‘Global Temperature’ — Why Should We Trust A Statistic That Might Not Even Exist?

[firstfloor]

Maybe - but that doesn't tell me that the mean is a valid measure for this.

[Teallaura]

FF2, I understand how the mean is used. Now, let's look at your example.

Robert Wadlow and Charles Stratton are both included in your sample. Now, what does the mean tell us about height? Or we simply use your wording - infants and giants are included in the sample. Numeric superiority won't save it - you got three hundred babies in a sample of 1000 - or any number at that proportion (random sampling doesn't guarantee no skew - just makes it less likely) - and the mean is now meaningless.

Oh, but we can weight it. Sure - you (general) can put your thumb on the scale - but won't you select for what you think is 'normal'? Yeppers - weighting is vulnerable to bias (see 2016 election polling). So is it really telling us something about human height - or something about your expectations of human height?

Well, we use a distribution curve ... that we got by averaging heights... Oopsies...

No? Great, Japan has a sampled distribution curve... Uh...

Better plan - skip the mean and use the median - THAT tells us something useful about the distribution where the mean does not. [Yeah, it's still gonna be skewed but I can fix that without weighting - categories for the win!]


Look, if you (general) don't know why you are using a exceedingly basic metric like the mean before you EVER plug it in, you did the whole thing wrong. I might (stress might because this is BAD form) accept that some metrics are going to have to be plugged in before you know if they work as expected - but it's a hellishly hard case to make - and even harder to prove that it's not mere conformation bias at work.

A metric, like the mean, is already well defined. Any statistician worth his salt knows why he is choosing it in the analysis of his data LONG before he has begun compiling that data. At the very least, he knows the reason why he's chosen it before using it. This 'using it to confirm it' stuff is exceedingly poor methodology.


We've been at this for pages. I do appreciate that you took this on and did so fair and square (unlike some externally seeded fruits I know. ) but I don't see a point in continuing. I don't know how to make the question any clearer and you don't seem to know how to answer it (that's not a dig - you did try and you weren't screwing around doing so - not everyone knows everything and heck, it may just be that I can't get the idea across well enough - or we're both losing track and talking past each other ).

Anyway, I appreciate the conversation! Catch you later!

[firstfloor]

Good information here:
https://climate.nasa.gov/vital-signs...l-temperature/

[Joel]

Name a complex system where someone other than you thinks using a mean isn't valid.

For one example, economists don't think it's valid to average prices to get an average price level.

This isn't just what I think. This is standard, textbook economics. A quick google search yields lots of results explaining the many problems. E.g.
http://textbook.stpauls.br/Macroeconomics/page_92.htm
https://www.investopedia.com/article...priceindex.asp
https://www.economicshelp.org/macroe...ems-measuring/
https://www2.econ.iastate.edu/classe...0Notes%206.pdf

For a taste of the problems:

It is not mathematically valid to compute the mean of the price of apples and the price of oranges because the units are wrong. You'd have to introduce arbitrary coefficients (scale factors) to convert them to a common unit.

To make a price "index", someone chooses a "basket" of goods (let's say 2 apples and 3 oranges), and watches how the price of that basket changes over time. But the choice of basket of goods is arbitrary (it's a form of choosing arbitrary coefficients in the form apples-per-basket), and usually not all goods are included. People have tried to make it more objective by trying to study what basket of goods an "average consumer" purchases. This pushes the problem back into new problems of how do you validly average the baskets of goods of various individuals. And even if that were possible at a particular point in time, the "average basket" will likely not be the same at any other point in time, so it doesn't measure the change in average price. Again, even if it were possible to calculate the average basket, it would be meaningful only to the "average consumer", who may not even exist. The resulting index would be less meaningful to each individual as that individual varies from the average basket. It is more meaningful (and simpler) for an individual to just monitor his own individual buying and selling over time.

Another problem commonly pointed out by economist is that it is difficult to impossible to include the effects of the changing quality of goods. For example, the price of electronics has surely fallen over the years, but how do you compare PCs today to PCs 10 to 40 years ago? If you include a contemporary PC in each year's basket, then your measure will fail to measure the fall in price of electronics. But if you consider them as different goods (because different quality), then you'd be comparing different baskets of goods over time, so your units of comparison would be wrong.

