Saturday, July 29, 2006

We still have a problem

Last year I posted an essay on Global warming. Since then I've been finding out more about the subject. Nothing that I have seen gives me any reason to change my belief that the problem is real.

In my earlier post I covered the evidence that human activity has led to increases in CO2 and the evidence that Global temperatures are increasing. I did not cover in detail the reasons for believing that the current temperature increases are caused by increases in CO2.

There are three main lines of argument.

The first is the claim that the pattern in space and time of the temperature changes is the one which would be expected from greenhouse gas forcing rather than from solar forcing.

The second is the claim that models taking into account human activity are consistent with the observations. Models which do not take into account human generated greenhouse gases and aerosols are not consistent with the observations.

The third is the claim that the current warming is unprecedented in the past few hundred years and probably in the past thousand years. That is the current changes are too great to be entirely due to the natural variability of the climate system.

I do not regard these arguments as equally strong. The first argument is quite strong, strong enough to carry the case by itself. The third is to my mind only supporting evidence and not conclusive by itself. The second is in between. I regard it as fairly strong but can understand how others might find it less convincing.

Solar variability is often proposed as an alternative explanation for Global warming. Solar forcing would act through an increased energy input and greenhouse forcing would act by reducing energy re-radiation. One would expect the patterns of temperature changes to be different for solar and greenhouse forcing.

The observed pattern for current warming matches that for greenhouse gas forcing not the one for solar forcing. Temperatures have risen more at night than during the day. They have risen more in Winter than in Summer. They have risen more in Polar regions than in the Tropics. The Stratosphere has cooled while the Troposphere, the surface and the oceans have warmed. These are all what would be expected from greenhouse gas forcing, not from solar forcing. I would say that the fingerprint of temperature changes is sufficient reason to believe that greenhouse gas forcing is the cause of Global warming.

It has taken a long time to create models of the climate that reproduce the temperature changes over the past century reasonably well. Models that successfully do this are all sensitive to greenhouse gas forcing and human generated aerosols. They also have to take into account solar variation an volcanic activity. These models attribute about half of the temperature increase in the first half of the Twentieth Century to solar variation and the rest to greenhouse gases. They attribute the temperature increase in the second half of the Twentieth Century to greenhouse gases. There isn't any current increase in solar output.

The General Circulation Models used to model climate are very complicated and it would require more time and resources than I have available to check them properly. However they are capable of making predictions. They correctly predicted temperature trends in the Troposphere. Measurements that disagreed with the predicted trends were found to be biased. When the biases were corrected the data was consistent with the predicted trends.

Models like these are an area where one just has to trust the expertise and integrity of the people doing the modeling at least as far as the details of implementation go. It takes a lot of time and effort to get a feel for the details. The basic concepts behind the models are more accessible. The models that can be created are constrained by the laws of physics. The models are available for criticism. People who think they are basically flawed can create their own. Yes it requires a lot of time and effort and resources. The thing is the General Circulation Models are the creation of scientists who have put the effort into trying to understand climate in detail. I trust them before I trust people doing hand waving and skipping over the details – which means most greenhouse skeptics. (Sorry but too many greenhouse skeptics are just looking for reasons to disbelieve Global warming rather than seeking the truth. This irritates me.) I do understand the dangers of a small group thinking the same way but I think the self-correcting aspect of science is working in this case.

There is a danger of the modeling assumptions being chosen to make model work out even when it is invalid and hence useless as a predictor. However the models used appear to be relatively insensitive to physically plausible starting assumptions. They can be checked out to some extent against palaeoclimatology data. In particular the Last Glacial Maximum has been used to put limits on the sensitivity of climate to CO2 changes. This produces results that are in rough agreement with the models. These are that a doubling of CO2 should bring an increase of average temperature of 2.9 ° C with a 95% confidence interval of 1.5 to 4.5 ° C. I need to look into how they get their confidence interval before I can comment on it. However the palaeoclimate data does seem to rule out very low (<1°>6° C) sensitivities.

