Having the same trend is “not warming” wouldn’t you say?

You have to include the “cold” years of 1992-1994 to get to the .2 per decade level.

13 years isn’t “consistent” with .2 per decade – it’s just “not inconsistent”! Wouldn’t that be the way to say it?

That we do not know if it is “consistent” or is “not consistent” , we only know it is not yet shown to be inconsistent. For that we have to wait until a long enough period can be averaged to get us that average per decade, correct? We have undefined, not applicable, lacking skill, or inappropiate time periods.

Why not just say that any readings less than 20 years are in an unknown status?

The argument about 8 years is an interesting one! The last time a period of 8 years had a negative trend slope started 20 years ago, from 1988-1995 at -.03 And flat for 9 years of 1988-1996! I would guess a volcano is involved? If no period of 8 years is sloping down or 9 years is flat without a volcano, and none since 1996 then any period of 8 or 9 years with no volcano should be sufficient, right?

But it is less than 20 years you say! But I say it does show an average of .2 those two year groups, 8 and 9! Just 10% off each group. If it is per decade, is not the last 10 years plus 3 minus 2 in the range of .2? As long as we do not expect it to grow 10% a year every year, it seems that around 8 years is enough? It just jumps at some point in time as the years get longer!

If 2008 keeps as it is going at .35 then since Jan 2000 the trend for 9 years will be .03! That has not happened for over 20 years. And since there was no volcano from 2000 until the figures for Apr 2008 came out, how do you explain it? Where is the greenhouse gas effect the last 9 years all will cry out!

But this is still too imprecise for me:

“Over this relatively short period the models’ predictions are unlikely to fail”

What do you mean by “unlikely”?

If 8 years won’t do it, then how long will?

We are still talking about 8-year trends in GAST, right? And about model ensembles, not, say, GCMs initialised with real data of the sort that people are beginning to run? Then I (almost) agree with the statement above. Over this relatively short period the models’ predictions are unlikely to fail, even if the underlying models have a significant bias. That’s exactly what I mean when I say 8-year trends in GAST provide a weak test for the models.

You write “it’s the *possibility* of the model failing the comparison that gives us more confidence when it doesn’t fail”

The “possibility” of a model failing is exactly the question I seek to answer when I have asked, in principle, “What would be “inconsistent with” the predictions of the models?”

Most (not all) answers suggest that nothing would be “inconsistent with” hence, there is no “possibility” for them to fail.

I think models should be verified every way they can be, by comparing their predictions (a very general term, not restricted to forecasts of what will happen in the future) with measurements at every opportunity. So I believe that comparing (say) predictions of GAST from the IPCC AOGCM ensemble in 2001 with measured data since is potentially valuable, but unfortunately (because of the way the predictions are expressed and the variability of the real world) it’s going to take a while until this comparison will give us any additional confidence in the model. Or not: it’s the *possibility* of the model failing the comparison that gives us more confidence when it doesn’t fail. In the meantime, we base our confidence (or not) in the models on the information we already have.

Since we are mostly or all scientists, I propose an experiment. Let’s sample people who think there may be bad effects but admit we don’t know the magnitude, and people who think that models are a pretty accurate basis for decisions and ask them:

1) What policies do you favor about reducing carbon and adapting to climate induced change?
2) And if you had 100 million to do research to help with climate change what percentages would you fund for what kinds of projects?

In this way, we could empirically test the hypothesis “model outputs matter in terms of policy and desirable research portfolios.” And we could test the sensitivity of different policy options and research portfolios to climate modeling.

You’re playing ‘word games’.

AGW is a very simple model – but a model none-the-less.

One looks at the Keeling curve. Interprets (and ‘confirms’ by computation) that the almost monotonic increase of CO2 (except for very regular seasonal fluctuations) is due mainly to human burning of fossil fuels. One also looks at the global temperature trend. One then takes into account ‘the greenhouse phenomenon’, and makes the conjecture (model, hypothesis, theory) that a good part of the observed warming trend is a function of (cause of) the anthropogenic atmospheric trend of increasing CO2.

Whether qualitative or quantitative – it’s a MODEL!

Len Ornstein

I am for improving human adaptability to weather extremes. This includes everything from constructing better buildings and smarter developments to the global adoption of modern farming methods, including crops genetically modified to produce the most efficient yields for a given region.

The only thing I am really opposed to is a global central authority artificially raising the cost of fuel through regulations and restrictions, based on an artificially generated crisis scheduled 100 year from now. History shows use that such central authorities are a very bad idea, prone to corruption and dismal inefficiency, always resulting in a huge increase in human suffering. If this authority proved to be an exception, there is still no discernable benefit from the regulations and restrictions. There would be no way to verify if the sacrifices were ‘worth it’!

There are many things we can do that make sense, but the lemmings are all charging for the cliff of global carbon mitigation, and the things that might really produce significant benefits are being largely ignored.

Nope.

“AGW” does not make predictions, models made by people with various assumptions etc. do.

As we have learned an enormous range of states are “consistent with” the models.

The Roger Sr.-motivated “update” makes the point that OBSERVED, recent, Ocean-’surface’ cooling is probably ‘responsible for’ the observed ‘flattening’ of the surface-air, temperature record, and that current GCM modeling of fluxes of heat, from ocean surface to ocean depths, is rather poor.

Current records of temperatures of the deep ocean are much less complete than of its surface. So we can’t yet tell whether the net heat content of the oceans plus atmosphere has none-the-less increased, as ‘predicted’ by AGW.

Does this seriously increase your personal skepticism about AGW?

Len Ornstein