And the very concept of a general or average "price level" requires the concept of the "classical dichotomy" of real and nominal price, which is rejected by various schools of economics, including Keynesians!
(https://en.wikipedia.org/wiki/Price_level
https://en.wikipedia.org/wiki/Classical_dichotomy)
And if you reject the concept of the "price level", then you certainly don't think indices tell us about it.


All this arbitrariness gives rise to many, conflicting, indices and "deflators" that purport to say something about price levels and inflation. They don't settle questions, but are themselves a matter of debate. Economists know to consider price indices only with a hefty grain of salt, and know that we cannot obtain the true average price--which is not agreed to even be a valid concept.

[oxmixmudd]

Good information here:
https://climate.nasa.gov/vital-signs...l-temperature/

Although there is nothing overtly wrong about asking the question "Why is global mean temperature a meaningful statistic", in fact it is a sign of an inquisitive and probing mind to ask such a question - , it was not asked in the OP-article with anything like an honest desire to understand it. So I wanted to add a little bit more to aid those that might actually want to know how we can know global mean temperature is not just some arbitrary number cooked up by bad math or poor data.

As has been discussed, the average temperature represents an average thermal energy level in the atmosphere itself, which in turn tells us about changes in the atmosphere's thermal retention properties. But there are other indicators of changing average temperature. One of those is the ocean. The seas are a much less dynamic heat sink than the atmosphere, and it takes a lot longer to effect change in it. In effect then, since the ocean, especially at depth, responds much more slowly, it requires a real, long term change in the average atmospheric temperature for it to react at depth. That is, momentary random fluctuations in atmospheric temperature have little long term effect on average global sea temperature. And especially rising averages that are artificially created by poor thermometer placement or improperly compensated for urban head island or 'artificial' compensations for things like TOB or 'homegenaity' calculations can not possibly produce REAL long term changes in ocean temperatures.

That is, the actual average air temperature MUST be getting warmer over the entire planet on average for heat transfer into the oceans from the surface air interface (and then later at depth) to occur.


So what do we see?

Two articles, one on what has been Hawaii's Warmest summer on record - due primarily to a warming Pacific Ocean

https://www.wunderground.com/cat6/Ha...Second-Warmest

and an article on how ocean warming is reaching the abyss:

https://www.earthmagazine.org/articl...-reaches-abyss

We see warming ocean temperatures across the globe AND we are seeing warming at depth as ocean currents take those warmer waters at depth.

from the Hawaii article:

average_sea_temperatures.jpg

As you can see, we are seeing the warmest ocean temperatures since 1854 almost all grouped together in the last 2 decades.

The second article discusses how ocean currents are taking this warm water down into the depths. Understanding that these changes REQUIRE Long term changes to atmospheric temperature because the ocean, again, takes a lot more time to absorb (and unfortunately to dissipate) heat than the atmosphere - especially at depth.

So YES - the mean global atmospheric temperature IS a meaningful statistic. It is in fact telling us what is happening to the heat retention properties of the atmosphere, and in spite of all the 'tobacco science' reasoning that has been and continues to be applied to the real science in disreputable means to discredit it - we can see by other objective, independent measures that the atmosphere is warming over time - which is exactly what the global mean temperature is telling us.


Jim

[Joel]

As has been discussed, the average temperature represents an average thermal energy level in the atmosphere itself

Is that true? I think the heat capacity of gasses varies with things like pressure, temperature and humidity.
(e.g. https://www.engineeringtoolbox.com/a...ity-d_705.html
https://www.engineeringtoolbox.com/a...es-d_1535.html)

And I pointed out earlier in this thread that when you have system containing non-uniform heat capacity, then you can have situations where average temperature and amount of thermal energy can even move in opposite directions.

So, because the atmosphere is not homogeneous, I think we need additional steps to know what is the actual relation between average temperature and the thermal content of the atmosphere.

