Guest post by William M. Briggs professional statistician
It is statistically appropriate to point to this year’s frigidity as evidence that the theory of man-made global warming is suspect.
Sure is cold out there, unusually so. By “unusual,” I mean the temperature is on the low end of the observed temperatures from previous winters.
Of course, we don’t have any more than about 100 years of reliable measurements, so it’s possible that the freeze we’re experiencing now isn’t as unusual as we suspect. But, anyway, it still sure is cold.
If you recall, a lot of global warming models predicted it would be hot and not cold, and to risk redundancy, it sure is cold. Does this dissonance between the models’ predictions and what is actually happening mean that those models are wrong?
No. But it sure as ice doesn’t mean that they are right.
Here’s the thing: No matter how cold the winter is, no matter how much snow falls, the global warming models will not be disproved. In technical language, they cannot be falsified by the observations.
Another way to say this is that the winter we’re seeing is consistent with what the models have been predicting. Again — does this consistency mean that the models are right and that the theories of man-made warming are true?
No.
Consistency is such a weak criterion that almost any imaginable theory of climate will produce predictions that are consistent with observations. The term is probabilistic: It means that what actually happens had to have some chance of occurring according to a model. If global warming climate models said, “It is impossible that this winter will see temperatures below X,” and temperatures did, in fact, drop below this threshold, then the models would be inconsistent with the observations. The model would be falsified.
But global warming climate models never make statements like that. They say that any temperature is possible, even if this possibility is low. Certain temperatures have probabilities as low as you like, but they are never precisely zero. (To anticipate an objection: “that number was practically zero” is logically equivalent to “she was practically a virgin.”)
Man-made global warming is just one of many possible theories of climate. Another is the Business-as-Usual Theory (BUT), which states that whatever happened last year will more or less happen this year, and so on into the future.
The winter we’re seeing is consistent with the BUT, which like the man-made global warming theory, never says any temperature is impossible. Further, BUT is corroborated more strongly by this winter than is the man-made warming theory. BUT’s predictions are closer to what we actually see.
“Stop right there, Briggs! You’re making the classical mistake of confusing weather with climate. The global warming models make predictions of climate and not weather. This winter doesn’t mean anything!”
I am not making that mistake, and it is you who are confused. Weather is climate. More specifically, aggregations of weather are climate. Means, averages, and distributions of daily weather comprise climate. That is, climate is a statistical phenomenon and depends for its existence on defining a reference time frame.
For instance, if “climate” is defined as the yearly mean temperature, then this year’s cold winter will produce a yearly mean temperature that is colder than average (as long as the coming summer isn’t abnormally hot: winter, of course, overlaps two calendar years and a hot summer can balance out a cold winter in the yearly mean).
So it is appropriate to point to this year’s frigidity as evidence that the theory of man-made global warming is suspect. If “climate” is defined as the decadal mean temperature, then this year’s cold winter will push the decadal mean lower. And it is still acceptable to point to this year’s winter as evidence against the man-made global warming theory.
Just as it was appropriate when the media trumpeted each and every “record setting high!” as evidence for that theory.
The difference is that one day’s temperature has little influence on a yearly mean — it is just one out of 365 other numbers that make up the average. One day’s temperature is thus weak evidence for or against any theory of climate.
But a slew of months with higher- or lower-than-average temperatures will push that yearly mean higher or lower. A season’s mean temperature is stronger evidence for or against any climate theory than is a day’s.
Back in the 1990s, when the yearly mean temperatures were increasing, this was touted as evidence for the man-made global warming — but those years’ temperatures also corroborated the Business-as-Usual theory. Which theory was better?
For the past decade, we have had a string of years with mostly decreasing temperatures. This is strong evidence against the man-made global warming theory, but pretty good testimony for the BUT. So far, the BUT theory is winning on points (there are other climate theories the BUT doesn’t beat). This doesn’t mean that BUT is true and that the man-made global warming theory is false, but it does suggest that this is so.
You can’t have it both ways. It is a mistake to extol evidence that supports the man-made global warming theory and to cry foul when presented with evidence which weakens that theory.
That so many do this says more about their desires than it does about any theory of climate.
Reposted from: http://pajamasmedia.com/blog/actually-weather-is-climate/
with permission from the author. Visit his website: http://wmbriggs.com/blog/
Mr Briggs you left the barn door wide open. I have to drive my truck thru it.
