After global warming began, the United States saw fewer extremely hot days

For decades, extreme heat has been presented as one of the most direct and damaging consequences of global warming, with potential impacts on human health, agriculture, infrastructure and energy systems. Yet one persistent methodological problem has remained: long-term temperature records are not equally homogeneous everywhere. Station placement, measurement standards, urbanisation and instrumentation changes can all affect the comparability of temperature data over time.

A new open-access paper published in Theoretical and Applied Climatology adds an important, but geographically limited, data point to this debate. Using an extended version of the United States Historical Climate Network, the study examined daily maximum summer temperatures and daily minimum winter temperatures across the conterminous United States from 1899 to 2025. Its central finding is counterintuitive: in this specific US dataset, several metrics of both extreme summer heat and extreme winter cold declined over the full period.

This does not mean that global warming has stopped increasing heat risks everywhere. The broader scientific assessment remains clear: the IPCC states that human-induced greenhouse gas emissions have increased the frequency and/or intensity of some weather and climate extremes, especially temperature extremes, since pre-industrial times. The key point is more nuanced: regional datasets can behave differently depending on geography, baseline period, local variability and the definition of "extreme".

Temperature risks: heat, cold and adaptation

Globally, non-optimal temperatures remain a major health burden. A large Monash University-led study estimated that more than five million extra deaths per year were associated with abnormal hot and cold temperatures in 2000–2019, with most deaths linked to cold exposure. At the same time, heat-related deaths increased across all regions over that period.

The temperature associated with the lowest mortality differs strongly by region. In warmer climates, populations may tolerate higher temperatures better; in cooler climates, the minimum-mortality temperature can be much lower. This "comfort point" can also shift over time as societies adapt through housing, healthcare, urban planning, behaviour and infrastructure.

Europe is a useful example of this complexity. Cold exposure still causes a large health burden, especially where housing quality and energy poverty are major issues. But Europe is also warming faster than the global average, and heat-related risks are rising. The European Environment Agency notes that both heat and cold increase preventable deaths, but warming is expected to increase heat-related deaths and not necessarily reduce cold-related deaths by the same amount.

What the US study found

The new US analysis focused on a relatively homogeneous long-term station network. According to the paper, the dataset includes more than a century of daily observations and extends the US Historical Climate Network back to 1899 and forward to 2025. The author calculated several metrics, including the hottest summer daily maximum values, the coldest winter daily minimum values, and the frequency of multi-day heat and cold waves.

The result differs from the simple expectation that all hot extremes should have become more frequent. Among the five years with the highest daily temperatures in the US record, only 2021 occurred after the modern period of rapid global warming had clearly begun. The other top years were 1930, 1934, 1936 and 1954, with 1936 standing out as the hottest in this specific metric.

Cold extremes declined as well. The coldest daily records were concentrated much earlier in the record, especially around the early 20th century and the 1930s. The study therefore argues that the US climate, at least by these specific daily-extreme indicators, has become less sharply variable at the tails: fewer record-level cold extremes and no clear rise in the most extreme hot days over the full 1899–2025 period.

One striking metric in the article is the narrowing of the 15-year moving gap between the hottest and coldest daily extremes. Over the 127-year period, that gap reportedly shrank by about 3.3°C, suggesting a reduction in the most severe continental temperature swings.

Heatwaves and cold waves

Single-day records matter, but from a public-health and infrastructure perspective, multi-day events are often more damaging. Heatwaves and cold waves put sustained stress on bodies, buildings, power grids, crops and transport systems.

The US study therefore also looked at periods of unusually hot or cold conditions lasting six or more consecutive days. According to the article, the highest frequency of heatwave days occurred in 1930–1944. It also reports that the combined frequency of hot and cold wave days declined from around 120 days per year in the 1930s to around 65 days per year after 1965.

The author interprets this as evidence that, in the conterminous United States, natural variability remains very large and can dominate long-term trends depending on the baseline chosen. If the analysis begins around 1960, heatwave days appear to rise slightly. If the full 1899–2025 record is used, the conclusion is different.

Why this should not be generalised to Europe

The US result is important, but it should not be mechanically transferred to Europe. Europe shows a different pattern. Copernicus notes that Europe has warmed at more than twice the global average since 1991, and that the frequency and intensity of extreme heat events are increasing. It also reports that 23 of Europe's 30 most severe heatwaves since 1950 occurred after 2000.

This is why the study is best read as a reminder about regional climate variability, not as a refutation of global warming or European heat-risk assessments. In Europe, the policy challenge remains two-sided: reduce heat exposure through urban cooling, early-warning systems, building adaptation and healthcare preparedness, while also addressing cold-related vulnerability through better housing, insulation and energy-poverty policy.

The broader lesson

The main takeaway is not that extreme heat is "no longer a problem". It is that climate-risk communication needs to be precise.

A long US station dataset can show declining daily temperature extremes over a specific historical period. At the same time, global and European assessments show increasing heat extremes and rising heat-related risks in many regions. Both statements can be true because they refer to different geographies, metrics and baselines.

For Europe, the evidence still points toward a future with more frequent, intense and dangerous heatwaves — especially in cities, southern regions, and ageing populations. For the United States, this paper adds a useful reminder that natural variability, historical drought periods such as the 1930s, and the choice of start date can strongly shape the interpretation of climate extremes.

illuminem Voices is a democratic space presenting the opinions of leading Sustainability Thought Leaders, their views do not necessarily represent those of illuminem.

The world needs sustainability knowledge. At illuminem, no interest group or shareholder can influence our work. Thank you for supporting our mission to make high-quality and independent sustainability information free for all. Every contribution helps. Thank you for donating today.