How the effect of climate change is measured in each heatwave
In June 2026, a large part of Western Europe experienced three days of extreme temperatures that, compared to the climate of 1976, would have been impossible without climate change at work. In parts of Europe, nighttime temperatures during heatwaves have come close to 30 degrees; on Spain’s Mediterranean coast they’ve brushed 40 at midnight. A team of researchers from Sweden, Denmark, the United States, the Netherlands, Ireland, and the United Kingdom analyzed the episode just days after it ended and calculated that the nighttime temperatures of that June would have been more than a hundred times less likely in the climate of 2003, and the daytime ones about ten times less likely [1]. This type of calculation is called attribution science.
What exactly it means to attribute an event to climate change
Until a few years ago, when a climatologist was asked whether extreme temperatures were caused by climate change, the only honest answer was that it couldn’t be known for certain. A weather event can happen by chance; there was no recognized way to isolate the cause.
That changed starting in 2004, when Peter Stott, Dáithí Stone, and Myles Allen published in Nature the first formal attribution study, focused on the 2003 European heatwave, which caused more than 70,000 deaths in France, Germany, and Italy. Their conclusion was that human-caused warming had, at minimum, doubled the risk of such an extreme summer occurring [2].
The method involves simulating the real world, with the full load of greenhouse gases emitted since the preindustrial era, and comparing it against a parallel simulated world, identical except for one detail: no emissions. The probability of the same extreme event occurring is calculated in both scenarios, and from there, it’s stated how many times more or less likely the event is in today’s climate. For example, a ratio of 30 means the event is 30 times more likely under current warming than without it.
The World Weather Attribution group, founded in 2015, turned this process into a standardized, published protocol that combines precise event definition, observational data analysis, climate model evaluation, and, when applicable, a social vulnerability analysis [3]. Thanks to that protocol, what used to take months or years to analyze is now resolved in days. More than 750 studies of this kind have already been published, and around 80% of them find a clear climate-change fingerprint in the event analyzed.
Europe, year after year
Europe is the most-studied continent. The summer of 2022, the hottest on record for the continent up to that point, caused 61,672 heat-attributable deaths, with Italy, Spain, and Germany leading [4]. In 2023, with an objectively milder summer, the figure dropped to 47,690, the second-worst of the last decade [5]. In 2024 it broke the record again, at 62,775 [6]. And in the summer of 2025, an analysis by Imperial College London and the London School of Hygiene and Tropical Medicine estimated that climate change had tripled expected mortality in twelve European cities: of the 2,300 deaths calculated for that period, about 1,500 were directly attributable to warming caused by fossil fuel burning [7]. A later analysis by the same teams, expanded to 854 European cities and the entire summer of 2025, calculated 24,400 total heat deaths, of which 68%, around 16,500, were attributable to climate change (a different methodology than the earlier ones was used, hence the difference in figures) [8].
A study published this very week in Nature Health calculates that a 2022-level heat mortality event is, in southern Europe, 26.5 times more likely today [9]. And the June 2026 episode isn’t the only one: it arrived weeks after another heatwave that had already broken May’s records, and both shared the same mechanism — a subtropical air mass coming from North Africa [1]. Across much of Western Europe, June is warming faster than any other month of the year, and daily maximum temperatures are rising at triple the rate of average global warming, while nighttime minimums are rising at double the rate. That means nights that no longer cool down enough for the body to recover from the heat accumulated during the day, a factor that weighs heavily on mortality [1].
This thermal stress is already forcing a redesign of infrastructure built for a different climate. In the UK, following the record 40.3°C recorded in July 2022, rail operator Network Rail rolled out the practice of painting stretches of track white to lower their temperature by 5 to 10 degrees and prevent the steel from warping through expansion — something that, in a single two-day heat episode, had caused thousands of train cancellations [10].
How much of this is avoidable? If Europe hadn’t implemented heat-prevention plans after the 2003 disaster, 2023 mortality would have been 80% higher than it actually was, and among people over 80, more than double [5]. Adaptation measurably reduces deaths. That’s why this year’s mortality figures appear lower.
