Last week we discussed the new paper by Prof. Qing-Bin Lu updating his “CRE” model comparisons against atmospheric data. That model asserts that halogenated greenhouse gases (like chlorofluorocarbons) are what drive short-term climate change, while non-halogenated ones are irrelevant… like CO2. And his idea of science is to test theories against one another by testing their predictions against evidence. Which might sound obvious but nowadays is controversial, especially in climate. So he makes this prediction: the Arctic has begun cooling and Arctic sea ice will not shrink further. Instead it will grow in the years ahead. Had anyone predicted that 20 years ago they’d have been laughed off the interwebs. And even today they’ll get sneered at… even though a new study (h/t NoTricksZone) acknowledges that yes, Arctic sea ice stopped disappearing 20 years ago, despite “overly alarmist” predictions that it would vanish. That study also admits that climate is complicated and sometimes doesn’t do what models predict. Well, IPCC models maybe. Of course in climate science you get to #HaveItBothWays so the authors argue that while most models predict ongoing Arctic ice disappearance, many predict a 20 year hiatus in Arctic sea ice loss even amidst a warming climate. But those ones also predict a rapid return to trend after 20 years. So we have two theories making opposite predictions about what will happen to the Arctic over the next five years, and one making no testable prediction. Let the science begin.

The new study, called “Minimal Arctic Sea Ice Loss in the Last 20 Years, Consistent With Internal Climate Variability” argues that the warming effect of greenhouse gases and the drop in Arctic sea ice from 1979 to 2000 “fueled predictions, which with hindsight look overly alarmist, that the Arctic would experience its first ice-free summer before 2020”. But the climate is complicated and there are natural variations at work that can lead to increased ice cover even when the air is warming.

The authors show that a trend line fit to the September minimum ice cover record over the past 20 years has a statistically insignificant slope. Meaning its wiggles and wobbles are not meaningfully distinguishable from zero. Is that result something IPCC models predicted? Um no. And it’s definitely the opposite of what a lot of scientists predicted. Though to be fair the authors dug through the data from a pile of models and found that there are at least some model runs (about 20%) where an interval of no change in minimum Arctic ice cover happens even during warming.

So most models predicted continued ice loss, but not all. But they’re not off the hook, because most of those models that predicted a slowdown over 2005-2024 also predict rapid melting over the next 5-10 years. Only about a quarter of the models that tracked the current hiatus in sea ice loss say the hiatus could run another 5-10 years. But for those ones the correct response isn’t that the science is settled and the melting is just briefly on hold, it’s that we have to wait a decade to see if they’re right.

Meanwhile most models failed to predict the hiatus in sea ice loss, and of the ones that did, most predict we should now see a rapid return to declining sea ice. Those will be the ones that will look the worst if Arctic sea ice continues to recover. And then what will they say?