Posted on 18 May 2011 by julienx2k2

The translations of two letters written by three Norwegian scientists (Erik Bye, Ph.D, Professor Ole Humlum and Kjell Stordahl, Ph.D) and published in Teknisk Ukeblad (TU) numbers 1011 and 1611 are available below. This post is a support material to our answer, available here.

Likelihood and climate models (TU 1011)

One of the central questions in the climate debate is how large the probability is for climate change being anthropogenic. The IPCC (Intergovernmental Panel on Climate Change) is operating with a large likelihood that CO2 is the main reason for global warming. And this is, again according to IPCC, CO2 that comes from the burning of fossil fuels, i.e. coal, gas and oil. In 2007, Professor Eystein Jansen, a lead author of the science part of the IPCC assessment report of the same year, stated during a discussion with Professor Ole Humlum in the NRK [Norwegian national TV and radio channels] show "Worth to know" that according to the IPCC there is a 90% probability that climate changes that have taken place during the last 50 years are mainly due to human activities. In the same show, professor Jansen explained further that it is not possible to compute such a certainty. Those 90% show only the share of scientists working with the IPCC supporting the idea that CO2 has a large effect on climate. In other words, in reality it has never been a question of certainty in a statistical meaning.

Forgetting this huge and complete misunderstanding, even a computation of the certainty of the reasons for present climate change is completely worthless. Especially when it is not demonstrated in the first place that what one's attempting to assess (here, the hypothesis that CO2 is responsible for global warming) is correct. The CO2-hypothesis as it is formulated right now does not even fulfills the requirements of science, so we are miles away from being able to calculate a meaningful certainty on it. The reason for its lack of scientific foundation is the lack of ability for the CO2-hypothesis to come with forecasts that can be checked now or within a couple of years' time. Usually it is said that we have to wait 50 to 100 years to check if the hypothesis is correct, which is completely unrealistic in a scientific context. Science must obviously follow its own established rules. If the rules are broken, the result becomes meaningless and unscientific. Therefore, we are in the awkward situation where a lot of planning today is done based on an idea that cannot be checked scientifically.

Hopefully the IPCC is not considered as a scientific organization but as a process. But there are many scientists participating in this process. Among others, there are many that are calling themselves climatologists. So it is a huge paradox for us that the process can still operate with such a false scientific certainty. The models that the IPCC is relying on are actually not based on an established and recognized theory. The whole thing is an hypothesis that is not verified, and it becomes completely wrong, yes, unscientific, to come up with a certainty estimation. This is a theoretical certainty, and there is no method to handle it. It would have been something else if "the uncertainty" had been an expression for the uncertainty in the climate models, disregarding the uncertainty in the measurements, but it is not the case.

And the big paradox we are facing lies first and foremost in that so many scientists accept this sort of theoretical unscientific certainty. Now its is known that a couple of scientists have withdrawn from the IPCC. But there are still hundreds left. What is it that prevents the scientists from protesting, or from reformulating the hypothesis they often base their own work on?

Still dominated by natural climate changes (TU1611)

We appreciate that Eystein Jansen (EJ) commented in TU1311 our correspondence published in TU1011, even if he means this is "strange". EJ criticizes that we are not concrete enough in our critique of the CO2 hypothesis. This was in the meantime not our goal. Our critique was more oriented towards a general scientific aspect. In addition we wished to point out that the IPCC's attribution of the current climate changes to humans with 90% certainty misses a statistical foundation, and looks more like a common vote.

We are followers of the classical scientific tradition, in which ideas for explanations are grounded in observations of the nature. On this basis a hypothesis is formulated, that is subsequently used to make predictions about other conditions in the nature. These can then be observed in the present or in a short time span. This way, the hypothesis is tested against the facts so it can be quickly decided whether it is worth developing further. Such trials for falsification are what either rejects a hypothesis, or confirms with time that it becomes a theory. The process of falsification is in our meaning at the very heart of scientific quality checking.