Climate change in Spain
The Mediterranean has for years been classified as a particularly sensitive hotspot for global warming, with projections also showing a sustained decline in rainfall, with southeastern Spain being one of the most affected areas [11]. In July 2026, seas across the planet recorded the highest surface temperature ever measured, and the western Mediterranean reached up to 6°C above normal in June of that year.
A team of researchers calculated that this warming, directly attributable to human activity, is around 2°C in the Mediterranean basins, far above the 1.4°C of average global warming [12]. That warmer sea doesn’t stay in the water: it feeds humidity and warmer nights for coastal populations, has been linked to torrential rain events like the 2024 Valencia flood, and is causing die-offs of coral, sponges, and other marine species that can’t tolerate these sustained temperatures [12]. The Mediterranean acts as an amplifier that worsens climate change’s consequences along the eastern coast.
Could Spain turn into a tropical climate? Unlikely. That would require every month of the year to average above 18°C, something not even the most pessimistic scenarios place on the horizon for Cádiz or Almería this century. What the data does show is an expansion of arid climates: the Hadley cell, the atmospheric circulation that carries warm air from the equator toward subtropical latitudes, where it descends as dry air and generates the planet’s great deserts. As global temperature rises, that belt of dry subsidence has shifted poleward between one and three degrees of latitude per hemisphere since 1979, at a rate of between 0.5 and 1 degree per decade [13]. That shift is what explains why southern Spain is getting increasingly dry, hot summers. The most likely destiny for southern Spain isn’t a tropical rainforest, but a drier, more extreme version of the Mediterranean climate it already has.
The rest of the world
In late June 2026, the northwest Pacific coast of the United States and Canada experienced an extreme heat episode that pushed thermometers to nearly 50°C in locations that had never before exceeded 40°C, followed by a wildfire that razed a large part of one of the affected towns. A study involving around twenty researchers concluded that this heat would have been practically impossible without human-induced climate change [15]. During those days, a marine biologist from the University of British Columbia documented, through field sampling across various coastal beaches, the heat deaths of more than a billion mussels, clams, sea stars, and other intertidal animals, when coastal rocks reached temperatures above 50°C during low tide [16].
In June 2023, Beijing exceeded 40°C for the first time for three consecutive days, an event with an estimated return period of 111 years under the current climate. The intensity of the heat carries a clear human fingerprint, but the exact magnitude of how much more likely the event has become remains within the margin of uncertainty.
In the depths of the 2024 austral winter, with months of darkness and temperatures that normally hover around -30°C, the interior of the continent experienced the most intense heatwave recorded in 46 years of satellite observation, with anomalies of up to 28°C above normal for more than two straight weeks. An international team showed that human-caused warming had more than doubled the probability of that episode and added 0.7°C to its intensity [19].
The physical limit of the human body
Wet-bulb temperature combines heat and humidity into a single value and measures the air’s actual capacity to allow sweat to cool the body. In 2010, climatologists Steven Sherwood and Matthew Huber proposed, on thermodynamic grounds, that a 35°C wet-bulb temperature marks the upper limit of sustained human survival: beyond that point, not even a healthy person, in the shade and with unlimited water, can avoid hyperthermia [20]. That threshold has already been briefly reached at some subtropical coastal locations, and episodes of extreme humid heat have more than doubled in frequency since 1979 [21].
Heat doesn’t usually kill through direct heatstroke, but by aggravating preexisting cardiovascular and respiratory conditions, which explains why most deaths occur among older people with prior health conditions. A study using data from 43 countries, published in Nature Climate Change, calculated that 37% of all heat-related deaths between 1991 and 2018 were already directly attributable to human-caused warming, with peaks of up to 76% in South American cities like Santiago, Chile [22]. Globally, cold still kills far more people than heat in absolute numbers, at a ratio of roughly nine to one according to the same type of studies [23], but that ratio is rapidly reversing, since heat deaths are rising steadily while cold deaths decline as average temperatures climb.
What we can do, and what’s already being done
2003 remains the reference case: that summer’s disaster led to the design of heat-prevention plans that, according to the most recent analysis, prevented 80% more deaths in 2023 than actually occurred [5].
