Global Warming Intensifies Extreme Day-to-Day Temperature Changes in Mid–Low Latitudes
- Apr 8
- 6 min read
A 2025 study finds that extreme day-to-day temperature swings have intensified since the 1960s across low and mid-latitudes.
When most people think about climate change and extreme temperature, they picture heatwaves — sustained periods of dangerous heat. But there is another dimension of temperature extremes that receives far less attention: the abrupt, large swings in temperature that occur between consecutive days. In their study, Liu et al. (2025) argue that these extreme day-to-day temperature changes (DTDTs) — defined as daily maximum temperature differences between consecutive days that exceed the 90th percentile of historical records — constitute a distinct and largely overlooked category of climate extreme.
Using six decades of observational data, reanalysis datasets, and CMIP6 model simulations, the study documents how these events have already intensified across the world's most populated regions, attributes the changes primarily to greenhouse gas forcing, and projects their continued amplification through the end of the twenty-first century.
Key Findings
A Distinct and Independent Extreme
Extreme DTDTs are shown to be largely independent of the 15 standard temperature-related indices maintained by the Expert Team on Climate Change Detection and Indices (ETCCDI). Across nearly all global land areas, only 1–8% of grid points show highly significant correlations between extreme DTDT amplitude and any individual ETCCDI index, meaning 90–99% of locations exhibit no meaningful association — confirming that extreme DTDTs represent a fundamentally distinct category of weather extreme.
Importantly, extreme DTDTs exhibit a stronger association with mortality than either minimum temperature extremes or the diurnal temperature range (DTR) — the existing standard metric for intraday temperature variability — in both the contiguous United States and eastern China. Above the 90th percentile threshold, mortality risk rises in a near-exponential fashion with increasing DTDT magnitude.
The DTDT index also shows significantly greater statistical instability than the DTR, and the two metrics respond asymmetrically to global warming: while minimum temperatures have risen faster than maximum temperatures — driving a well-documented decline in DTR — the DTDT index has been increasing, reflecting a divergent and physically distinct process.
Observed Trends Since 1961: A Divided Planet
Analysis of Berkeley Earth observational data and two reanalysis products (ERA5 and NCEP/NCAR) covering 1961–2020 reveals a globally consistent but geographically divided pattern of change. Low- and mid-latitude regions have experienced robust increases in the amplitude, frequency, and total intensity of extreme DTDTs, while high-latitude regions — particularly the northern high latitudes — have seen a decrease.
The amplitude, frequency, and total intensity of extreme DTDTs in low and mid-latitudes increased at rates of approximately 0.03–0.04°C, 0.4–0.5 occurrences, and 4–5°C per decade, respectively.
The regions showing the most pronounced increases are the western United States, eastern China, South America, and the Mediterranean. The increases in total intensity in these regions are particularly striking: 11.1°C per decade in the western USA, 9.4°C per decade in eastern China, 12.4°C per decade in South America, and 7.1°C per decade in the Mediterranean.
These trend patterns are consistent across all four seasons and are robust to different dataset choices and percentile thresholds (95th, 98th, and 99th percentile results confirm the same spatial pattern).
Changes in extreme DTDT variability account for approximately 80% of the total increase in DTDT at low and mid-latitudes and roughly 70% of the decrease at northern high latitudes.
Attribution: Greenhouse Gases Are the Primary Driver
CMIP6 model simulations under combined anthropogenic and natural forcing (ALL) successfully reproduce the observed spatial pattern of change, with a pattern correlation of 0.47 between the Berkeley Earth observations and the multi-model ensemble mean — statistically significant at the 99% confidence level.
Critically, simulations driven by greenhouse gas forcing alone (GHG-only) closely replicate both the ALL-forcing results and the observational data, strongly implying that GHG emissions are the dominant driver. Anthropogenic aerosol (AA) and natural forcing (NAT) runs produce predominantly insignificant changes.
Formal optimal fingerprinting analysis — using single-signal, two-signal, and three-signal detection approaches — confirms that the GHG signal can be clearly separated and detected from the effects of aerosols and natural variability. The observed increase in extreme DTDT amplitude across both Berkeley Earth and ERA5 datasets during 1961–2014 is robustly attributable to anthropogenic GHG forcing.
