Discussion > The New Law of Large Numbers
Little boys of a mathematical bent used to play at who could get their p-value highest up the wall. Then they grew up and became climate scientists.
There’s a graph of ocean heat content going the rounds which converts hundredths of a degree into thousand trillion trillions of Joules. Then there’s the Skeptical Science measure of sunshine in Hiroshimas per second.
Anyone playing this game is a charlatan. They should be called out on it at every opportunity.
geoffchambers
There's an example of the misuse of large numbers in an emotive and dubious comparison here.
Richard Betts
Joanna, I too have been contemplating the fondness of climate scientists for any number greater than 95, and amended a bad joke I thought up a couple of years ago accordingly:
Q. How many climate scientists does it take to change a lightbulb?
A. One to change the bulb, one to hold the stool (because there is bound to be a tipping point) and another 95 nearby so there is an overwhelming consensus that there has been a change.
lapogus
Richard - care to make any comment on the UN's now disappeared claim of 50 million climate refugee produced by climate change by the year 2010? http://www.bbc.co.uk/news/magazine-23899195
And it is a fact that thousands of old people do die of hypothermia (or require hospitalisation for malnutrition related diseases) each winter in the UK, because they have to choose between heating and eating.
lapogus
Richard Betts, like you I am not fond of these sorts of figures and this kind of argument (as I said in the comments). While it is certainly true that being too cold is worse than being too hot (on average), it doesn't strike me as a particularly productive discussion. It reeks of "rolling out the corpses" in a competitive way. Yuk.
That said, it can be seen as a reaction to "the new law of large numbers" being used by CAGW alarmists, such as when they make stupid (and unsubstantiated) claims like that 750 million people are under threat because of human-caused melting of Himalayan glaciers.
johanna
My question is, what is it about "climate science" and large numbers?
Traditionally I think it is a common sign when somebody, often a politician, is trying to convey a "message". Or simply to exaggerate an existing message. Bigger numbers have a larger impact.
My guess is that the tactic may often work on a significant fraction of the population, but that the more numerate and knowledgeable members of the audience might actually react in the opposite way... I know I do when I suspect this.
The use of inappropriate, or frankly stupid, units of measurement tends to confirm this. Hence some members of the green-climate-lobby like to use heat or energy measured in "Hiroshima bombs".
And the "missing heat" allegedly skulking in the oceans; Instead of representing it as a tiny fraction of 1 degree temperature change, it was converted into the terrifying joules of heat instead, allowing a number of the order ~10000000000000000000000 to be used to frighten the horses. Even if the reporter missed several zeros off the end, that number would still be just as likely to fulfil it's prime objective with the target audience. I regard it as duplicitous.
michael hart
How about the oceans absorbing 4 Hiroshimas/m^2/s? That's scary. Actually the sun emits 1.325*10^17 Hiroshimas/second which means the oceans are absorbing 1.2% of the Sun's heat/mm^2/second. Scary now?
geronimo
"There's an example of the misuse of large numbers in an emotive and dubious comparison here."
Whether we like it or not, numbers are the lifeblood of knowledge, I guess Joanna was referring to the multiple exaggerated numbers we get from the environmental movement, and their foot soldiers in the war against humanity, that we see on the BBC news every day. I don't know if the 7800 number in this article is exaggerated, certainly there will be a residual number that would have occurred regardless of the artificial rise in energy prices caused by the environmental movement. Let us say 80% would have died anyway, and rejoice in the fact that only 1560 people died from fuel poverty because of green energy taxes. Or better still let's say 99% would have occurred anyway and only 78 died because of the global warming scare. Does that make it better Richard?
geronimo
Geronimo, 4 Hiroshima explosions each second per square meter of ocean would be scary... think about it. I suspect you've not got your sum right. Here's my attempt: according to every scientist's favourite reference, Wikipedia, Hiroshima released energy of about 12.5 kilotons of TNT (52 x 10^12 J).
The figure I've seen bandied about for how much 'extra' heat might be going into oceans is of the order of 0.5 W per square metre. So 0.5 J every second. That would be the equivalent of about 10^–14 Hiroshimas per square metre of ocean per second. Less scary. Using Wikipedia again, I get a total of 3.6 x 10^8 square kilometres of ocean or 3.6 x 10^14 square metres. That would then be about 3.6 Hiroshimas per second for the whole of all the oceans - near enough your value of 4.
Jeremy Harvey
And let's not even get started on "Manhattans" or "Rhode Islands", which Anthony Watts informs us are favoured metrics in the US.
In Australia, we tend toward "football fields" and "Sydney Harbours."