In effect then, since the ocean, especially at depth, responds much more slowly, it requires a real, long term change in the average atmospheric temperature for it to react at depth...That is, the actual average air temperature MUST be getting warmer over the entire planet on average for heat transfer into the oceans from the surface air interface (and then later at depth) to occur.

I thought (as in the article I posted earlier in this thread) that the explanation for the hiatus in atmospheric warming was that the ocean depths were warming instead of the atmosphere (and ocean surface) warming, not because of the atmospheric warming.

[Terraceth]

from the Hawaii article:

[ATTACH=CONFIG]39679[/ATTACH]

As you can see, we are seeing the warmest ocean temperatures since 1854 almost all grouped together in the last 2 decades.

I know this is the pettiest of complaints, but I still need to say it...

Why in the world is that graph in reverse timeline? The standard is that the left-most bar is the earliest and the right-most bar is the most recent, i.e. it moves from oldest to newest, left to right. It goes against every convention to do it the opposite and put the most recent on the left side, and just makes it look weird.

[Sparko]

I know this is the pettiest of complaints, but I still need to say it...

Why in the world is that graph in reverse timeline? The standard is that the left-most bar is the earliest and the right-most bar is the most recent, i.e. it moves from oldest to newest, left to right. It goes against every convention to do it the opposite and put the most recent on the left side, and just makes it look weird.

I would like to know why it is measuring surface temperature which would be mostly affected by things like sunshine, and air temp. If we are looking at how much of a heat sink the oceans are, I would think we should be measuring the temperature below the surface and averaging that. But maybe I am missing something?

[oxmixmudd]

I would like to know why it is measuring surface temperature which would be mostly affected by things like sunshine, and air temp. If we are looking at how much of a heat sink the oceans are, I would think we should be measuring the temperature below the surface and averaging that. But maybe I am missing something?

Here is a pretty good set of articles with information about sea surface measurement. But surface doesn't necessarily mean the first inch, or even feet, it refers to a depth of up to 10 meters (33 ft). At 10 meters there is little to no variation due to the day/night cycle, with the most variation occurring around the first millimeter - about 2.5 degrees. Solar infrared doesn't penetrate very deep into the ocean, so the actual amount of direct warming from the sun is not all that much, even at the surface.

https://skepticalscience.com/ocean_t...ure_part1.html

https://skepticalscience.com/ocean_temperature_part2.html


And there are programs such as ARGO which are looking at temperatures down to a depth of 2000m, but there is nothing like a record from 1854 to work with.

Jim

[firstfloor]

Post reports podcast, the south Atlantic blob, 11 September. Time to wake up Christians; your God is giving you a good kicking, and He does not like the way you are ignoring Him.

[oxmixmudd]

Post reports podcast, the south Atlantic blob, 11 September. Time to wake up Christians; your God is giving you a good kicking, and He does not like the way you are ignoring Him.

Not sure why you would characterize it like that. Humanity is the cause, not God. But it is an interesting phenomenon. Thanks for pointing it out.

Jim

[firstfloor]

Not sure why you would characterize it like that. Humanity is the cause, not God. But it is an interesting phenomenon. Thanks for pointing it out.

Jim

Are some Christians confused? Is climate change a sign of the end of days, or something else?
I found this evangelical call to action:

[oxmixmudd]

Are some Christians confused? Is climate change a sign of the end of days, or something else?
I found this evangelical call to action:

Chritians are not a monolithic group - that much should be obvious (catholic, baptist, methodist, presbyterian etc. etc.).

But many are very much confused as regards science and faith and how to differentiate between people that are just feigning some sort of commitment to their ideals to gain their support and those that represent some element of truth in one area but that promte other agendas contrary to their ideals.

Jim

[firstfloor]

Chritians are not a monolithic group - that much should be obvious (catholic, baptist, methodist, presbyterian etc. etc.).

But many are very much confused as regards science and faith and how to differentiate between people that are just feigning some sort of commitment to their ideals to gain their support and those that represent some element of truth in one area but that promte other agendas contrary to their ideals.

Jim

[oxmixmudd]

Not worthy of comment FF. But if you change your mind and decide to discuss more seriously, I'll be willing to talk.

Jim