In July 2026, the Spanish government presented the Climate Shelter Network, more than 180 buildings belonging to the General State Administration open to any citizen every year from May 15 to September 30, as part of the State Pact against the Climate Emergency. The national map maintained by the Ministry for Ecological Transition already identifies 1,100 shelters across the country, combining public and private initiatives [30]. Barcelona remains the municipal benchmark, with more than 500 spaces spread across its ten districts and the goal of having one within five minutes of anyone by 2030 [26]. Other smaller cities are building their own networks: Jerez, after recording 45.8°C in August 2025, launched 49 shelters in 2026 [31], and this summer Ourense is unveiling Galicia’s first urban climate shelter, a €340,000 green corridor funded by the Xunta [32]. The private sector is starting to join these networks rather than staying on the sidelines: Aigües de Barcelona converted its customer-service offices into micro-shelters, and the startup Jungle Roofs, still at a very early stage with no confirmed clients, proposes converting 30% of the city’s flat rooftops into green roofs [33].
Another approach is to act directly on the most vulnerable buildings, such as schools. The European LIFE myBUILDINGisGREEN project, coordinated by the Royal Botanical Garden (CSIC), installed green roofs and facades at a school in Solana de los Barros, in Extremadura, and at two Portuguese schools: the building’s surface temperature dropped 20°C during heat peaks, and inside the classrooms — which used to exceed 32°C in June — it now hovers around 27°C [28]. Along the same lines, Aragón is renaturalizing schoolyards in Zaragoza as part of the European CARDIMED project [34], and Castellón is doing the same in 18 schools with ERDF funding [35]. Vegetation cools by intercepting solar radiation, preventing it from reaching surfaces like concrete and asphalt. These surfaces absorb heat and then release it into the air.
Seville remains a pioneer: its ProMeteo project, launched in 2022 with the Adrienne Arsht-Rockefeller Foundation Resilience Center, names and categorizes heatwaves by their health risk, the same way hurricanes are named [25]. The Community of Madrid activates a plan every year, from May 15 to September 15, with three alert levels coordinating schools, hospitals, transport, and care homes, with nearly €18 million invested solely in air conditioning for public schools in the 2025-2026 school year [29]. And at the Mediterranean scale, the European HEATSAFE project works with Cartagena, the metropolitan area of Barcelona, and cities in Greece, Bulgaria, Italy, and Albania to build vulnerability maps and shared local plans [36].
In India, the SEWA union, the Arsht-Rock organization, and the insurer Blue Marble created a parametric extreme-heat insurance policy in 2023 for women in the informal sector exposed to temperatures exceeding 49°C. The insurance pays out automatically as soon as sensors confirm a dangerous heat threshold has been crossed for two consecutive days, with no need to prove a loss. It started with 21,000 women in 2023 and now covers 50,000 across three Indian states [27].
At the individual level, what actually reduces risk is simple: check on elderly people who live alone, avoid physical exertion between noon and 5pm, and hydrate before feeling thirsty.
It’s reasonable, and plausible, to think that rapid attribution could arrive in the coming years — a report published days after an event and integrated directly into weather-warning systems.
Adaptation plans have proven they reduce mortality significantly, and that’s genuinely good news. But they have a physical limit: nobody can design a shelter against a wet-bulb temperature that sustainably exceeds 35°C, because beyond that point, no human behavior can compensate for thermodynamics.
REFERENCES
[1] World Weather Attribution (2026). “Fossil fuel emissions have rapidly worsened European heatwaves in just a few decades.” Published June 26, 2026. worldweatherattribution.org — Recent. Rapid analysis, not peer-reviewed as of this article’s writing.
[2] Stott, P.A., Stone, D.A., Allen, M.R. (2004). “Human contribution to the European heatwave of 2003.” Nature 432, 610–614. doi:10.1038/nature03089 — Reliable. Foundational study of the attribution field; compared real and counterfactual climate simulations to estimate the risk of exceeding a summer temperature threshold.
[3] Philip, S., Kew, S., van Oldenborgh, G.J., Otto, F., Vautard, R., van der Wiel, K., King, A., Lott, F., Arrighi, J., Singh, R., van Aalst, M. (2020). “A protocol for probabilistic extreme event attribution analyses.” Advances in Statistical Climatology, Meteorology and Oceanography 6(2), 177–203. doi:10.5194/ascmo-6-177-2020 — Reliable. Peer-reviewed protocol standardizing the rapid attribution process.
