Greenhouse Effect Experimental Designs

[BrianG]

.... (i.e. doubling the concentration of CO₂ in the atmosphere will cause an average surface temperature increase of 3.29 degrees.)...

 

Plus or minus 60%, right?

[modest]

Plus or minus 60%, right?

 

The paper I was summarizing gave no error margin. As I said, the paper is quite old and there are many factors not considered in the model.

 

The subject of the error margin of the climate sensitivity parameter is extremely robust. The concise answer would be found in the 4th IPCC report section 2.3 and is that doubling CO₂ in earth's atmosphere is likely to increase the global average surface temperature in the range of 2 to 4.5 ºC with the best estimate being 3 ºC and is very unlikely to be less than 1.5 °C. The term "likely" means > 66% probability and "very unlikely" means < 10% probability.

 

You can find the 4th IPCC report here,

Climate Change 2007 Synthesis Report

and an explanation of the terms "likely" and "very unlikely" here,

Guidance Notes for Lead Authors of the IPCC Fourth Assessment Report on Addressing Uncertainties

 

For a more in-depth answer here are some papers addressing the topic,

 

• "Quantifying uncertainties in climate system properties with the use of recent observations" Forest, C.E.; Stone, P.H.; Sokolov, A.P.; Allen, M.R.; Webster, M.D.; Science 295 (5552): 113–7
  
• 
  
• "Climate sensitivity constrained by CO2 concentrations over the past 420 million years" Royer, Dana L.; Berner, Robert A.; Park, Jeffrey; Nature 446 (7135): 530–2
  
• "The Climate Sensitivity and Its Components Diagnosed from Earth Radiation Budget Data" Forster, Piers M. de F.; Gregory, Jonathan M.; Journal of Climate 19 (1): 39–52
  
• "Using multiple observationally-based constraints to estimate climate sensitivity" Annan, J.D.; Hargreaves, J. C.; Geophysical Research Letters 33
  
• "Constraining climate forecasts: the role of prior assumptions" Frame, D.J.; Booth, B.B.B.; Kettleborough, J.A.; Stainforth, D.A.; Gregory, J.M.; Collins, M.; Allen, M.R; Geophysical Research Letters 32
  

 

~modest

[BrianG]

So, the effect of doubling [ce]CO_2[/ce] is less certain, "doubling [ce]CO_2[/ce] in earth's atmosphere is likely to increase the global average surface temperature in the range of 2 to 4.5 ºC" so the average temperature increase is 3.25 C plus or minus 1.25 C, or 38% on top of a starting certainty of 66% or 1.07 C. This is a far cry from the certainty of the charge on an electron, 0.3% Electron - Wikipedia, the gravitational constant, ±0.027 × 10−11 Gravitational constant - Wikipedia .

[modest]

So, the effect of doubling [ce]CO_2[/ce] is less certain, "doubling [ce]CO_2[/ce] in earth's atmosphere is likely to increase the global average surface temperature in the range of 2 to 4.5 ºC" so the average temperature increase is 3.25 C plus or minus 1.25 C, or 38% on top of a starting certainty of 66% or 1.07 C. This is a far cry from the certainty of the charge on an electron, 0.3% Electron - Wikipedia, the gravitational constant, ±0.027 × 10−11 Gravitational constant - Wikipedia .

 

No doubt, the uncertainty in the climate sensitivity parameter is greater than the uncertainty in certain other things.

 

~modest

[BrianG]

No doubt, the uncertainty in the climate sensitivity parameter is greater than the uncertainty in certain other things.

 

~modest

 

On most other things that are defined by experimental science, do you think? Can you name some areas with more uncertainty than climate change mitigations's sensitivity? I've named two with much, much less.

[modest]

I've named two with much, much less.

 

You can't really say that with certainty. You're comparing uncertainty in change in degrees Celsius to, for example, uncertainty in Coulombs which aren't really directly comparable. They have different dimensionality. For a good comparison you would want to choose two things which have uncertainty in temp.

 

Also, just as a side note, the ±0.3% uncertainty that you mention comes from Robert Millikan and Harvey Fletcher's 1909 oil drop experiment which they reported as 1.5924 (± 0.0017) x 10^-19 C. The current value is 1.602176487 (± 0.000000040) x 10^-19 C, which is much, much less uncertain. ±0.0000025% to be exact.

 

On most other things that are defined by experimental science, do you think?

 

That's difficult to answer. Earth is certainly the best-understood climate system of any planet we know. If you changed the composition of Jupiter's atmosphere, for example, what would be the change in temp?... far more uncertain. Other things... the composition of a comet or even the composition of Earth's core—there's more uncertainty in a lot of things in the empirical sciences than there is in Earth's climate sensitivity parameter... are most things more uncertain? I don't know—that would be rather difficult to quantify. It doesn't seem like a worthwhile scientific characterization.