“(that number was practically zero” is logically equivalent to “she was practically a virgin.)”
Was the number derived “digitally”?
For George E. Smith:
“..Anybody knows that the incoming solar TSI is more like 1366 Watts per m^2 and not any 342 number which Trenberth asserts.
And no 1366 W/m^2 time shared over the surface does not produce the same effect as 342 received 24/7 by each square metre of the earth from pole to pole.”
It’s pretty simple. And the theory matches the measurements. It’s not, on average 340 W/m^2
The TSI is received against the “2d radius” of the earth, so the total energy incoming is:
E(total) = S x pi x r^2
where S is the solar “constant” of 1367W/m^2 at the distance of the earth from the sun
r is the radius of the earth
Some is reflected out by the clouds, earth’s surface – the planetary albedo, A – which is about 30%
And the 70% received in is shared out (unevenly) over the area of the earth which is 4 x pi x r^2
Actual average shortwave solar radiation absorbed by the earth’s surface =
S/4 (1-A)
= 240 W/m^2
The actual number at any one time depends on real angle of the sun at the particular point so is dependant on latitude, time of year and time of day.
continuing.. for George E. Smith:
I’d be amazed if Trenberth had got such a basic number wrong- what are you citing?
The averaged value across the globe is 240W/m^2. At a given point at a given time, it is quite possible to be 340W/m^2.
Good Morning Mr. Briggs:
Your mission, should you decide to accept, is to take several of the “Average Temperature” data sets, to analyze them in terms of their “standard deviation” over time.
You are to show that changes over the last 10, 50, 100, 150 years DO NOT FALL
far enough outside of 2 SD’s to have ANY statistical significance.
Should you, or any of your “SM” (Statistical Methods) team be denounced, or derided, you may safely dispense of your critics as complete statistical morons.
Good Luck Mr. Briggs, this Email will Phil Jones delete in 5 seconds…
5
4
3
2
1
Psssstttt!!! —-
“Beware of Greeks bearing gifts” and statisticians without data.
Briggs writes: “For instance, if “climate” is defined as the yearly mean temperature …”
We know that weather is going to change in a fairly predictable way as the earth circles the sun. These SEASONAL changes do not represent changes in what scientists call “climate”. One year’s average annual temperature is a single data point that tells us almost nothing about “climate”. From a statistical point of view, if we don’t know anything about the natural variation in average annual temperature, we can’t say anything meaningful about how average annual temperature is changing.
Briggs does better with: “If “climate” is defined as the decadal mean temperature, then this year’s cold winter will push the decadal mean lower.”
With ten years of average annual temperature, we can say something meaningful about the mean and standard deviation of the average annual temperature and therefore talk sensibly about how the climate is changing. So how cold has “this year’s cold weather” been? What does Briggs mean by “this year’s cold weather”? Briggs appears to be referring to the cold weather shown on the map of the Northern Hemisphere for December of 2009 (without any temperature scale). December 2009 represents about 0.4% of the data needed to compute the global climate for a decade (120 months times two hemispheres). If the blue color on the map represented December temperatures that were an absurd 5 degC colder than average and that blue color covered a remarkable 50% of the Northern Hemisphere, the mean land temperature for the whole decade would be reduced by only 0.01 degC by the December data. (Current data shows global warming averaging about 0.2 degC per decade between 1975 and 1995 and IPCC projections of future climate change are at least this big.)
If “this year’s cold weather” means December 2009 weather, Briggs IS making the classic mistake of confusing weather and climate. Confusing weather and climate is wrong when either the alarmists or the skeptics do it.
If “this year’s cold weather” means a longer period than December 2009, where is the data? Compared with the rest of this decade, every reader of this blog should already know that 2009 was not an usually cold or hot year. The decade’s climate isn’t changed by a normal year. Compared with the previous century’s or half-century’s climate, 2009 was unusually warm.
Let me put it another way:
Black walnut wood is beautiful. Back in the early 80s, When weather changed, and the permafrost layer marched south during the late 1970s, the “climate zones” in North America shifted south. Many Black Walnut trees growing near St Louis, Missouri were killed by the “cold-line” moving south. A local saw mill harvested these “southern” trees, and processed their wood into T&G flooring, and I did my whole house in the hardwood floors.
