Bart,
Did you notice the following reference to uncertainties in that report?
Major uncertainties
"Key remaining uncertainties relate to the precise magnitude and nature of changes at global, and particularly regional, scales, and especially for extreme events and our ability to simulate and attribute such changes using climate models."
Yes, I've seen it, and what's more, I've understood what they are saying.
For instance, and I'll focus on one subject that you used/misinterpreted from the IPCC reports;
"
Key Finding 2The frequency and intensity of
extreme heat and heavy precipitation events are increasing in most
continental regions of the world (
very high confidence). These
trends are consistent with expected physical responses to a warming climate. Climate model studies are
also consistent with these trends, although models
tend to underestimate the observed trends, especially for the increase in
extreme precipitation events (very high confidence for temperature, high confidence for extreme precipitation). The
frequency and intensity of
extreme temperature events are
virtually certain to increase in the future as
global temperature increases (
high confidence). Extreme precipitation events will
very likely continue to increase in
frequency and intensity throughout
most of the world (high confidence)."
Then follows the sentence that explains some the difficulties you cherry-picked from the report.
"Observed and projected trends for some
other types of extreme events, such as floods, droughts, and severe storms, have more variable regional characteristics."
Because of the more variable
regional characteristics, it's harder to qualify it with high Confidense for a
global environment. Also, because they are extremes, their frequency of occurrence is generally lower than that of other events, which has an effect of the statistical 'confidence level' (meaning that on a local basis they are too infrequent to have reliable statistics, with a high confidence level), hence the lower
confidence levels that are
given to the data sets. In contrast, the
likelihood of these extreme events happening is "
very likely" to increase.
That key finding, which you quoted out of context, is then further qualified as follows:
"
Major uncertainties Key remaining uncertainties relate to the
precise magnitude and
nature of changes at
global, and
particularly regional, scales, and
especially for extreme events and our ability to simulate and attribute such changes using climate models."
So, given the
regional differences which reduces the confidence level of the dataset, it is indeed harder to predict the
exact Magnitude of such events especially at a
Global level. That's a very logical result and not hard to understand for those who understand the terminology that is used. Again, 'Confidence Levels' are about datasets and consensus, 'Likelihood' is about the probability of occurrence.
So while a Confidence level of a dataset may sometimes be lower (due to a more limited number, or the changing location, of observations), Likelihoods can be very or extremely high. That's almost statistics 101.
Another thing is that as instrumentation improves, it is found that the improved quality and frequency is usually fully in line with the trend from earlier observations, hence the higher Likelihood, despite the missing observations in the past (which lowered the confidence level of the earlier data-sets).
Sometimes better quality data leads to a discovery that leads to adjustments of prior data-sets, e.g. in the case of ocean temperatures measured too close to the exhausts of the ship, compared to actual water-temperatures measured with growing numbers of buoys.
Other remaining issues that affect confidence levels in data-sets are when local conditions are influenced by unwanted bias sources, e.g. expanding cities which create their own (heat-island) micro-climate, or when rising water levels start flooding observation posts near those water-bodies and the observation posts need to be relocated (which reduces the confidence level of the data).
Cheers,
Bart
