[LINK TO THE HOME PAGE OF THIS SITE]

THIS POST IS A CRITICAL EVALUATION OF THE WAY THE STATISTICAL PROPERTY OF VARIANCE IS INTERPRETED IN CLIMATE SCIENCE WITH THE COMMONLY SEEN “LOOK HOW BAD IT COULD BE” AND THE “WORSE THAN PREVIOUSLY THOUGHT” ARGUMENTS.  

1. Climate science is corrupted by activism. As an example, consider how the activism needs of researchers corrupt the interpretation of the statistical property of variance. In statistics, as in information theory, high variance implies uncertainty and therefore low information content. Uncertainty means our information lies somewhere between knowing and not knowing such that the higher the variance the less we know.
2. In this context high variance is undesirable because it degrades the information we can derive from the data. However, high variance also yields large confidence intervals that invite the confirmation bias of the researcher to interpret the portion of the interval that supports prior beliefs and activism needs. Thus, high variance can be interpreted not as uncertain information but as certainty of the danger of how extreme it could be. This anomalous understanding of variance is the norm in climate science where the word “could” is used and understood as certainty in the context of the precautionary principle.
3. This interpretation of uncertainty takes on the incredible power of information when the precautionary principle is invoked. The precautionary principle says in effect that if there is any chance at all that this could happen it must be treated as a certainty and it is the norm in certain risk assessment models as well as in superstition. The role of the precautionary principle in superstition and confirmation bias is explored in a related post [LINK] .
4. In climate science, the precautionary principle is subsumed in most cases in terms of a risk assessment assumption that the alternative to the appropriate level of  climate action is the destruction of the planet. This assumption leads to a perverse interpretation of uncertainty in climate science such that uncertainty becomes transformed into certainty of extreme values. The climate science position that the less they know the greater the risk of harm by climate change follows from this assessment.
5. This principle is pervasive in climate science such that it follows that the the truth of AGW climate change theory and it’s catastrophic impacts and consequences implicitly becomes the null hypothesis and the null hypothesis of the truth of climate catastrophe can be rejected only if convincing evidence can be provided by climate deniers against it. This methodology is claimed to be science although it is exactly the opposite of science.
6. It works exactly the other way around in science where the null hypothesis is the absence of the effects predicted by theory and it is necessary for empirical tests to provide convincing evidence against it so that the null can be rejected. It is only then that we can accept the alternate hypothesis that the proposed catastrophic AGW climate change theory is correct.
7. An uncertainty problem that has been difficult and contentious for climate science is that of climate sensitivity (ECS), a proposed relationship between the logarithm of atmospheric CO2 concentration and surface temperature at the appropriate time scale that sits at the foundation of AGW theory. A large range of values for the ECS parameter and the appropriate time scale in the observational data is found in the literature as described. in a related post [LINK] . Such a large range of observed values means that both high warming at low CO2 & low warming at high CO2 are possible and that therefore the ECS is not a useful concept.
8. But in climate science the high end of the range is what matters because it provides the condition that should be taken into account in terms of the precautionary principle. The ECS uncertainty issue is a high profile example of a parameter which contains little useful information because of uncertainty but is treated as certainty of high values.
9. It is this faux understanding of variance in climate science that leads to the oft repeated pattern in which first a climate “tipping point” is declared, meaning that we have run out of time to act on climate and that therefore the planet is about to be destroyed. The declaration is quickly followed by its retraction once climate scientists realize that they have removed all motivation for climate action. The retraction is invoked with the statement that “there is still time” if we act quickly and decisively to limit or eliminate our use of fossil fuels. The pattern likely derives from the failure of climate science to understand the concept of variance in statistics.
10. The climate science view of uncertainty facilitates confirmation bias because the interpretation of the uncertain information can become biased by the beliefs, biases, and activism priorities of the researcher.
11. Briefly, the odd interpretation of uncertainty in climate science is that large uncertainties are not interpreted in terms of the absence of information but as information about how extreme it COULD be. For example a  2017 paper climate science uses a “climate-carbon-cycle” model to find that that the carbon budget for limiting warming to 1.5C is 920 to 1,980 gigatons of carbon dioxide emissions for the period 2016 to 2100. In any other science discipline an uncertainty this large would imply that “we don’t know”. There is not enough useful information in an estimate with such a large variance. But the climate science conclusion is that a budget of the extreme value of 920 gigatons give us a pretty good chance of staying below the 1.5C target“.
12. WORSE THAN PREVIOUSLY THOUGHT: Another form of this uncertainty issue is the odd bias in climate science which holds that forecasts that turn out to be wrong on the safe side such that the actual event was worse than what climate science had predicted, is presented proudly proclaimed as “WORSE THAN PREVIOUSLY THOUGHT” with the prediction error interpreted as “EVEN MORE RIGHT THAN PREVIOUSLY THOUGHT“. This  biased interpretation of uncertainty provides more evidence of a deeply seated CONFIRMATION BIAS principle in climate science methodology.
13. If it is worse than previously thought it does not mean that climate science was even more right than previously thought. It means that they were wrong and that implies that their understanding of the phenomena of nature at issue is flawed. Yet in all these situations, confirmation bias creates correctness out of error as in the interpretation of large confidence intervals in terms of its most harmful end.

BRIEFLY, UNCERTAINTY DOES NOT MEAN OH! LOOK HOW HIGH IT COULD BE. IT MEANS WE DON’T REALLY KNOW. THE LESS WE KNOW THE HIGHER IT COULD BE AND IN PERFECT IGNORANCE IT COULD BE AS HIGH AS INFINITY

BECAUSE THE ANSWER IS NOT CONSTRAINED BY INFORMATION 

RELATED WORKS

CIRCULAR REASONING IN CLIMATE SCIENCE: [LINK] 

CLIMATE SCIENCE CORRUPTED BY ACTIVISM: [LINK]

CONFIRMATION BIAS IN CLIMATE SCIENCE [LINK]