[4] Ballester, J. et al. (2023). “Heat-related mortality in Europe during the summer of 2022.” Nature Medicine 29, 1857–1866. doi:10.1038/s41591-023-02419-z — Reliable.
[5] Gallo, E. et al. (2024). “Heat-related mortality in Europe during 2023 and the role of adaptation in protecting health.” Nature Medicine 30, 3101–3105. doi:10.1038/s41591-024-03186-1 — Reliable.
[6] Janoš, T., Quijal-Zamorano, M., Shartova, N., Gallo, E., Méndez Turrubiates, R.F., Beltrán Barrón, N.D., Peyrusse, F., Ballester, J. (2025). “Heat-related mortality in Europe during 2024 and health emergency forecasting to reduce preventable deaths.” Nature Medicine. doi:10.1038/s41591-025-03954-7 — Reliable.
[7] Clarke, B. et al. (2025). “Climate change tripled heat-related deaths in early summer European heatwave.” Grantham Institute, Imperial College London / London School of Hygiene & Tropical Medicine, July 2025. — Recent. Rapid study not published in a peer-reviewed journal at the time of release, as confirmed by the Science Media Centre.
[8] Grantham Institute, Imperial College London / London School of Hygiene & Tropical Medicine (2025). “Summer heat deaths in 854 European cities more than tripled due to climate change.” September 2025. — Recent. Same methodological status as [7].
[9] Achebak, H. et al. (2026). “Extreme event attribution for heat-related mortality due to anthropogenic climate change across Europe.” Nature Health. doi:10.1038/s44360-026-00193-z — Recent. Published days before this article was written, no independent replication yet.
[10] Network Rail (2022-2025). Track-painting operational practices following the UK’s July 2022 record of 40.3°C; statements and rail-buckling incident data collected by Network Rail and British media. — Reliable. Widely corroborated institutional and journalistic source, a verifiable operational fact, not a scientific study.
[11] Cos, J., Doblas-Reyes, F., Jury, M., Marcos, R., Bretonnière, P.-A., Samsó, M. (2022). “The Mediterranean climate change hotspot in the CMIP5 and CMIP6 projections.” Earth System Dynamics 13, 321–340. doi:10.5194/esd-13-321-2022 — Reliable.
[12] World Weather Attribution (2026). “Climate change is driving unprecedented European ocean temperatures, with severe impacts for marine life.” Published August 19, 2026. doi:10.25560/132103 — Recent.
[13] Lucas, C., Timbal, B., Nguyen, H. (2014). “The expanding tropics: a critical assessment of the observational and modeling studies.” WIREs Climate Change 5(1), 89–112. doi:10.1002/wcc.251 — Reliable. Systematic review of the literature on the expansion of the dry subtropical belt since 1979.
[14] Agencia Estatal de Meteorología, AEMET (2025). CLIVAR-Spain 2024 Report, presented 03/20/2025. aemet.es/es/noticias/2025/03/informe_clivar_24 — Reliable. Official institutional source.
[15] Philip, S.Y., Kew, S.F., van Oldenborgh, G.J. et al. (2022). “Rapid attribution analysis of the extraordinary heat wave on the Pacific coast of the US and Canada in June 2021.” Earth System Dynamics 13(4), 1689–1713. doi:10.5194/esd-13-1689-2022 — Reliable.
[16] Harley, C. (University of British Columbia). Field estimate following the June 2021 heatwave, reported by CBC News and other outlets starting July 2021. — With reservations. A researcher’s estimate based on field sampling replicated across several beaches, widely cited, but without a peer-reviewed publication quantifying the total figure with the same rigor as a formal study.
[17] Zachariah, M. et al. (2023). “Attribution of 2022 early-spring heatwave in India and Pakistan to climate change: lessons in assessing vulnerability and preparedness in reducing impacts.” Environmental Research: Climate. doi:10.1088/2752-5295/acf4b6 — Reliable.
[18] Qian, C., Ye, Y., Jiang, J. et al. (2024). “Rapid attribution of the record-breaking heatwave event in North China in June 2023 and future risks.” Environmental Research Letters 19, 014028. doi:10.1088/1748-9326/ad0dd9 — Reliable. The calculated probability ratio did not reach statistical significance, though the attributable increase in intensity was robust across both methods used.
