Guan, ShoudeShoudeGuanHuang, MengyaMengyaHuangCai, WenjuWenjuCaiZhang, ZhengguangZhengguangZhangLin, I-II-ILinKim, Hyun-SookHyun-SookKimZhou, LeiLeiZhouLin, XiaopeiXiaopeiLinXu, ZhaoZhaoXuJin, Fei-FeiFei-FeiJinMei, WeiWeiMeiWang, QianQianWangZhou, ChunChunZhouMeng, ZeZeMengTian, JiweiJiweiTianZhao, WeiWeiZhao2026-02-262026-02-262026-01-08https://www.scopus.com/pages/publications/105027245961https://scholars.lib.ntu.edu.tw/handle/123456789/736024Sea surface temperature directly beneath tropical cyclones is crucial for their intensification. In the long term, global warming heats the surface oceans, intensifying tropical cyclones, whereas concurrently with a cyclone, inner-core surface cooling is induced by the cyclone itself curtailing its intensification. However, the magnitude of cyclone-induced cooling, or the trend in storm-local sea surface temperature, remains uncertain. Here we provide a quantification using global surface drifter data from 1992 to 2021. We find that storm-local sea surface temperatures are rising at 0.29 ± 0.07 °C per decade—about twice the average rate in tropical cyclone-active regions despite enhanced cyclone-induced cooling; furthermore, the magnitude of cyclone-induced inner-core cooling is far smaller than previous estimates. The inner-core cooling measured by drifters is −0.68 ± 0.04 °C, substantially less than microwave satellite estimates (−1.05 ± 0.06 °C). State-of-the-art climate models tend to overestimate inner-core cooling while underestimating storm intensity. These findings offer observational benchmarks for models and suggest that current projections may underestimate the strength, frequency and impacts of major tropical cyclones.climate modelingcoolingdrifterglobal warmingsatellite altimetrysea surface temperaturestormtropical cyclonejournal article10.1038/s41561-025-01879-x