Jump to content
Objectivism Online Forum

Chaotic nature of the climate

Rate this topic


brian0918

Recommended Posts

I've been involved with a meetup group email debate regarding AGW, and got onto the topic of the complexity of the climate. Feel free to use any of the following in your own arguments. If any of my reasoning is bad, please let me know.

The person I was responding to relied on a paper from 1991 rebutting AGW skeptics' claims (Kellogg's "Response to Skeptics of Global Warming").

Coming from a physics background, I am skeptical of the certainty of their models' conclusions due to the complexity of the system, so with regard to Kellogg's "Response to Skeptics of Global Warming", I fit into section 2. However Kellogg doesn't even attempt to refute my position, except to say that all five models that existed in 1991 all predict an increase of 4 to 9 degrees Fahrenheit if we maintain our present course over the next half century.

So my first question to you: almost 20 years later, do they still predict such a large increase in temperature in such a short time-span?

If the answer is no, then the original 5 models' conclusions - of which Kellogg was so certain - were actually wrong, or the data on which they relied was inaccurate. Either way, those scientists were incorrect in their "error analysis" - ie, in the stated amount of uncertainty in their measurements and modeling.

Now, I should explain why I claim that the climate system is extremely complex. In my physics courses and labs we covered the topic of "chaotic systems" - which are systems so complex that slight changes in the accuracy of your initial measurements lead to complete unpredictability of the system within a short period. For a system to be considered "chaotic", there are certain requirements in the number and type of variables that affect the system. In our labs, the systems involved pendulums and water drops from a faucet, with only a few variables affecting it.

Suffice it to say that the Earth's climate is extremely chaotic - it has an immense number of interdependent variables and large uncertainties in measurement. The 1997 paper I link to above, which examines the chaotic nature of the global climate in predicting global temperature changes, concludes: "there are severe limits to climate predictability at all scales [of time]."

The "butterfly effect" is the popular example of this. You can see this yourself by your weatherman's inability to accurately predict the weather more than a few days ahead of time - or sometimes no more than a few hours ahead of time. And they have also even been wrong in that case.

Now expand the local weatherman's area of focus from your one county to the entire globe, and increase the time scale from a few hours or days to decades and centuries. In order to make any claim to certainty, the initial measurements have to be... obscenely accurate - accuracy unheard-of in meteorology, climatology, dendrochronology, etc.

And if the initial measurements are slightly off, the whole model becomes useless. Even a tiny error of 0.0001% in initial measurements can quickly lead to 100% uncertainty, as we have seen with these models' inability to even predict the cooling of the last several years.

Like physicist Richard Feynman in regards to social science, I am likewise skeptical of the sweeping certainty and conclusions of climate scientists, for the exact same reasons. They both deal with chaotic systems, they both have large uncertainties in initial measurements, and they both assume that that doesn't render their models entirely useless.

So, in summary - even if climate scientists' models correctly simulate the Earth's climate - which even they admit is not likely, any slight inaccuracy in their initial measurements can produce absolutely wrong results. And if the models don't even simulate the Earth's climate correctly, then - due to the nature of chaotic systems - the models are also completely useless.

In looking up rebuttals to this, I found people who simply misunderstood what was meant by "chaos" - it doesn't mean "wild and crazy, shooting way up and down". I did find one analogy in New Scientist that initially seemed reasonable:

You cannot predict the exact path a ball will take as it bounces through a pinball machine. But you can predict that the average score will change if the entire machine is tilted.

My response to this, left in the Comments section:

Your pinball scenario is a false analogy, because it assumes that all other variables - ie, the positions, size, and number of the knobs, bumpers, bouncing plates, the points awarded, etc - are all fixed, *independent* variables.

Unfortunately for your argument, "tilting the table" doesn't necessarily have the same effect as in your analogy. To fix your analogy - tilting the pinball table would have to warp all of those variables, because they are not independent, but interdependent. It would be as if tilting the table caused the bumpers to all shift around, change in size, warp in shape, change in point value awarded, etc, all in a way that has not been accurately modeled.

So, your analogy inadvertently points out the simplicity of the current climate models. "If we just assume that all of these variables are independent or fixed, then we'll get this result" - except for the slight problem that none of those variables are independent or fixed, but are dynamic and highly interdependent.

Edited by brian0918
Link to comment
Share on other sites

Here is a good analogy I recently thought of regarding the reasoning of AGW proponents:

GW scientists model the Earth's climate to the best of their ability. They then compare their model to reality. They see a warming in the 90s that they can't explain. So they attribute it to the only thing they can think of that they haven't accounted for - the activity of man. That is the source of the conclusion that GW, if it is real, is caused by man (and thus is AGW).

Does this sound familiar?

When I encounter this argument, I immediately think of the God of the Gaps, in particular the Irreducible Complexity argument of creationists. Namely, we cannot currently understand the level of complexity of organisms - ie our models don't match reality - so they attribute the complexity to the only thing they haven't accounted for - God.

This is how God of the Gaps always works: wherever our understanding is incomplete, the answer is "God did it". For AGW proponents, wherever the model fails to predict reality, the answer is "Man did it". Neither group bothers to consider the validity of their model. The argument is simply an argument from ignorance, a logical fallacy - "I don't know what caused X, so I will attribute X to cause A."

