oblong

I am aware of what a model is in general but as you mention I was using the word specifically in reference to GCMs.

I was only talking about the first kind of model which represents a nonlinear climate with various feedbacks - and of course they are not perfect. The output from many runs can give certain relationships such as climate sensitivity. If the relationship between co2 increase and temperature turns out to be logarithmic according to these nonlinear models, then perhaps that actually is the case. I don't see why it must be automatically a fault that a linear relationship has been found within a nonlinear system.

It also assumes certain conditions on earth. The climate sensitivity may come out differently for an earth in total ice cover for example, or an earth with a different atmosphere, or different configuration of continents. The linear relationship between co2 and temperature is an output of the model set to physics similar to today's climate.

I don't see what simplification they have made to the physics that has implicitly linearised the earth's climate. The model represents a nonlinear climate, and it just happens that there are linear relationships that can be found in it.

I am not aware of the second type of model, and it doesn't look any different than a description of the output of the first. If the relationship between co2 and temperature really is logarithmic with about 3C warming per doubling co2 - at least on the current earth - I don't see how that cannot be true just because the climate as a whole is non-linear.

GCMs incorperate knowledge of climate physics and in doing this they can reproduce the 20th century trends well. Going back further they are less able to reproduce climate. Partly that's not suprising because there is less data available for both understanding the physics involved and for validation, and in some cases because conditions were different than todays. In the specific case of interglacial warmings and glacial coolings the causes and feedbacks may have been different than anything observed in today's climate. Either way the GCMs are clearly not perfect, but they are not without validation either, especially for today's climate.

warrenm2

That'll now be illegal in 2 years time - brilliant rent seeking by the low energy bulb manufacturers (thats an economics term BTW)

oblong

The fact that climate models cannot reproduce interglacial and glacial warmings only reflects that current scientific knowledge cannot fully explain the causes behind them.

In constrast climate models do fine at reproducing the 20th century temperature trend reflecting that current scientific knowledge is able to explain the causes behind the 20th century temperature trend.

The don't predict global chaos. They project temperature rise given different conditions:
Images from “The Greenhouse Effect and Climate Change”

Whether say a 3C temperature rise by 2100 will cause global chaos is not something the models say.

Janspeed

iTrader: (1)

As long as there is an Ozone layer, I will be making a hole in it, with the taxes I pay I have every damn right!!

matty01

Get a grip people , the whole climate change thing is so obviously a scheme dreamt up by the American gov. to give them a reason to start to lay the plans to cut oil dependency, ie. dependency on the middle east, what other possible reason could there be for the US gov. to pass new car emission laws [ie oil usage] and give themselves a hard time, other than to save the planet if climate change does turn out to be down to us, by the time we are sure it's down to humans it'll be too late, and people will be asking why nobody did f all about it.....

cynical - Definitions from Dictionary.com

ps If you hav'nt seen 'An Inconvenient Truth' by Al Gore than i suggest you go and rent it now ,blockbusters have it, and then see what you think , it's pretty compelling stuff and hard to deny ,unless you're a fukcwit.....

Sprint Chief

OK, just a simple semantic confusion, happens all the time!



...and this is why the Paschen's law analogy is so important. For the uninitiated, Paschen's law is something that should be very close to every scooby drivers heart! It defines the voltage required to generate a spark across a gap of a given distance, at a given air pressure. Paschen wasn't the first to investigate this: others performed experiments in which they varied just one parameter at a time in controlled experiments (just like Arrhenius varied the concentration of CO2). But they could not determine the true relationship - which was a nonlinear combination of these factors. In fact, the only way to find out the true combination was to build the experiment and make observations while varying both parameters.

Models are built on parameterisations based on simple experiments (and, unfortunately, dissipations and adjustments made solely to make the model "stable") that may not correctly relate various aspects of the climate.


But the model is tuned (through dissipations, flux adjustments and / or parameterisations) to be stable and behave like current climate. Hindcast results are incestuous, implicitly, because the models are tuned diagnostically to match recent measurements. This is no bad thing (it represents the models greatest utility) but a direct consequence is that the models are fundamentally unusable for future prediction.



Just to be clear: many aspects of GCMs are not linearised (although they also do not reflect the dynamics of climate well, even though they mimic some recent climate metrics), but the climate sensitivity model as presented by the IPCC (reducing the model inputs to forcings) has an implicit linearisation.



See Paschen's law explanation above. I know I keep harping on about this but it really is a good lesson regarding the dangers of linearising complex systems.


Current GCMs only have diagnostic merit; they can be used to show some of the things we don't understand. They do not have predictive power for several basic reasons; (1) climate is an initial-value problem; (2) climate exhibits self-similar behaviour and long-term persistence which is not captured by the models (3) many of relationships that drive climate are still poorly understood (e.g. cloud formation), so the models cannot capture these aspects and (4) the future climate is dependent on external variables that cannot be known, e.g. volcanic eruptions, solar output etc.

Furthermore, GCMs produce much unphysical behaviour. I'm not fully up to speed on these things, and problems vary from model to model, but amongst the problems include:
- Failure to simulate interglacial entry/exit (as mentioned above)
- Split ITCZ
- Having to change parameterisations as a function of resolution to keep models stable (strong evidence of unphysical behaviour)
- Dissipations and flux adjustments needed to match modern climate
- Atmosphere having the viscosity of tar to make model stable

The bottom line is these models have their limits and they are being used way beyond these limits, and then subsequently to define a linear model (the climate sensitivity model) which is then used to "prove" that only greenhouse gasses could possibly explain the recent warming. Greenhouse gasses are a plausible explanation for recent warming, but there are other, equally plausible, competing theories. Scientific knowledge is a long way from understanding which is correct, and needs a fundamental breakthrough to make headway. GCMs will probably never answer these questions.

Janspeed

iTrader: (1)

So many damn quotes!!!

LOL **** Global warming!

CrisPDuk

I've seen 'An Inconvenient Truth', admittedly it is compelling stuff. However I found certain parts of it quite easy to dispute, and I like to think I'm not a fcukwit


Here's just one for instance, why do the global warming advocates usually only quote global temperatures from the late 1800's onward, when reliable, accurate thermometers have been in existance since the early 1800's?

Is it by any chance related to the earth's temperature being somewhat warmer prior to the 3 major volcanic eruptions that occurred at the turn of the last century, which dumped tens of millions of tonnes of light blocking dust into the upper atmosphere

When this sucker blows in the next couple of decade or so,

Krakatau, Indonesia

the last thing on anybody's mind will be Global Warming