In 2018, with Hurricane Florence bearing down on the Carolinas, a group of scientists cranked up their supercomputer and announced that global warming had caused the hurricane to carry 50 percent more rain than it otherwise would have. But a little-noted asterisk said their analysis was just a preliminary estimate, rushed into print while the hurricane was still news. With the benefit of more time to check the numbers, the same scientists have revised their estimate... downwards. Way downwards. Their computer now says Florence might have been 5 percent wetter. Or possibly not wetter at all. Oops.

Their original analysis also declared Hurricane Florence was 80 kilometers wider than it would have been without global warming. But once the errors were fixed the new estimate was about 9 km, plus or minus 5 km. Another factor-of-ten error. Note the huge headlines not reporting the change.

Also, is it just us or do these kinds of errors all seem to go in the same direction? And while the authors were lauded for publishing the corrected analysis, we can't help but ask why they rushed the study into print in the first place. Obviously they wanted to catch the news cycle around Hurricane Florence while it was still charging towards the coast at full strength. But meeting a news deadline hardly seems like a priority for science research.

Let alone one that trumps getting the right answer.