Edited by brian0918
Link to comment
Share on other sites

I see one problem with this argument: there is some evidence that centennial climate variability is not chaotic.

Chaos is not an irreducible primary, as many scientists (unfortunately) believe. One way in which chaos arises is by the interaction between multiple modes of the system. This is known as non-linear dynamics. The timescale of chaos, thus, has a strong dependence on the time-scale of system-modes. (Timescale of lower-order chaos is generally double or quadruple than the time-scale of system modes). What is the longest timescale of earth's ocean-atmospheric modes? The longest "natural" variability known is El Nino - Southern Oscillation (ENSO), which has 3-5 year irregular period. ("Natural" means without any external forcing - solar, volcanic, CO2 etc.)

An important question in this regard is: what is the longest possible timescale for ocean-atmospheric modes? If we look at ENSO, one likely hypothesis is that its timescale is determined by the time it takes for oceanic (Rossby) waves to cross the Pacific. In fact, there is plenty of evidence that wave-adjustment is the mechanism by which large-scale ocean-atmosphere system adjusts (from one state to another). Thus, the timescale of ocean-atmospheric modes may have an upper limit determined by (i) size of the earth, and (ii) propagation speed of slowest waves. A simple calculation implies that this number could not be larger than 15 years.

Thus, the timescale of chaos resulting from the (non-linear) interaction of these slowest modes could not be greater than 30 years. If this calculation is right, climate trends over longer timescales could only be due to external forcing (solar, volcanic, CO2 etc.). Since, earth's climate does not have an internal source of energy, this means that over longer timescales, it is necessarily a damped system sustained by external forcing. Simplest example is the hemispherical winter-time cooling due to decrease in solar forcing. In this case, chaos does not play a role because inter-hemispherical modes cannot interact due to earth's rotation. (I tried to convey this previously and got something in return from Zip :D )

The question for climate science is: is CO2 large-enough forcing for centennial climate variability? I agree with you that the models are basically useless. The reason, however, I think is not chaos. Rather, it is the extremely uncertain cloud and water vapor feedbacks, magnitude of air-sea exchange fluxes and inaccurate convective parameterization schemes.

(Note: Some would argue that Pacific Decadal Oscillation (PDO) is the longest (20-30 year period) and not ENSO. However, I am unsure whether PDO is a real dynamical oscillation or is just a residual left from irregular ENSO, or even worse, a manifestation of changes in observational technology and coverage.)

Edited by Rockefeller
Link to comment
Share on other sites

Chaos is not an irreducible primary, as many scientists (unfortunately) believe.

Of course not, that wouldn't make any sense. The fundamental physics involves particles interacting with eachother, of course, but it is hopeless to simulate all the particles on the Earth, so we attempt to model it using simplifications. Unfortunately, for chaotic systems, simplifications and/or uncertainty in initial conditions and variables results in total non-predictive value.

The reason, however, I think is not chaos.

Let me correct you here: I don't believe "chaos did it" is the irreducible answer either. It is simply that there is no evidence that the current models accurately reflect reality, and the reality of chaos means the models and measurements have to be extremely accurate, or the whole effort is pointless. It shouldn't be impossible to model reality, but it is clearly more difficult than the current models claim.

I agree that the fundamental question that climate scientists have to answer is whether the CO2 injection is large enough forcing on the system to dwarf chaos, as happens with the tilting of the Earth to produce the seasons. Even so, there are other external processes that could also force the system, as you say - the Sun's output, volcanic activity, changes in the Earth's orbit/tilt/rotation, the poorly-understood physics of the inner Earth, etc.

Edited by brian0918
Link to comment
Share on other sites

  • 1 month later...

Chaos is an emergent property of system whose dynamics are non-linear and whose temporal behavior is sensitive to initial conditions. The non-linearity of convective systems was discovered (or re-discovered) by Ed Lorenz in the 1950s.

Please see:

http://en.wikipedia.org/wiki/Edward_Norton_Lorenz

for a brief summary of Lorenz' work.

Lorenz re-ignited interest in chaotic dynamic systems which were studied by Henri Poincare back in 1905. There were no electronic computers back then and Poincare could only do so much with manual calculations and qualitative descriptions of chaotic dynamics. Lorenz' discovery was accidental when he did two computations with electronic computers with slightly different initial conditions. This was in connection with a system of deterministic non-linear differential equations describing atmospheric convection. His luck accident re-iginited chaotic dynamics and the rest is history. By the way, this is the origin of the "butterfly effect", a phrase coined, I believe, by Lorenz.

Bob Kolker

Link to comment
Share on other sites

  • 3 weeks later...
...Even so, there are other external processes that could also force the system, as you say - the Sun's output, volcanic activity, changes in the Earth's orbit/tilt/rotation, the poorly-understood physics of the inner Earth, etc.

Your "etc." gets me to thinking of those cute little picture simulations showing the Milky Way rotating on an axis, and moving along its own path in space. How much warning do we receive that the Milky Way is about to encounter some kind of "cold patch" or "hot spot" and to what extent does our position in the solar system protect us from the effects of the galactic orbit and rotation?

Link to comment
Share on other sites

There is at least some speculation that our orbit around the center of the milky way (about 250 million years or so) runs us through thicker or thinner patches of cosmic dust and that affects the climate.

I don't know what our past history of interactions with other galaxies is; I don't know if anyone does at present.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...