The last six months of 2023 weren’t just hotter than we’ve seen before, they were so much hotter it’s alarming. September in particular spurred Zeke Hausfather to call it absolutely gobsmackingly bananas, and compared to previous Septembers it rather was:
I’ve used the data from ERA5 since they’ve already reported their December value.
I know how easy it is to get such a startling result, even when there’s not really a “startling” underlying cause. Data sets which offer a lot of choices of subsets — such as, “which month shall we look at?” — are especially vulnerable to this. So, I can only approach the September result as suggestive but not conclusive.
We can also look at the average over each half-year:
The value for the 2nd half of 2023 is definitely a record-setter, and yes it’s surprisingly high, but it’s not quite what I’d call “bananas.” It’s very interesting to look, not just at the September-only graph, but at the same for all 12 months:
February and March, and to a lesser deree January and April, show the extreme heat from the previous el Niño back in 2016. This is because the el Niño which started in 2015 had its peak impact the following year during those months of winter/spring, as el Niños often do. The year 2023, however, shows extreme heat in September through December, in spite of the fact that the el Niño had only begun and its biggest impact is generally delayed by months.
If this el Niño follows the pattern of its predecessors, then 2023’s high heat is the precursor of its full effect in 2024. Just as 2015 set a new record, but was only the precursor to 2016, which broke that record by a large margin. If that is repeated, 2023 might be “crazy hot” but 2024 will blow it away.
The adjusted data are also very revealing on a month-by-month basis:
This makes a good case that the adjustments have done a fair job removing the influence of el Niño; those spikes in 1998 and 2016 are mostly gone. But the extremes over the last six months of 2023 are still there. According to the adjustment procedure, 2023’s record-breaking heat was not because of el Niño.
If the future is like the past, el Niño‘s hottest heat is coming this spring to a planet near you. That’s why some, including James Hansen and colleagues, are expecting temperature in early 2024 to be “off the scale,” even by the standard set in 2023. If so, it will be further strong evidence (but not quite proof) that the reason 2023 and 2024 were so hot, is that global warming picked up speed.
We’re already dangerously close to the 1.5°C threshold, and the litany of billion-dollar weather disasters in the USA this year is testament to just how costly conditions already are. The hotter it gets, the worse things get, and if that is happening faster now than it was, our wild ride will be too fast and too furious.
As for this spring blasting past existing world records … time will tell. Unlike most testable forecasts from climate science, in this case we won’t have to wait very long.
You can help this blog! Feel free to
donate at my wee dragon
Open Mind 
Thanks. James Hansen thinks we’ve already effectively shot the 1.5°C target. It’s hard to disagree with him. You said we passed 2°C on two days, though it must have been very close on many more days as ClimateReanalyzer has more than 7 days with an anomaly of 1.4°C or more, with 1979-2000 as the baseline. Hansen thinks 2°C will be shot within a decade.
I’m just hoping that there is something we don’t yet understand about this year’s exceptional warmth (land and sea). I note that ClimateReanalyzer is now showing Jan 3rd as almost not the warmest. Fingers crossed.
PS I noticed “deree” is missing a “g” in the post!
Regarding 2024: the question is did this ENSO cycle just fire early (so a shift to an earlier peak, in which case we’ll soon see temperatures come back down to “normal”*) or as you state, is the peak still to come, in which case we will get another couple tenths of a degree warming in Jan-March, blowing away all previous temperature records.
The ClimateReanalyzer Jan 3rd data is tentatively suggestive of the latter, but too early to tell for sure – back in December, there were a few days when the temperature was the second warmest, and December as a whole was still a record breaker by a fair margin.
In any case, thank you as always for your great analyses.
*Where “normal” is still way too warm relative to preindustrial.
Whoops. I meant, the ClimateReanalyzer Jan 3rd data is tentatively suggestive of the former (timing shift for the peak). But in any case, still way too soon to tell.
Excellent post!