Several different kinds of data were combined to give reconstructions of temperatures over the past several centuries. There was a reconstruction of the past 400 years showing a rapid increase over the past 100 years after about 300 years of relatively little variation in temperature. There was also a reconstruction over about 1000 years again showing the past 100 years as being a quite different pattern from the rest. This is the infamous hockey stick graph so called because the rapid 20th Century temperature rise looks like the blade of a hockey stick.

There have been criticisms of some of the statistical methods used to create this graph. It has been claimed that a particular statistical technique Principal Components Analysis was misapplied. It was further claimed that this misapplication could lead to the creation of a hockey stick form even when there was only random variation. Defenders of the reconstruction claimed that even if the PCA was performed in the way that critics suggested the results were only marginally different and that the critics had made an error of their own in extracting too few principal components. I've looked at the methods used and the PCA was definitely performed incorrectly and there was a risk of creating a spurious pattern. However they extracted enough principal components to minimize the effects of their mistake. The critics probably did extract too few principal components. Thus I can see how the mistake in the analysis might not have greatly affected the conclusions. Later analyses using different methods also show anomalous 20th Century warming. Here I think that the pattern extracted is probably real but some of the methods are less rigorous than they should have been.

The real problem has been that none of the hockey stick's originators said that they made a mistake. They at least publicly say that it was a valid choice. It wasn't. This is almost certainly ego speaking. You had applied mathematicians, physicists and other scientists doing their own statistics and making mistakes. They misapplied some methods and were unaware of newer statistical techniques. And they still think they can get away with doing palaeoclimate analysis without one of us helping. (One of us meaning statisticians.) Skill in creating deterministic models does not necessarily help when randomness comes in as it does with palaeoclimate data. But all this is beside the point. The hockey stick was never proof that that that human activity had caused something unprecedented. It was only evidence that something different from the previous pattern had happened in the 20th Century. That something could have been previously undisplayed natural variation. And without a better time series analysis they couldn't even be certain that the blade of the hockey stick did represent something new.

The greenhouse skeptics don't realize that the onus of proof is on them. Increases in CO2 should lead to Global warming unless some feedback mechanism stops or reduces it. We know this from Physics and the properties of CO2 . The main feedbacks should be changes in water vapor and in cloud cover. Increasing temperature should lead to increases in water vapor which would amplify the greenhouse effect. The biggest uncertainties in current models concern changes in cloud cover. This could be either a positive or a negative feedback. The skeptics need to come up with and demonstrate the existence of a negative feedback nearly as strong as the greenhouse gas forcing mechanisms. This has not happened.

I think the evidence for anthropogenic Global warming is strong enough to prove it beyond reasonable doubt. Some of the evidence has not been publicized enough. In particular the CO2 concentrations and carbon isotope ratios from ice cores are not well enough known. These conclusively prove that human activity has led to large increases in CO2 concentration. The temperature trend data is well known. The reasons for attributing Global warming to greenhouse gas increases are not well enough known in particular the fingerprint of the greenhouse forcing is not well enough known. This is easy to understand and very strong evidence.

Some greenhouse skeptics simply don't want to believe in anthropogenic Global warming either because they don't like the changes they would have to make or because they are hostile to environmentalists. I can understand these viewpoints but this sort of skepticism is simply wishful thinking. They end up arguing dishonestly, clutching at straws and cherry-picking evidence. And some skeptics have been misled by myths spread by the first group. For example the claim of Global warming on Mars. (There isn't. It's only a local effect and not driven by solar forcing.) Or the claim that a major volcanic eruption produces more CO2 than years of human activity. (It doesn't. Total human CO2 output dwarfs that of volcanoes.)

And people who are concerned about greenhouse warming should not come across as feeling morally superior. If you do people are likely not to listen.

We are in a mess and have to do something about it. And the people trying to resist this conclusion should realize that they are letting their opponents set the agenda. If this happens the measures taken will be worse than they need to have been.

2 comments:

Brett_McS said...

Tell it to the Chinese!

Lloyd Flack said...

What has that got to do with the price of fish?

Either increases in greenhouse gases increase Earth's temperature significantly or they don't. The Universe does not seem to care in the slightest about politics.