.com/2019/10/the-real-climate-crisis-is-not-global-warming.pdf
RECORD LOW TEMPS UP TO 50 DEGREES BELOW NORMAL THREATEN TO ABSOLUTELY WRECK THE REST OF THE HARVEST SEASON
October 31, 2019 by Michael Snyder
http://endoftheamericandream.com/archives/record-low-temps-of-up-to-45-below-zero-threaten-to-absolutely-wreck-the-rest-of-the-harvest-season
MINUS 45 DEGREES IN OCTOBER? AN ARCTIC BLAST IS BREAKING RECORDS ACROSS WESTERN AND CENTRAL US
https://www.usatoday.com/story/news/nation/2019/10/30/arctic-cold-blast-breaks-temperatures-october-utah-wyoming-colorado/4098089002/
Forecasts, shmorecasts: I’m more concerned about how much rain (or snow) we’re going to get, because the local watershed is loaded nearly to the. 35 miles north of Chicago, Lake Michigamu is pounding the lake shores good and hard, and the rivers are full to the brim from rains to the north of me back in August and September.
The rivers, creeks, flood catchment zones – all of it – are full to the brim and there is little to no room for any more. Where there was flooding on the Chicago River’s tributaries last summer and fish were swimming in the streets, that hasn’t gone down yet.
I’ve asked this before and not gotten an answer: why is no one looking at the humidity levels in the atmosphere? That is as important as temperature? Humid summer: nothing unusual. But a humid winter? The snow that fell yesterday was large, wet flakes, sometimes in clumps. Snap out of it, people. Go look at the snow when it falls. Large clumpy snowflakes mean more moisture in the air. (Fewer upper respiratory issues, too.)
More moisture in the air more precipitation. Temperatures only determine what form it takes when it falls out of the clouds.
Sara – I saw your previous question but I have not studied this issue, so made no comment.
Thanks, Alan. I think it’s a valid thing to study. It does have a direct effect on the weather.
Correction for Sara:
I did study humidity here, in Section 6 of this paper.
CO2, GLOBAL WARMING, CLIMATE AND ENERGY
by Allan M.R. MacRae, B.A.Sc., M.Eng., June 15, 2019
https://wattsupwiththat.com/2019/06/15/co2-global-warming-climate-and-energy-2/
Excel: https://wattsupwiththat.com/wp-content/uploads/2019/07/Rev_CO2-Global-Warming-Climate-and-Energy-June2019-FINAL.xlsx
[excerpt]
6. The sequence is Nino34 Area SST warms, seawater evaporates, Tropical atmospheric humidity increases, Tropical atmospheric temperature warms, Global atmospheric temperature warms, atmospheric CO2 increases (Figs.6a and 6b).
Other factors such as fossil fuel combustion, deforestation, etc. may also cause significant increases in atmospheric CO2. However, global temperature drives CO2 much more than CO2 drives temperature.
Think about it. The wx models go screwy after a few days. Hurricane predictions have *wide* path uncertainties and landfall is unpredictable until landfall is actually happens. Both of these are based on predictions from “initial conditions” that are measured quite precisely. None of these models can predict weather a year into the future.
Yet the climate models claim to be able to do so. The prediction for the end of Year0, based on initial conditions at the start of Year0, becomes the input for the start of Year1, etc, etc, etc.
We continually hear the excuse from the climate alarmists – “weather is not climate and climate is not weather”. And yet there is a *direct* relationship between the two. When it is wet it is cooler, when it is dry it is hotter. This is all based on things like albedo, evapotranspiration, humidity, etc. If the climate models can’t get all of this right then no amount of parameterization and fudge factors can fix them. And as Pat Frank has shown, the models are all nothing more than a complex way of representing a linear equation, y=mx+b. And yet, as Richard Courtney demonstrated so well, the Earth’s climate has long term cycles associated with it – none of which are adequately represented in the linear equation outputs of the models.
If you can’t trust the weather models past about 10 days then there is no reason to trust today’s climate models all. The physics are just too complicated to reliably model using our present knowledge level.
