Science1 publisher2 min readPublished
A grid-point analysis dates regional warming speedups from the 1980s through the 2010s
Claudie Beaulieu's team fit statistical models at every grid point of 55 years of surface temperature data and found warming speedups arriving in separate regional waves, the most recent in midlatitudes such as southeast China.
The Scientist · Science desk

What happened
- Researchers at UC Santa Cruz, Clemson University and the NOAA Geophysical Fluid Dynamics Laboratory fit statistical models to multiple observational surface-temperature data sets covering 1970 to 2024.
- Lead author Claudie Beaulieu said the regional speedups emerged in distinct waves, with onset timings running from the 1980s through the 2010s depending on the geographical area.
- The detected onset timings are consistent with declining sulfur dioxide emissions from Chinese coal-fired power plants, though the authors say the mechanism behind every speedup still needs study.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- decision A plan dated off a global-mean acceleration gives every jurisdiction the same start year; on these results an authority whose local speedup began in the 1980s and one whose speedup began in the 2010s are designing against records of very different length.
- constraint Heat return-period tables built on a steady warming assumption will run behind local reality after an onset date, because a faster rate lifts record-breaking odds more than proportionally.
- exposure The sharpest exposure falls on populous regions with low socioeconomic development, where a revised extremes baseline meets the least capacity to rebuild for it.
- capability With the dashboard, a local agency can check the warming rate over its own map location rather than wait for the next verdict on the global average.
Thirty years separate the earliest wave from the latest. Take the middle of the 1980s and the middle of the 2010s and the gap is 30 years [17], inside a record that runs 55 years, from 1970 to 2024 [16][2]. Two grid cells can both be flagged as accelerating and disagree about the start date by more than half the length of the window being analyzed. Earlier work by members of the same team asked one question of the global mean: is an acceleration detectable at all [15].
The team fit models separately at each location in several observational surface-temperature data sets, then looked for coherent spatial and temporal patterns in where warming rates shifted [2]. "Fitting our models at each grid point allows us to identify regions with similar warming behavior," said co-author Rebecca Killick, a professor of statistics at Clemson University [7].
The detected timings are consistent with declining sulfur dioxide emissions from coal-fired power plants in China [9], and the authors note that localized drivers such as regional aerosol pollution reductions may play a role in some of the speedups but not all [8]. Beaulieu and her co-authors say pinpointing the physical mechanisms behind every regional speedup requires further study [10]. For a planner that gap is the difference between a speedup that will persist and one that was a one-time step, and the paper reports when and where rates changed [18].
The onset date matters for planning. Steady warming on its own shifts the baseline for the frequency and severity of heat waves, and an increase in the rate of warming can raise the likelihood of record-breaking events disproportionately [11]. Ecosystems are often more sensitive to rapid temperature shifts than to gradual ones [13]. The authors single out regions with large populations and low socioeconomic development as especially vulnerable to accelerated warming and the temperature extremes that come with it [12].
The most recent accelerations show up in Northern Hemisphere midlatitudes, southeast China and the Gulf of Mexico among them [3], with the eastern Mediterranean, Bolivia, the North Pacific and New Zealand also flagged [4]. Beaulieu, the lead author and an associate professor of ocean sciences at UC Santa Cruz [6], said she hopes the work, "along with an interactive public dashboard to enable the public to continue monitoring local rates of warming by clicking on a map," will help target efforts where they are most needed [14]. The paper is "Space-time signatures of surface warming accelerations since 1970," in Nature Communications [19].
What to watch
- Whether modeling work confirms the aerosol link for southeast China, turning an alignment of dates into a demonstrated mechanism.
- Whether the onset dates hold as 2025 and later observations enter the dashboard, especially for the regions dated to the 2010s.
- Whether national and regional adaptation assessments begin citing local onset years instead of a global acceleration date.