Article by Dr. Christopher Callahan and Prof. Andrew Dessler republished from Carbon Brief under a CC license. 

In June 2026, a record-breaking heatwave swept across Europe, with France among the first and hardest hit countries. 

In a new analysis, we estimate that the extreme conditions caused more than 2,700 heat-related deaths in France.

We also show how France’s extreme temperatures in June exceeded projections from climate models.

Our findings illustrate the human toll of extreme weather as the world warms.

We also highlight the challenges in projecting the magnitude of future heatwaves and their impacts on people.

Outpacing projections

For most of this century, Europe has seen summer heat extremes that outpace projections from climate models. 

Several different factors likely explain this trend, including reductions in planet-cooling aerosols as nations have cleaned up their air pollution, as well as changes in atmospheric circulation patterns, which models struggle to represent.   

In June 2026, daily high temperatures averaged across France reached 36.9C, shattering the previous June record set in 2022 by 2.4C.

[For more on the impacts and coverage of Europe’s June heatwave, see Carbon Brief’s explainer.]

The rise in observed temperatures in France has outpaced projections made by climate models, with June maximum temperatures more in line with what was expected for the 2070s.

This is illustrated in the figure below, which shows how France’s average maximum daily high temperature for June recorded in 2026 (black line) compares to climate model projections (blue and orange lines).

Counting the death toll of climate change

The downstream impacts of these extreme temperatures are lethal. 

Scientists are able to estimate the death toll of high temperatures in many locations, depending on the availability of mortality and climate data. 

There are several ways to do this. 

One option is to examine death certificates to see which deaths have been directly recorded by physicians as related to heat. However, there is strong evidence that this method significantly undercounts heat-related deaths, as most death certificates do not consider environmental factors such as heat when diagnosing the cause of death.

Alternatively, it is possible to calculate the rate of total (“all-cause”) mortality in a given time period relative to previous time periods – for example, by comparing the total number of deaths in June 2026 compared to the average of previous Junes. This “excess deaths” figure can be used as an estimate of the deaths from a heat wave. 

Using this approach, Public Health France attributed around 2,000 deaths in France to the extreme heat in the week of 22-28 June.

Finally, scientists can use long-term data on overall mortality and correlate changes in mortality with changes in temperature to understand the statistical relationship between the two. 

Research published in Proceedings of the National Academy of Sciences in 2025 that used this third approach found that mortality rates in France increase rapidly in cold or hot conditions as daily maximum temperatures depart further from approximately 20C.

This pattern of a U-shaped response of mortality to temperature – shown in the figure below – is very consistent across time periods and regions around the world. 

To calculate the death toll of the June 2026 heatwave in France, we compared observed temperatures over 12-29 June to their baseline average over 1980-2025. 

The difference between these two temperatures helps us understand how many more people died than they would have in the absence of such extreme conditions.

Over 12-29 June, we found that France has experienced around 2,700 heat-related deaths above the average baseline. Day-to-day heat-related mortality rates rose from less than 100 to almost 300 on the hottest days of 24 and 25 June.

This is shown in the graph below, which illustrates the cumulative total heat-related deaths seen in France over the two-and-a-half week period. The inset shows how heat-related deaths fluctuated on a day-to-day basis during this time.

Recent analysis by World Weather Attribution has already shown that human-caused climate change increased the frequency and intensity of the June heat wave across Europe.

Meanwhile, previous research has shown there is substantial evidence that heat-related mortality in Europe has already been elevated by greenhouse gas emissions. 

As a result, we can be confident that at least some of the more than 2,700 deaths already seen in France are directly due to the burning of fossil fuels.

Calculating climate risk

In April, the UN-led body responsible for coordinating the work of climate modelling centres – the Coupled Modelling Intercomparison Project (CMIP) – unveiled a set of seven new emissions scenarios.

These are designed to replace the previous scenarios that have been used by scientists to understand how the climate might change in the future. They will feed into the upcoming seventh assessment report (AR7) of the Intergovernmental Panel on Climate Change (IPCC).

The range of future emissions in the new CMIP scenarios is smaller, with scenarios of very high or very low emissions no longer on the table.

The retirement of the very-high emissions scenario – known as “RCP8.5” – led to certain commentators in the media and in politics, including US president Donald Trump, arguing that the risks of climate change had been “overstated”.

[For more on false and misleading claims around the new emissions scenarios, see Carbon Brief’s factcheck.]

Our analysis of June’s heat-related deaths in France suggests that, even if the most severe emissions pathways are no longer needed, climate impacts are taking a heavy toll on society. 

Moreover, the temperatures seen in France show that climate models continue to underpredict the magnitude of heatwaves for a particular level of global warming.

This is because greenhouse gas emissions are only a first step in estimating the impacts of climate change.

The second step is converting emissions to changes in the climate at both the global and local levels – or hazards. This includes heatwaves, flash floods and droughts.  

The third step is to determine how changes in the hazards will affect local populations. This can be determined by calculating people’s exposure and vulnerability to hazards.

Substantial uncertainty persists at every stage of this sequence. 

For example, scientists do not know exactly how the global climate will react to ever-rising greenhouse gas emissions – nor the extent to which global temperature increases will drive local climate hazards. We also do not know how climate change at a local level impacts human health outcomes.

Managing the future of heat risk

Almost all heat-related deaths are preventable. 

Adaptation options, such as air conditioning, heat action plans and social support for isolated people, will be crucial as the climate moves away from the typical conditions that people are used to.

Our previous research showed that France made a lot of progress reducing heat-related mortality after the deadly 2003 summer heatwave by taking many of these actions. 

Adaptation can reduce deaths, but it cannot eliminate the risk created by continued warming. 

Without a move away from fossil fuels, future heatwaves will keep testing the limits of public health systems and more people will die.

Article by Dr. Christopher Callahan and Prof. Andrew Dessler republished from Carbon Brief under a CC license. 

• Media reaction: How climate change intensified Europe’s record-breaking June heat
• EM-DAT: Trump aid cuts could close database storing ‘world’s memory of disasters’
• Media reaction: UK and Europe’s ‘mind-boggling’ May heat and climate change
• Traditional models still ‘outperform AI’ for extreme weather forecasts