“The wx models”
Why is weather abbreviated as “wx”? I don’t get it.
When I was doing high-speed morse code on the amateur radio bands 50 years ago, wx was the abbreviation used for “weather”. I suspect it came from usage on the old landline telegraph systems 150 years ago. It’s the same kind of abbreviation as TX for transmitter and RX for receiver.
Tim
And “DX” for long Distance.
Joe Bastardi points this out all the time. EVERY long-term forecast is mostly above avg. Only until the time comes very close (couple weeks away) does it change to a more reasonable forecast.
Rhetorical questions:
What does this say about trying to forecast even beyond a month or two (like decades into the future)?
Why are forecasts beyond a couple weeks ALWAYS too warm?
How can the model-sychophants not acknowledge these severe/obvious model-issues?
So the world may not end in 12, 100, 5000 or more years. 12 years was a weather forecast or was it a wish?
I still rely on the forecast even if it is not always right at my location.
I’ve been noticing this for years. Notice how the longer range forecasts rarely ever show large swaths of the country as predicted to have below or much below average temperatures. The forecasts are almost always for above average temperatures. This despite the fact that significant cold events still occur on a regular basis (usually explained by a “polar vortex” due to climate change…).
This is called the perfect positive feedback!
Precisely the factor at play both in the CO2 “tipping” point trope as well as in the herd mind of its adherents.
Outside of that, positive feedback is rare in natural processes and a trap for buying into the CAGW mania.
Environment Canada provides the forecasts I usually depend on. Current Conditions are usually pretty close but not perfect as I live 25 miles from either local weather stations.
Summer thunderstorms can develop within just a few miles in this area so are impossible to predict accurately. They can only state possibilities.
Forecasts of highs and lows are usually only somewhat accurate for the current day and the next.
We are on the prairies in the lee of the mountains so the difficulties of accurate predictions are immense.
That does not mean they do not have value as very large weather systems can be fairly accurately tracked and future movements predicted. The timing is often considerably off, however.
They provide real time satellite and radar imagery and with a little practice we can use them plus a knowledge of local conditions to be reasonably accurate in predicting conditions in our own locality for a few hours. Usually that is all you really need to be prepared for what the weather brings.
As for long term climate models, I cannot see how there is any possibility of them being accurate. I think it is going to get colder because white men are building warmer houses.
I did not install my window AC this year because I predicted a cool summer. Was I ever right. I think the same trend will continue and we will have a cold winter. I put my winter tires on a month earlier than usual and was I ever right again. Three snowstorms before Halloween. It would have been nice to have been wrong.
My prediction for tonight,(after having pulled two teeth today) tonight is going to be a long one possibility of a anomolus high temperature certainly uncomfortable, the next few days will be sore with periods of relief from a mixture of fluids mixed with Panadol,with experience the forecast for the next few days is pointless with the ever threat of a high pressure knock to the side of my face resulting in a chaotic unpredictable few days.
You’ll excuse me if I dont care about the weather for one night at least.
50% chance of precipitation means that if 100 things fall out of the sky, 50 of them will be rain. 🙂
My long range solar based prediction for the Arctic Oscillation, was AO turning positive from the 2nd October, turning negative from the 16th October, and turning positive again from the 27th October. November is negative from the 5th, turning positive from the 15th. December strongly negative from the 12th, turning positive from the 22nd. And watch out for the very cold February, especially the first half.
https://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/ao.shtml
In business, we have known about the forecast trumpet, sometimes called the “trumpet of doom” for some time.
Bottom line. The further out the forecast, the more it is unreliable.
https://www.linkedin.com/pulse/forecast-trumpet-doom-david-j-joe-armstrong/
Nice article by Cliff to clarify something many folks don’t understand.
As an operational meteorologist for 37 years, I’ve made over 10,000 weather forecasts. The first of many lessons learned was to not get married to your forecast or the model output it’s based on. On tv the first 11 years, that meant not being overly confident in many situations that you were fairly confident about but could see the outside chance of being wrong with.