[19] Tang, H. et al. (2026). “Unprecedented 2024 East Antarctic winter heatwave driven by polar vortex weakening and amplified by anthropogenic warming.” npj Climate and Atmospheric Science 9(1), 122. doi:10.1038/s41612-026-01392-x — Reliable.
[20] Sherwood, S.C., Huber, M. (2010). “An adaptability limit to climate change due to heat stress.” Proceedings of the National Academy of Sciences 107(21), 9552–9555. doi:10.1073/pnas.0913352107 — Theoretical framework. Thermodynamic model establishing the 35°C wet-bulb physiological limit.
[21] Raymond, C., Matthews, T., Horton, R.M. (2020). “The emergence of heat and humidity too severe for human tolerance.” Science Advances 6(19), eaaw1838. doi:10.1126/sciadv.aaw1838 — Reliable.
[22] Vicedo-Cabrera, A.M. et al. (2021). “The burden of heat-related mortality attributable to recent human-induced climate change.” Nature Climate Change 11, 492–500. doi:10.1038/s41558-021-01058-x — Reliable.
[23] Zhao, Q., Guo, Y., Ye, T. et al. (2021). “Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: a three-stage modelling study.” The Lancet Planetary Health 5, e415–e425. doi:10.1016/S2542-5196(21)00081-4 — Reliable.
[24] Oberlandesgericht Hamm (2025). Ruling of May 28, 2025, Saúl Luciano Lliuya v. RWE AG case, file I-5 U 15/17. — Historical source. Judicial ruling documented by the court itself and by multiple specialized legal sources; not a scientific study.
[25] Seville City Council / Adrienne Arsht-Rockefeller Foundation Resilience Center (2022-2024). ProMeteo Sevilla Project. — Reliable. Evaluated in a two-year ex-post study (Zenodo, 2024).
[26] Barcelona City Council (2026). Xarxa de Refugis Climàtics. barcelona.cat/barcelona-pel-clima — Reliable. Official institutional source.
[27] Blue Marble, Arsht-Rock, SEWA (2023-2025). Extreme Heat Income Micro-Insurance, Gujarat, India. — Reliable. Jointly documented by the insurer, the NGO, and the union; figures updated by the World Economic Forum in 2025.
[28] Royal Botanical Garden-CSIC / EU LIFE program (2018-2024). LIFE myBUILDINGisGREEN Project. — Reliable. Case study published on AdapteCCa, the official platform of the Ministry for Ecological Transition.
[29] Community of Madrid (2026). High Temperature Episodes Action Plan 2026. — Reliable. Official institutional source.
[30] Government of Spain / MITECO (2026). Climate Shelter Network, presented July 27-28, 2026, as part of the State Pact against the Climate Emergency. — Reliable. Official institutional source (La Moncloa, MITECO).
[31] Jerez de la Frontera City Council (2026). Municipal Climate Shelter Network. jerez.es/refugiosclimaticos — Reliable. Official institutional source, corroborated by multiple local media.
[32] Xunta de Galicia / Ourense City Council (2026). Galicia’s first urban climate shelter, Avenida Otero Pedraio. — Reliable. Institutional agreement documented by the Xunta de Galicia.
[33] Jungle Roofs (2024-2026). Green-roof startup in Barcelona. — With reservations. A real, verifiable company, but at a very early stage: no confirmed clients as of the latest available information, self-funded with €15,000.
[34] Government of Aragón / CARDIMED project, Horizon Europe programme (2026). Renaturalization of schoolyards in Zaragoza. — Reliable. Public tender documented by the Government of Aragón.
[35] Castellón City Council / SEO-BirdLife / Biodiversity Foundation-MITECO, co-funded by ERDF (2026). “Castellón Naturaleza en Red” Project. — Reliable. Institutional source and formal confirmation of the 18 schools involved.
[36] Federation of Municipalities of the Region of Murcia / Interreg Euro-MED (2025-2026). HEATSAFE Project. — Reliable. European project with partners from Spain, Greece, Bulgaria, Italy, and Albania, institutionally documented.