Physical Mechanisms: Drying Soils and Rising Pressure Variability
The study identifies a chain of thermodynamic processes linking GHG-driven warming to increased extreme DTDTs at low and mid-latitudes. Anthropogenic warming reduces mean soil moisture and increases variability in both sea-level pressure and soil moisture — changes that together amplify day-to-day fluctuations in cloud cover, downwelling radiation, and precipitation.
In the tropics and southern subtropics (40°S–20°N), enhanced variability in sea-level pressure and meridional wind emerge as the dominant contributors, reflecting how global warming intensifies tropical convective activity and synoptic-scale atmospheric disturbances.
In the Northern Hemisphere subtropics (20°N–40°N), the picture shifts: decreased mean soil moisture and increased soil moisture variability play the leading role, as drier soils reduce surface heat capacity and amplify temperature swings.
The decrease in extreme DTDTs at northern high latitudes follows a different mechanism — Arctic amplification, which warms northerly cold air masses more rapidly than southerly warm air masses, weakening the meridional temperature gradient and reducing subseasonal temperature variance.
Future Projections: Intensification for 80% of the World's Population
CMIP6 models project continued and widespread intensification of extreme DTDTs under both SSP 2-4.5 and SSP 5-8.5 scenarios through 2050–2099, affecting low- and mid-latitude regions and broad land areas in the Southern Hemisphere — regions that collectively encompass over 80% of the global population.
Under the high-emissions scenario (SSP 5-8.5), the spatially averaged frequency, amplitude, and total intensity of extreme DTDTs in low and mid-latitudes are projected to rise by approximately 17%, 3%, and 20%, respectively, by 2100.
In the future, changes in extreme DTDTs will account for approximately 80% of the total increase in DTDT amplitude and virtually all (∼100%) of the total intensity change in low and mid-latitudes.
Record-Breaking Events and Shrinking Return Periods
Two record-breaking spring extreme DTDT events occurred in 2022: eastern China on 16 March (temperature swing of 22.9°C, or 3.16 standard deviations above the mean) and the western United States on 20 May (20.3°C, or 3.18 standard deviations above the mean). These surpassed the previous records for each region — 20.4°C in eastern China and 20.0°C in the western USA.
Return period analysis reveals a dramatic acceleration in the likelihood of such events. Events of this magnitude occurred once every 1,000–3,000 years during 1950–1985. By 1986–2021, the same events were occurring once every 40–60 years — a reduction in return period by a factor of tens to hundreds, driven by anthropogenic warming.
A Climate Risk Hiding in Plain Sight
Despite their strong associations with mortality, disease, ecosystem disruption, agricultural stress, and even economic contraction, extreme DTDTs remain absent from the standard toolkit of climate monitoring indices. This study makes the case that they deserve formal recognition alongside more familiar extremes:
Public Health: Extreme DTDTs trigger and exacerbate a range of health outcomes — respiratory and cardiovascular mortality, influenza spread, allergic responses, and immune suppression — with mortality risk rising near-exponentially above the 90th percentile threshold. The health burden falls disproportionately on the populations of low and mid-latitudes, where these events are already intensifying.
Ecosystem and Agricultural Stress: Terrestrial ecosystems and agricultural systems are sensitive to high-frequency temperature variability, and intensifying DTDTs threaten to compound the stresses already imposed by rising mean temperatures.
Forecasting and Prediction: Abrupt day-to-day temperature swings challenge the skill of modern weather forecasting systems, underscoring the need to improve prediction capabilities specifically for this type of event.
Policy and Monitoring: The findings call for DTDT extremes to be incorporated into global climate monitoring frameworks alongside the ETCCDI indices, and for adaptation strategies — particularly in public health and agriculture — to account for this distinct dimension of climate change.
When Tomorrow Feels Like a Different Season
The warming world is not only getting hotter on average — it is becoming more volatile from one day to the next, in ways that existing climate indices do not fully capture. As highlighted in a recent study, for the billions of people living in low and mid-latitudes, the danger is not only in the heat itself, but in the unpredictability of daily temperature swings that tax the human body, stress ecosystems, and disrupt the rhythms of agriculture and society. As greenhouse gas emissions continue to drive these extremes toward new records, the 2022 events in China and the United States offer a preview of what is becoming increasingly ordinary.
Sign up for our newsletter or connect with us on social media to stay up-to-date with our latest posts and permaculture inspiration.
Explore our inspiring series and posts:
Love the post? Share it with your circle, inspire your people: