> based on studies of global climate models I'm Jack's complete lack of fucking surprise. It looks, also unsurprisingly, that the "models" that they use to get these high end guesses rely on the utterly preposterous RCP8.5 forcings. It's all just utter nonsense. ... wait a second... it's even fucking worse... This appears to be the main paper cited: https://www.pnas.org/doi/10.1073/pnas.1817205116 From the abstract: > Future sea level rise (SLR) poses serious threats to the viability of coastal communities, but continues to be challenging to project using deterministic modeling approaches. Nonetheless, adaptation strategies urgently require quantification of future SLR uncertainties, particularly upper-end estimates. Structured expert judgement (SEJ) has proved a valuable approach for similar problems. Our findings, using SEJ, produce probability distributions with long upper tails that are influenced by interdependencies between processes and ice sheets. They are using "Structured Expert Judgement"... which is effectively just the "best guesses, with wide error bars" of some selected "experts", they then take the tails of these ranges to use as the upper bounds. They are fucking using the most wild guess of the most alarmist expert they could find. From the supplimental material of that paper: > An expert’s statistical accuracy is the P-value (column 2 in Table S2) at which we would falsely reject the hypothesis that an expert’s probability assessments are statistically accurate. Roughly, an expert is statistically accurate if, in a statistical sense, 5% of the realizations fall beneath his/her 5th percentile, 45% of the realizations fall between the 5th and 50th percentile, etc. **High values (near 1) are good**, low values (near 0) reflect low statistical accuracy. >Six of the 13 US experts had a statistical accuracy score above 0.01. This is a high number for SEJ studies, especially considering the fact that 16 calibration variables were used, constituting a more powerful statistical test than the traditional number of ten calibration items. Two of the eight EU experts had a statistical accuracy score above 0.01, which is in line with most SEJ studies For a metric where values near 1 are "good" and values near 0 are "bad", they are taking any "expert" with a score over 0.01 as good. What a fucking joke. I also downloaded the SEJ data files. They are literally just asking the "experts" to type in their guesses of sea level rise and then using this as "data".