The Hiroshimas thing is weird. Almost nobody even knows what it means, in quantitative (let alone relative) terms. But, Hiroshima is bad, OK?
johanna
Here in the UK, expanses of territory are measures in Wales'
TheBigYinJames
good talking point johanna
in my work we model components for aircraft structure which a CNC machine then uses to produce the component/part (very accurately, usually to +/- 0,2 mm).
in the old days (when m/c by an operator) we would work to maybe +/- 0,4 mm tolerance & fettle to suit on build.
for ref - 0,1 mm = 0.004 of an inch approx.
large & small numbers are meaningless to most people outside certain fields, when I hear billions spent on xyz I have some idea what this means by relating it to my expected salary for life, but even then I kind of know i can't grasp the full meaning.
big numbers or small, this is outside most peoples experience.
you can become an "EXPERT" in their eyes if you sound autharative enough & that's what's happened.
dougieh
Johanna, I think your distaste for large percentages of the 97% variety is because the number shows that most scientists disagree with you. If you got a figure like 97% of the public don't care about climate change, your impression of the numbers would be different.
Note that for men at least, large is almost universally preferred over small.
Chandra
Jeremy, you could be right, so I'll give you the numbers I used.
One Hiroshima is 60*10^6 joules so 4 is 240*12 joules.
Total output of the Sun is 360*10^26 joules
So the Sun gives out 360*10^26/240*10^12 4*Hiroshimas/second = 15*10^13 4*Hiroshimas/second
Divide this by total area of the oceans, 335,258 * 10^9 = 461 * 4 Hiroshimas/second/m^2
4/461 = 0.88%
So I appear to have overestimated a bit. But please check for consistency.
geronimo
Troll alert. DNF.
johanna
geronimo, Wikipedia agrees with you that the sun has a total power (or Luminosity) of 3.846×10^26 W, so that number of J per second leave the sun. But not all of them come to the earth! Near the earth, the sun's irradiance is at ca. 1,300 W/m^2, and the earth has a cross-section of pi * r^2 = ca. 1.3 x 10^8 km^2 or 1.3 x 10^14 m^2. So roughly 1,300 x 1.3 x 10^14 = ca. 1.7 x 10^17 J of solar energy are incoming to the earth each second - roughly 3200 Hiroshimas/second.
Jeremy Harvey
Jeremy Harvey: "So roughly 1,300 x 1.3 x 10^14 = ca. 1.7 x 10^17 J of solar energy are incoming to the earth each second - roughly 3200 Hiroshimas/second."
Not so! That is the flux density between sun and space at a distance of one astronomical unit. What is incoming to the Earth is dictated by its temperature and albedo. The only way to determine how much is coming in is to measure what is radiated out (assuming a system in balance).
ssat
Johanna: What do others think?
I think what they lack in quality they try and make up for with quantity. 95% is after all, nothing more than a try, and that by the most rabid.
ssat
Jeremy you're right I'd forgotten the albedo effect which should halve (roughly) the effects of the sun so 4* Hiroshimas is roughly 1.76% of the total hiroshimas from the sun. Either way the point is that Cook and Nuccitelli are trying to scare the bejasus out of us.
geronimo
TheBigYinJames (Oct 11, 2013 at 7:16 PM)
Here in the UK, expanses of territory are measures in Wales'sI’d always assumed that’s because 90% of Brits couldn’t give a tinker’s thingy about Wales. (“An area the size of Wales might disappear” = “What Me Worry?”)
geoffchambers
ssat, sorry, you're right. I think my earlier calculation, yesterday at 1:12, accounted for the 4 Hiroshimas/second, as propagated e.g. by Skeptical Science. In my next comment, I was replying to geronimo who was suggesting that the total incoming energy flux from the sun might be of the order of 4 Hiroshimas/second/m2, and did another sum to see. I was talking about incoming energy flux to the planet Earth at the top of the atmosphere - before things like albedo kick in.
On topic: the units you choose for sums can reflect convenience, desire to be understood, pedanticism, rhetoric (as in the desire to make a number seem very big/small), or ideology (as in the icky use of Hiroshimas/s by SkS). A lot of the last two in climate science.
Jeremy Harvey
As part of the innumeracy which has accompanied so much of "climate science", post-modern version, we constantly encounter very large numbers parading as proof. So, the utterly discredited and now withdrawn (sunk without trace by Jean S and Steve McIntyre) Gergis et al paper claimed that it must be right because they had done thousands of computer runs of the data. Similarly, AR5 claims to be believable because of the sheer number of papers reviewed by large numbers of people. This peculiar mode of thinking is crystallised in the absurd 95% certainty number, an echo of the equally ridiculous 97% of scientists meme.
My question is, what is it about "climate science" and large numbers?
Firstly, any fule kno that running crap through a computer thousands of times does not transform its qualities. Secondly, the suggestion that 95, or 97 percent of people agree on anything (apart from an empirical fact such as whether or not it is raining ... and even then ...) is only found in the lies of totalitarian regimes. It enhances understanding not a jot.
If I woke up from a long hibernation today, a la Rip van Winkle, people who used large numbers like these - indeed, who felt that they had to use numbers like these to support their views - would attract deep suspicion.
What do others think?