The CO2 hypothesis predicts higher temperatures in 50 or 100 years, and assumes a confidence in these long term predictions without falsification process. Many articles take on the duty of supporting this hypothesis, but well meant support in itself has no room in science. EJ cites as support for the CO2 hypothesis many points, among other that the warming since 1975 has been larger over land than seas. We agree with this. The problem with this type of argument is in the meantime that such increases would be observed whatever the cause of global warming, has it something to do with CO2 or not. A test must naturally be performed in conditions that are unique for the actual hypothesis. We are not seeing the defenders of the CO2 hypothesis proposing forecasts that can be tested in the present. Therefore the CO2 hypothesis remains as unscientific.

Beyond that, we can follow the challenge posed by EJ to be more concrete about some details in the CO2 hypothesis. We do not fight the need to predict global warming if more CO2 is present in the atmosphere. The problem lies in the fact that the effect of CO2 in itself is relatively weak, and that a potentially large hypothetical effect on the temperature can be reached only if many other processes unfold with the CO2 increase.

A supposedly important process is increased water evaporation, since water vapor is the most important greenhouse gas. The CO2 hypothesis assumes that the quantity of water vapor in the atmosphere will increase in pace with CO2 causing a slight temperature increase. Unfortunately analyses show that the quantity of water vapor in the atmosphere has decreased since 1948, in direct opposition with this basic condition [for the CO2 hypothesis to be true]. Cloud cover as well is important for the radiative balance, but this one cannot be modelled correctly either. This is not so surprising, since cloud formation takes place on a mm scale or less, whereas the global climate models (GCMs) have a horizontal definition of 150km.

A fundamental uncertainty is linked in the meanwhile to the question of whether the CO2 hypothesis is in any way empirically relevant; one can immediately realize that other hypotheses can be considered.

For example we could have a look at our own interglacial period, from which we have good temperature and CO2 data from ice cores. The diagram above shows the temperature and atmospheric concentration of CO2 through most of our current interglacial. The top figure shows the temperature changes at the top of the inlandsis in Greenland, which usually is considered as a rough proxy for the global temperature, with the exception of some parts of the Southern atmosphere, Antarctic included. The left hand scale shows the temperature at the top of the inlandsis, while the right hand side scale shows the approximate change of the mean global temperature relative to 1961-1990. The data series stop around 1855, and we have suggested how the overall temperature change until now would be with a dotted red line.

The lower part of the figure shows variations of the atmospheric CO2 as deduced from ice cores drilled on Dome C in the Antarctic. The atmospheric CO2 is usually considered to be well mixed (the differences being up to 4%), so that a reconstruction from the Antarctic provides with a good estimate for the variations of the global atmospheric CO2. These data cover a time span up to 1780; therefore the recent increase to 390 ppm is not seen. The latter would lie well above the scale limits, even if the simultaneous temperature increase (dotted red line) isn't especially anormal. The effect of CO2 is obviously not large.

On the figure it can be clearly seen that the atmospheric CO2 hasn't had a mentionable net effect throughout the current interglacial either. For instance, CO2 has exhibited during the last 6-7000 years a regular increase while the temperature has been dropping. The increase from about 260 to 280 ppm should according to the CO2 hypothesis result in a global temperature increase of about 0.3 degree. In reality the temperature has sunk by 0.8 degree. No variation of the CO2 concentration is observed related to the significant temporary temperature increases that have taken place every 1000 years for the last 4-5000 years. Other factors than CO2 must have been determining the recorded temperature evolution.

We respect the work put into building GCMs. In the meanwhile we do not have a reciprocal trust in the usefulness of their results. The net effect of CO2 is probably overestimated. What is crazy is supposedly that the assumed companion processes (among others, water vapor and clouds) do not take place as programmed in the GCMs. The past shows with high clarity (the figure above) that the current climate changes are not especially worth noting. Similar or larger changes have taken place earlier due to many different reasons, and these show generally no clear relationship with the simultaneous changes in CO2 concentrations.

We mean consequently that an obvious alternative explanation for the current climate is that it is mainly steered by natural variations. We have a larger (and growing) confidence in this explanation than in the CO2 hypothesis. And we are completely sure that the uncertainties (the one of the IPCC as well) must be computed following classical statistical rules.