>
Average global temperature starts to drop unexpectedly, and continues to drop for a decade or more (so that we can rule out yearly statistical noise), despite the increased level of CO2.
No, it doesn't have to drop. It just has to not rise. Which is exactly what has been happening for close to two decades now.
Furthermore, the level of CO2 has not just increased--it has increased by close to the "business as usual" scenario in the models--i.e., the worst case scenario, which the models predicted would show significant temperature rise that has not happened. In other words, the models have already been falsified.
> falsifying established science is hard.
Climate science is not "established science" in the way that GR or QM is. Falsifying GR or QM within their established domains of validity would be extremely hard, yes--but that's because their established domains of validity are nailed down by mountains of evidence from controlled experiments that matches the theoretical predictions to ten or more decimal places. Climate science is not even close--not even within orders of magnitude--to that kind of accuracy.
In fact, we should not expect climate science to be anywhere close to GR and QM in accuracy, because it's still a fairly new science, its domain is very complex, and it's not possible to do controlled experiments in most of the domain. We can't re-start the Earth's climate of fifty years ago with slightly different initial conditions and see how it performs.
None of this, in itself, disqualifies climate science as a science, or means that climate scientists are not doing important work. But it does mean that climate science simply cannot make the kinds of claims that GR or QM make about what is going to happen if we do or don't do particular things. The honest thing for climate scientists to do would be to admit that.
This[1] is the first Google result for "has the global warming stopped".
In addition, this is an interactive page that lets you plot yearly global temperatures up to now. Now look at this graph and please tell me if temperature did not rise in the 21st century.
Finally, this page [2] compares IPCC's 2001 prediction with observed temperatures up to 2010. Look at the graph titled "2001 Projections vs. Observations", and see how closely the prediction matches reality starting 2001. (Which is not surprising, considering that it's only 10 years.) And now ponder what would happen if this "trend" continues for another 90 years.
> Now look at this graph and please tell me if temperature did not rise in the 21st century.
These aren't "graphs", they're histograms, and it's impossible to read any trend lines off of them. So they're useless for this discussion.
> Look at the graph titled "2001 Projections vs. Observations", and see how closely the prediction matches reality starting 2001.
First of all, since the IPCC TAR was published in 2001, anything before 2001 is not a "projection", it's a "fit of model to past data". (And notice that the model does not fit all of the past data very well; before the mid-1900s in Figure 2 there is significant divergence. A model that can't retrodict the past can't be expected to predict the future.)
Second, these graphs are useless because they just give the "prediction" as a single line, with no error bars and no differentiation between the different sets of model runs that were based on different assumptions for how CO2 would rise. In other words, these graphs are spin, not science.
Real Climate is much too biased a source, and has been caught too many times in the past being disingenuous, for me to take anything they say seriously. Sorry.
Also note the graph showing the increasing mismatch between models and observations.
Of course many will object to my choice of source, just as I objected to RealClimate as a source. But the data set used is an "official" one (the RSS satellite data), though not one you hear as much about.
No, it doesn't have to drop. It just has to not rise. Which is exactly what has been happening for close to two decades now.
Furthermore, the level of CO2 has not just increased--it has increased by close to the "business as usual" scenario in the models--i.e., the worst case scenario, which the models predicted would show significant temperature rise that has not happened. In other words, the models have already been falsified.
> falsifying established science is hard.
Climate science is not "established science" in the way that GR or QM is. Falsifying GR or QM within their established domains of validity would be extremely hard, yes--but that's because their established domains of validity are nailed down by mountains of evidence from controlled experiments that matches the theoretical predictions to ten or more decimal places. Climate science is not even close--not even within orders of magnitude--to that kind of accuracy.
In fact, we should not expect climate science to be anywhere close to GR and QM in accuracy, because it's still a fairly new science, its domain is very complex, and it's not possible to do controlled experiments in most of the domain. We can't re-start the Earth's climate of fifty years ago with slightly different initial conditions and see how it performs.
None of this, in itself, disqualifies climate science as a science, or means that climate scientists are not doing important work. But it does mean that climate science simply cannot make the kinds of claims that GR or QM make about what is going to happen if we do or don't do particular things. The honest thing for climate scientists to do would be to admit that.