 

Can you name some areas with more uncertainty than climate change mitigations's sensitivity?

 

Areas? Climate sensitivity is a prediction of a future empirical measurement given certain factors that affect the system involved. If, for example, the sun increases in solar irradiance significantly then we will find the 3 ºC prediction very low. But, there's just no way for us to currently know with certitude that the sun will not do that. Based on past behavior of the sun we can say with some confidence what it's likely to do, but there is uncertainty.

 

This same idea would be universal to any complex system. What is the population of Earth going to be if our food production doubles? There may-well be a correlation between food production and population, but that kind of prediction will have uncertainty. You understand?

 

~modest

[BrianG]

...Climate sensitivity is a prediction of a future empirical measurement given certain factors that affect the system involved...

 

Isn't climate change happening now? Aren't they monitoring climate sensitivity as [ce]CO_2[/ce] levels change? The uncertainty is much, much greater than the certainty of the boiling point of water.

 

Are you saying climate sensitivity is purely theoretical, without measured, observed historical verification?

[modest]

Isn't climate change happening now? Aren't they monitoring climate sensitivity as [ce]CO_2[/ce] levels change? The uncertainty is much, much greater than the certainty of the boiling point of water.

 

Are you saying climate sensitivity is purely theoretical, without measured, observed historical verification?

 

I'm saying that the boiling point of water is more uncertain 420 kilometers below the surface of the clouds on Jupiter than is the climate sensitivity parameter. The boiling point of water is not constant, and you continually display your unfamiliarity with the topics you wish to refute.

 

There are paramaters which are less uncertain that the climate sensitivity parameter, and I could easily give you examples, but any complex system will have uncertainty. Given a measurement,

Y = f(X

₁

, X

₂

, . . . , X

_n

)

Where Y depends on X₁, X₂, . . . , X_n the uncertainty is given by,

y = f(x

₁

, x

₂

, . . . , x

_n

)

Any complex system will have may x's... many factors. This is not unique to climatology. If you're looking for a political tag line, I can't help you. The uncertainty in the climate sensitivity parameter is as I have given,

The subject of the error margin of the climate sensitivity parameter is extremely robust. The concise answer would be found in the 4th IPCC report section 2.3 and is that doubling CO

₂

in earth's atmosphere is

likely

to increase the global average surface temperature in the range of 2 to 4.5 ºC with the best estimate being 3 ºC and is

very unlikely

to be less than 1.5 °C. The term "likely" means > 66% probability and "very unlikely" means < 10% probability.

If you need help characterizing that politically then I can't help.

 

~modest

[BrianG]

I'm saying that the boiling point of water is more uncertain 420 kilometers below the surface of the clouds on Jupiter than is the climate sensitivity parameter. The boiling point of water is not constant, and you continually display your unfamiliarity with the topics you wish to refute.

 

There are paramaters which are less uncertain that the climate sensitivity parameter, and I could easily give you examples, but any complex system will have uncertainty. Given a measurement,

Y = f(X

₁

, X

₂

, . . . , X

_n

)

Where Y depends on X₁, X₂, . . . , X_n the uncertainty is given by,

y = f(x

₁

, x

₂

, . . . , x

_n

)

Any complex system will have may x's... many factors. This is not unique to climatology. If you're looking for a political tag line, I can't help you. The uncertainty in the climate sensitivity parameter is as I have given,

If you need help characterizing that politically then I can't help.

 

~modest

 

 

I concede the science of extraterrestrial climate (exoclimatology?) is a field with more uncertainty than climate change mitigation and climate sensitivity to [ce]CO_2[/ce]. Thank you for your kind offer, but I'd like to keep this thread focussed on science and the nature of experimental tests to verify or disprove theory. I see I hold the minority view, and you are doing a yeoman's job for scientific debate. Thank you for your continued effort.

 

Your equation is very interesting:

y = f(x

₁

, x

₂

, . . . , x

_n

)

 

It looks reminds me of the Drake equation. "T.J. Nelson states:

The Drake equation consists of a large number of probabilities multiplied together. Since each factor is guaranteed to be somewhere between 0 and 1, the result is also guaranteed to be a reasonable-looking number between 0 and 1. Unfortunately, all the probabilities are completely unknown, making the result worse than useless."

[modest]

I see I hold the minority view, and you are doing a yeoman's job for scientific debate. Thank you for your continued effort.

 

Sure. I hope not for naught.

 

Your equation is very interesting:

y = f(x

₁

, x

₂

, . . . , x

_n

)

 

It's explained here,

Essentials of expressing measurement uncertainty

the example continues with the link at the bottom of the page.

 

It looks reminds me of the Drake equation.

