How Can Climate Decades into the Future be Predicted When Weather Forecasts Beyond 3 Days Are Unreliable?

It's an issue of short-term noise vs. long-term signal again.

In the short-term, the weather is extremely chaotic and very difficult to predict. But over the long-term, the short-term variations tend to average out. A good example is ENSO - we can barely predict what it's going to do over the next few months, but over the long-term the El Niños and La Niñas cancel eachother out, and they have no effect on the global temperature trend or climate changes.

The analogy I like to use is with gambling in a casino. There's no way to predict if you'll win a given bet, or come out ahead on a given day or a given week. But if you play long enough, it's a pretty safe prediction that you'll end up losing money. Or with the stock market - it can do anything on a given day or given week. But over the long-term, it's a safe bet that it will go up.

The long-term predictions are easier to make because the short-term noise becomes less prevalent the longer you look into the future.

What deniers like eric don't understand - because they don't want to understand it, because they need AGW to be wrong - is that climate models don't make projections over timespans of less than a decade. Over those short periods, short-term effects like ENSO, which are impossible to predict, dominate. The reason the planet hasn't warmed much over the last 8 years is that there have been mostly La Niña cycles over that period. No, climate models didn't predict that. Nor did they attempt to.

My first comment is that I don't really believe you are asking a question, but making a statement. YA is intended for actual questions.

Otherwise I think you're making a good point, that most people seem not to understand. I think they also don't understand how much weather forecasts have improved in the last 20 years.

EDIT: One answerer said:

"Because according to the AGW crowd, you don't need to know all of the boundary conditions. You just need to know that CO2 is driving the warming and thus your models are perfect. Or at least that is what I seem to see the consensus as doing.

"If you disagree with the AGW crowd then you are an outcast."

I don't understand how people that don't believe in AGW expect to convince other people of their beliefs by lying. No one has EVER claimed that the models are perfect, or that boundary conditions are not needed. Perhaps it is only the liars that become outcasts.

There is some validity to that assertion. However, it does downplay the necessary role in climate models of internal processes in controlling the exchange from in-system to out-system. that is, the balance of system input and output (heat gain and loss) is affected by internal processes, such as internal heat transfer (ocean and water circulation) and atmosphere and ocean composition, among other factors.

How these factors affect the input-output must be presumed. They, like the weather, are difficult to predict and so the models tend to fudge factor them in (a generic relationship is predicated based on recent climate behavior and this generic relationship is used for future extrapolations).

There is still an awful lot of slop in the calculations.

One thing that seems to get lost in all of the climate model discussion, or rather that gets downplayed, is that modellers tend to incorporate the same sort of fudge factors (they all read the same studies and use those studies in developing their model conditions), so of course they tend to arrive at the same results. That is, when many different models arrive at similar results, it is often presented as proof of model validity, when in fact all it is proving (in my view) is that the models are all more or less based on the same presumptions.

They can't and people really have to understand what the word "prediction" actually means. Now when the prediction comes from "experts" people loose the fact it's a prediction, more likely to be true because they are "experts", no, it's not ,more likely it's a "valued" prediction. The models use variables that are "predicted" and what is ended up with is a prediction based on variables.

And by definition variables are subject to change, therefore so is the "prediction.

OttawaMike said,

My comment is a picture: http://c3headlines.typepad.com/.a/6a0105%E2%80%A6

Now, lest anyone get the idea (from the above link) that Dr. Orrin Pilkey doubts that global-warming is a problem, let's take a look at what his views are regarding global-warming and sea-level rise (from http://www.islandpacket.com/news/local/story/59465...

############

However, Orrin Pilkey, a professor at Duke University, contends global warming is driven by escalating carbon dioxide emissions, which cause the ocean to heat up and glaciers to melt.

As the ocean warms, it expands. As a result, sea levels are expected to rise 2.5 feet by 2050, and up to 5 feet by the end of the century, said Pilkey, an earth and ocean sciences teacher at Duke's Nicholas School of the Environment.

"Sea level rise is much more serious than flooding," Pilkey said. "(With those rises) you have to look at storm surges, waves that would begin to penetrate into the community and other problems such as the impact on drainage."

############

And people should note that climate-models are *not* the most important sources of information that show how much CO2 emissions will impact the climate. The best source of that information is the direct physical evidence uncovered by paleoclimatogists. Advances in isotopic analysis have allowed paleoclimatologists to link CO2 changes with major climate upheavals in the Earth's past.

Of particular importance in this respect is an event known as the "Paleocene Eocene Thermal Maximum". Analysis of the PETM event has given us a pretty good idea of what could happen to the Earth's climate over the next couple of centuries if fossil-fuel (in particular, coal) emissions are not reigned in. If we were to provoke another PETM-like event, it would be a civilization-ender.

People who try to cast doubt on global-warming by nitpicking at climate-models (which actually don't do that badly on a global scale) are deliberately ignoring the most powerful evidence -- the direct physical evidence left by major climate events in the Earth's past.

In fact, there is strong evidence linking most of the great extinction-events in the past (Permian-Triassic, Triassic-Jurassic) to rapid rises in CO2 levels (most likely caused by massive "basalt flood" eruptions). Now, we can do nothing to prevent future "basalt-flood" volcanic events that could cause climate-changing CO2 increases, but that doesn't mean that we should provoke a CO2-induced climate-change disaster of our own.

Edited to add: I'd like to put in a plug for the book "Under a Green Sky". It was written by Dr. Peter Ward, a paleontologist who is one of the world's leading authorities in the mass-extinction events of the Earth's past. In it is a fascinating description of how paleontologists distinguished CO2-driven extinction events from the asteroid-impact Cretaceous extinction event. Think of the book as a combination adventure/science/"Sherlock Holmes whodunnit" book. You can get in paperback (cheap) at amazon.com. Highly recommended.

I agree with Ottawa Mike. The Earth's climate is far too complicated and our understanding of it to immature to make any long term predictions.

Ten years ago, none of the models predicted the slow rise in temperatures that has occured. Now they are going around saying why they were wrong, but they still insist that temperatures will resume their steady upward trend.

http://www.ncdc.noaa.gov/oa/climate/research/2008/...

People on this board who are saying that the models have been accurate up until now, are not pushing science, they are pushing an agenda.

Because according to the AGW crowd, you don't need to know all of the boundary conditions. You just need to know that CO2 is driving the warming and thus your models are perfect. Or at least that is what I seem to see the consensus as doing.

If you disagree with the AGW crowd then you are an outcast.

Because it is a lie. the best forecasting is about 2 weeks.

all weather forecasts are pretty unreliable

My comment is a picture: http://c3headlines.typepad.com/.a/6a010536b5803597...