When/if you are right 9 times out of 10 people will remember the one time that you were wrong.
Fresh out of college and feeling over confident in models and my ability to interpret them……and the human weakness of having cognitive bias, i hung my hat on a few forecasts that in later years would have been revised much quicker.
After experiencing the real world, the objective is NOT to be right. The objective is to recognize when you are wrong ASAP or even might be wrong…..then adjust the previous forecast to dial in the changes.
Be very humble. When you are wrong, which happens a lot, you have to be careful about explaining why. The real reason is almost always because you believed the weather models. If your explanation seems to blame something other than yourself, it can come off the wrong way but do try to explain it vs pretending it didn’t happen or acting like you were right if partly wrong.
The best meteorologists are experts at paternities recognition. I note in today’s world with models much improved that almost all local NWS forecasts are just regurgitating hourly numerical weather output for that spot…….even going out days in advance.
Something like “rain likely between 4-7am, then chance of rain between 7-10am, followed by a chance of showers and a possible thundershower.
I understand that this gives some value to people who want to know exact times if they have an outdoor event but often, it suggests our ability to time weather with much mote precision than possible.
A forecast like that, which is common, tells me the forecaster is only passing on the model data and may not be actually looking at the pattern and recognizing some things that can assist him or her in beating the models.
Different models have different bias and most models will vary from the other ones, especially in the week 2 period.
Giving the numerical output from just the US model in a forecast 7 days from now is a recipe for a lot of busted forecasts.
Looking at the overall pattern, spacing between large scale features, the value of indices of things like the NAO or AO, consistency of solutions and between models are all considerations.
Despite the difficulty of getting it right after 1 week, there are plenty of fairly high confidence week2 forecasts. Not with exact timing of a storm on day 10 for instance….with rain but like what cliff stated…..the average and very likely chance for it to be wet or very warm during that period. It may rain on days 8-9-11 and temperatures 10 degrees above average but your daily forecast for rain on day 10 will be wrong.
If day 10 is a Saturday and somebody has a huge event planned outside, that may be the only day they care about.
You could come in the next day and tell them that it will rain that day and not the next.
Despite the implied weather model weaknesses, weather models are the main reason that meteorologists can get most of the weather right for the first week display enough skill to be useful out to 2 weeks. In some cases, even beyond that.
I have enjoyed every hour of every day analyzing developing weather patterns since 1982….all thanks to weather models.
It’s like having a magic crystal ball that allows us to peer into the future and see weather patterns, features and the resulting consequences that will happen in form of the wind, snow, temperature or other things affected which we will experience.
And every new day results in a fresher sampling of initial conditions and an opportunity to fine tune yesterday’s forecast. The more extreme the weather pattern, the bigger the adrenaline rush in appreciating the enormity of the physical laws and powerful forces at play.
Despite loving the most extreme weather patterns the most, our job is to do the best possible in alerting people to the risk of high end, life threatening events……which sometimes can be seen in a week 2 forecast period with a high risk outlook.
Heat waves and cold outbreak risks can be seen with high risk and decent confidence for a large area in week 2. They are huge air masses that may be over 1,000 miles wide. They can change a bit but the forecast often is still good.
Excessive rain or severe weather events are tough to pinpoint at that time frame because they may be targeting a more limited geographic area…..with small changes having big impacts.
Hurricane forecasting skill has improved a great deal but obviously the skill drops off fast after the first few days.
Speaking of extreme weather. I have not observed an increase in most of it the last 4 decades. In fact, warming the highest latitudes the most and decreasing the meridianal temp gradient has decreased some extreme weather as in violent tornadoes.
However, the atmosphere holding more water vapor has increased high end rain events and amounts. Heat waves in some places are slightly worse too….dough. If the planet is warmer this has to be true.
Regardless as somebody that estimates crop production/yields and energy use for heating and cooling for a living, the benefits from the modest warming and especially the massively beneficial increase in CO2 and in other areas outweighs the negatives for most of life on this greening planet during the current climate optimum…..by a wide margin.
I have spent the last 37 years observing it.