Why did the “climate zones” — where trees grow — shift, if not a climate change. Well, in the late 80s and 90s, the climate zones shifted north. Now, in the “20-teens” we are seeing the climate zones shift south again.
Why aren’t moose wandering around in Palm Springs? Why do southern pine trees not grow in Connecticut with the northern pine species? It’s the climate, stupid!”
Yeah, “weather isn’t climate” but long-term cold weather moves the climate zones south, and long-term warm moves them north.
BTW, since we are warmer than the Little Ice-Age, isn’t that a good thing? We are not as warm as the MWP. Is that bad? Why is it that, even though the AGW crowd has yet to prove that rising CO2 forces rising temps, that they get a free-pass on their assertion that “change is BAD!” What a normative load of BS propaganda. Are they saying that the 1970s were the most perfect for the planet? Was it the 1980s that were “perfect.” Please tell me so I can laugh out loud!
Newt Love (my real name) newtlove.com
Aerospace Technical Fellow: Modeling, Simulation & Analysis
“The most famous person nobody has heard of!”
And nevertheless, the UAH global January temperarure is breaking all records… So if a “frigid” earth (which is not happening) would be a sign of the planet going colder, is a “sizzling” earth a sign that warming is actually well confirmed?
I think this is the only time I’ve ever seen “Business as Usual” defined as a climate model in itself. Everyone else uses it to describe an unchanged rate of human GHG emissions. There are them climate models based on BAU emissions, but they certainly don’t say “whatever happened last year will more or less happen this year, and so on into the future”.
Climate versus Weather
Perhaps this is a false dichotomy: ‘weather’ is the steady accumulation of individual numbers (days and their particular conditions) while ‘climate’ is a category decision (‘cooling’, ‘warming’). An old philosophy conundrum is ‘How many hairs do you have to pluck from a man’s head before you can call him bald?’ To decide when the climate is category ‘cooling’ or category ‘warming’ a particular threshold has to be decided beyond which the definition is moved from one to the other; this is achieved by the steady accumulation of individual ‘weathers’ (hairs). The decision to be made is simply at what figure do you set the threshold? The figures bandied about that I have seen are so small they seem to enter into the pettiness of Swiftian ‘Big Endians’ versus ‘Small Endians’. Without a definition of where the threshold lies this pettifogging (weather pun) controversy will run and run.
OMG… but…but… if we can’t draw a clear distinction between weather a climate then all of Stephen Schneider’s precautionary principle doomsday “insurance policy” scams/metaphors will become forever muddled and meaningless. Say it ain’t so!
I never understood why the fact that a particular day, say January 20th,was the hottest or coolest in 150 years was worth talking about, and might even make the headlines on the evening news.
It strikes me that a record for a running 31 day average would make a lot more sense and provide something to lead the nightly news with.
Davidmhoffer,
“If we use temp of earth average at 300K, and get a 3 degree rise, earth radiance would go up by 18 w/m2? But IPCC estimate is CO2 doubling = 3.7 watts, tripled by H2O = 11.1 watts. so unless they are claiming another 7 watts from somewhere else? Or I am missing something else?”
Yes, you are missing the fact that in order to get a 3c rise from 3.7 watts, the climate models incorporate positive feedbacks which supposedly add to this value.
scienceofdoom,
“Actual average shortwave solar radiation absorbed by the earth’s surface =
S/4 (1-A)
= 240 W/m^2
The actual number at any one time depends on real angle of the sun at the particular point so is dependant on latitude, time of year and time of day.”
This is true, but it is important to remember that the actual value received at specific points on the earths surface will vary enormously. At night there will be zero insolation, but during high noon in the tropics on a clear day the value will be close to the maximum – probably > 1000watts.
When you consider this variation taking place all the time – zero to 1000watts and everything in between – going up and down each time clouds come and go, then the idea that you can somehow cause the planet to overheat by adding a couple of watts to the mix, is somewhat fancifull.
Yes, you are missing the fact that in order to get a 3c rise from 3.7 watts, the climate models incorporate positive feedbacks which supposedly add to this value
I included that. CO2 doubling = 3.7. Positive feedback from water vaper = 7.4 watts. IPCC mentions other feedbacks, NONE of which are in the same order of magnitude as those two. Now earth isn’t an ideal black body, but with radiance increasing commensurate with temperature raised to power of FOUR…. the negative feedback from plane old physics ALONE is larger than the temp change the IPCC claims is possible. so either I missed something else, or their claim is WAY over stated.