 

I can see where you get that, but these two things have a different meaning,

[math]y = f(x_1, x_2,... x_n)[/math]

[math]f(x) = x_1 \times x_2 \times ...x_n[/math]

The first is meant to represent any equation which depends on x₁, x₂... x_n, not that they are products. In fact, they usually are summed. That is to say, in the first equation above the uncertainty is usually (but not always) found by summing the uncertainty of each variable. It's explained at the link. Or, I guess that part is not so much explained at that link. It is here,

Wikipedia -- Law of Propagation of Uncertainty

 

~modest

[BrianG]

T.J. Nelson's criticism still applies, see http://hypography.com/forums/environmental-studies/21736-greenhouse-effect-experimental-designs.html#post289660 for citation.

 

Aren't experimental tests required to reduce the "complete unknowns"? How do you propose to reduce uncertainty?

[freeztar]

Aren't experimental tests required to reduce the "complete unknowns"?

 

It depends on how you define "experimental tests". If we get a better telescope into space and make better measurements of galaxies far away, thereby reducing the uncertainty of certain factors such as age, composition, etc., then do we not reduce uncertainty without an actual textbook example of an experiment, per se? How is climate science different in this regard?

 

How do you propose to reduce uncertainty?

 

By continuing to study and make observations, refining our models along the way.

[BrianG]

It depends on how you define "experimental tests". If we get a better telescope into space and make better measurements of galaxies far away, thereby reducing the uncertainty of certain factors such as age, composition, etc., then do we not reduce uncertainty without an actual textbook example of an experiment, per se? How is climate science different in this regard?..

 

Then you've experimented on telescope design(optics), not astronomy. If you find a better thermometer, same thing. It would seem that the theory, increased [ce]CO_2[/ce] is causing catastrophic climate warming should be testable, how do we do it?

[modest]

T.J. Nelson's criticism still applies, see http://hypography.com/forums/environmental-studies/21736-greenhouse-effect-experimental-designs.html#post289660 for citation.

 

His criticism of the Drake equation?

 

Yes, he's spot on about the Drake equation. :naughty:

 

There are a couple terms in the Drake equation which are constrained fairly well by observation such as the rate of star formation, or even, to a much lesser extent, the proportion that have planets. But, most of the terms are a complete unknown. As such, the equation does not give a confident prediction... at all. Most astronomers, I'm quite sure, would tell you that its purpose is not to make a confident prediction, but to explore possibilities.

 

Any case, here's a thread on the topic: The Drake Equation Revisited. Criticizing the Drake equation is off topic here.

 

Aren't experimental tests required to reduce the "complete unknowns"?

 

I'm not sure if you are still talking about the Drake equation or if we're back on the climate sensitivity parameter, but the answer is the same either way. Yes, experimental tests are most useful in science. For example, testing the spectroscopy of carbon dioxide and knowing its absorbance at different wavelengths at different pressures is useful to both climatology and astronomy.

 

Also, generally speaking, an observation is just as effective at testing an hypothesis or a theory as is an experiment. Eddington's 1919 solar eclipse observation, for example, confirmed a prediction of Einstein's general relativity. In the terminology of Popper, the prediction makes the theory testable and the observation makes the theory tested. The experiments in the opening post are examples of the greenhouse effect being tested. Observations also test the theory... for example... the emission temp of Venus is much lower than the actual temp of Venus which is a confirmed prediction of the greenhouse effect.

 

In the philosophy of science this distinction you would have between experiment and observation is... well, it's not as you would have it.

 

~modest

[Erasmus00]

So, the effect of doubling [ce]CO_2[/ce] is less certain, "doubling [ce]CO_2[/ce] in earth's atmosphere is likely to increase the global average surface temperature in the range of 2 to 4.5 ºC" so the average temperature increase is 3.25 C plus or minus 1.25 C, or 38% on top of a starting certainty of 66% or 1.07 C. This is a far cry from the certainty of the charge on an electron, 0.3% Electron - Wikipedia, the gravitational constant, ±0.027 × 10−11 Gravitational constant - Wikipedia .

 

You are comparing a prediction to a direct measurement. The gravitational constant or the charge of the electron have an uncertainty that only depends on the instruments used to measure.

 

Now imagine a calculation that uses both the gravitational constant AND the charge of electron (maybe an electron radiating as it orbits). This calculation will depend on the uncertainty of both, and will be larger. Predictions of climate science depend on lots of input, and each factor's uncertainty then adds to the total uncertainty.

[BrianG]

You are comparing a prediction to a direct measurement. The gravitational constant or the charge of the electron have an uncertainty that only depends on the instruments used to measure...

 

Are you saying climate sensitivity to [ce]CO_2[/ce] isn't a direct measurement, it's a prediction? We have no direct experience of [ce]CO_2[/ce]'s greenhouse effect on atmospheric temperature?

[lemit]

Are you saying climate sensitivity to [ce]CO_2[/ce] isn't a direct measurement, it's a prediction? We have no direct experience of [ce]CO_2[/ce]'s greenhouse effect on atmospheric temperature?

 

Check. Congratulations.