Are you looking at the graph for the DMI temperatures for the last 10 years or so?
I wonder as the area of the curves above the green line and below the red line has been consistently greater than the area of the curve above the red line and below the green line for the last 10 years or so.
This information is linked on this website.
It shows evidence that the north polar region is warming.
Excellent article Mr. Briggs. One small correction though, where you say, “Weather is climate.” Weather is NOT climate, because climate does not exist as a separate entity. Climate is merely the word we use to identify the average of weather. The fact of the matter is you can NOT see climate. You can not see or measure climate. You can only see AND measure weather. You can then average your measurements and call the average anything you want, i.e., “climate” in this case, but whatever you call it still does NOT exist as a separate entity. It IS an average. Measure it again, and you have a NEW, DIFFERENT average. Additionally, as “weather” by anyone’s definition consists of many components, only one of which is air temperature, if all that you are measuring IS temperature (by what ever means) at relatively few points in the total atmosphere, you don’t even have a close approximation of “average” global weather or, if you want to name that figment of our collective imaginations something, global “climate”.
Just as I thought climate science comes out of the BUT
Good work Briggs, I guess your not bored after all.
Could we back off from the “climate is an average” notion. If you look up the common labels for types of climate, such as Mediterrian or Marine West Coast, you will find a very interesting set of “patterns” – such things as dry summers, wet summers, cold winters, mild winters, and so on. A “hot summer” doesn’t become non-hot because someone has calculated a global average temperature (from garbage data) that is a half a degree larger or smaller than 50 years ago. Show me that the yearly pattern of weather is different in Athens, Greece (or Athens, GA), or Forks, Washington than it was 50 years ago. This monstrosity of a number called global average temperature anomaly, to be of any use at all, would have to be based on good data. No one has such.
“davidmhoffer (06:13:15) :
[…]
the negative feedback from plane old physics ALONE is larger than the temp change the IPCC claims is possible. so either I missed something else, or their claim is WAY over stated.”
Exactly. The small extra wattage induced by man-made GHG rise completely drowns in the natural fluctuation.
‘This monstrosity of a number called global average temperature anomaly, to be of any use at all, would have to be based on good data. No one has such.’
Are you forgetting the University of Alabama in Huntsville?
The Merriam Webster online dictionary defines:
Main Entry: re·ify
Pronunciation: \ˈrā-ə-ˌfī, ˈrē-\
Function: transitive verb
Inflected Form(s): re·ified; re·ify·ing
Etymology: Latin res thing — more at real
Date: 1854
: to regard (something abstract) as a material or concrete thing
“Climate” is a reification of Weather. Weather is real. The atmosphere is real. Solar radiation is real. Evaporation of water from the oceans is real. A whole host of other processes are real. They produce Weather on an ongoing basis. Making measurements on the Weather and computing statistics on the measurements and choosing among the statistics which ones you think are important gives you Climate.
Good post. Don’t discount the laymen. We know the CO2 from burning goes someplace, and I think the isotope signature shows the increase to be coming principally from burning, but, having said that, there are undoubtedly places where rising ocean deep water, rich in dissolved CO2, becomes warm at the surface and releases CO2. So a trend of additionally cold and CO2 rich water reaching the surface might have some bearing on the trend of observed CO2 concentration, yet I doubt it is very significant.
There are many people who argue that the amount of CO2 we produce through burning is small compared to the seasonal variation thus it matters not, but this isn’t a very pertinent point–no one can deny that we produce a fair amount of “new” CO2. The ultimate pertinent point is…does it all matter enough to justify the alarm and expense of rescue missions?
The issue might be that the Stefan Law does not apply in this case. The LWIR goes out though a narrow “water vapor window” and in order to get a temperature to the fourth power factor for emitted power, you need the full spectrum. Depending on how the spectrum peak fits within this window or not results in effects ranging from a temperature to the 4.6 power to almost no effect at all.
Dr Anthony Fallone (17:32:02) :
I know of factor analysis in social sciences and biology although I have never engaged in it. It seems the difference is that we engineers identify the factors with specific physical causes, or interactions, and you social scientists can see the influence of factors in the data, but may not be able to connect them with a specific physical effect.
We can learn a lot from one another, though, can’t we?