Thinking · Evidence

Europe counted 34,600 heat deaths within weeks: India cannot produce that number.

Abhijith Magal · 3 October 2026 · 8 min read

The heat deaths India data gap is easiest to see from Madrid. On 1 October, a Bloomberg compilation of official estimates put this summer's heat toll in five Western European countries above 34,600, as the region closed its hottest June-to-August on record, 2.5 degrees Celsius above the 1991-2020 average per Copernicus. Spain's health institute, the Instituto de Salud Carlos III, reported 5,805 excess deaths attributable to heat between 15 May and 30 September, its deadliest summer since its records began in 2015. France's public health agency put its toll at 7,824 across three heatwaves, the highest since at least 2015. Germany's estimate ran to about 16,000, the UK counted 2,877 for May and June alone, Belgium 2,112.

Hold the numbers a moment less than the dates. Spain's figure covers a season that ended on 30 September and was public on 1 October. The mechanism is the point: Carlos III compares each day's total deaths against the expected baseline, factors in temperature thresholds for each of the country's 50 provinces, and publishes the attributable excess continuously. France runs the same logic. Europe does not count heat deaths by reading cause-of-death lines on certificates, where heat rarely appears; it counts the dead who would not have died, statistically, in a cooler week.

The number India does not have

I went looking for the Indian equivalent, an in-season, excess-mortality-based heat death count for summer 2026, and could not find one. What exists instead is a familiar pair. Official tallies of certified heatstroke deaths, which arrive months late and count only the narrow clinical category, historically in the hundreds for seasons in which independent studies estimate far more. And research estimates like the Climate Impact Lab's projection of 15,800 additional deaths for this El Nino season, covered in the projection piece, which model what the official system will never record. Between the two sits no operational count.

The pattern should be familiar to readers of the dengue registers piece: India's surveillance and certification systems disagreed on dengue deaths by a factor of nearly seven, and the states that counted best looked worst. Heat is the same failure at larger scale, because heat kills almost entirely through other diagnoses, the cardiac arrest, the renal failure, the stroke, so a certificate-based count structurally misses it. Only an excess-mortality method sees it, and India does not run one in season.

The cost of not counting is allocative. A state that cannot see its heat mortality within the season cannot direct cooling, outreach or workplace protections to the districts where people are actually dying, cannot evaluate its heat action plan against anything, and will under-budget next year because the official toll stayed in three digits. Europe's post-2003 history runs the other way: counting 70,000 dead forced the warning systems and hot-weather services that now hold its tolls down.

The method is replicable, and the wait is optional

Nothing in Spain's machine needs inventing. Daily all-cause death counts, a baseline model, local temperature thresholds. India's Civil Registration System records deaths with known but measurable completeness gaps, several states register digitally with short lags, and IMD grids the temperatures. A state public health directorate, or a serious academic partnership with one, could run a Carlos III-style model for a pilot state within a season or two, and publish during the season rather than after it. That is the kind of public instrument a climate-health intelligence practice exists to push toward, and I intend to keep pushing.

A programme does not have to wait for the state instrument. Any cohort-based health programme can count its own ledger through a heat season: deaths in the enrolled cohort against the cohort's cooler-month baseline, and treatment days maintained through hot spells, the metric from the chronic-care piece. A few thousand patients counted properly beats a nation estimated annually, for the decision the programme actually faces.

One limit. The European figures come from a news compilation of agency estimates using differing methods and periods, the UK's covering May and June only, so the 34,600 is a floor built from unlike parts and the country figures should not be compared against each other too precisely. On the India side, absence of evidence was my own search; if a state is running an in-season excess-mortality heat count I have missed, I want to know about it, and I will correct this page.

Common questions

How many heat deaths did Western Europe record in summer 2026?

More than 34,600 across five countries, per a Bloomberg compilation of agency estimates published 1 October 2026: about 16,000 in Germany, 7,824 in France, 5,805 in Spain, 2,877 in the UK for May and June, and 2,112 in Belgium, in the region's hottest June-August on record.

How does Spain count heat deaths so quickly?

The Instituto de Salud Carlos III compares daily all-cause deaths against expected baselines and attributes the excess using temperature thresholds set for each of Spain's 50 provinces, publishing continuously through the season. The method counts statistical excess, not heat mentions on death certificates.

Why does India have no comparable heat death count?

India's official figures count certified heatstroke deaths, a narrow clinical category reported with long lags, and no agency runs an in-season excess-mortality model. Heat kills mostly through cardiac, renal and cerebrovascular events, so certificate-based counting structurally misses the toll.

Can the excess-mortality method work in India?

The ingredients exist: Civil Registration System death records, with measurable completeness gaps, digital registration in several states, and gridded IMD temperature data. A state-level pilot modelled on Carlos III is feasible, and individual health programmes can count their own cohort's mortality and treatment continuity through heat seasons without waiting for it.

Abhijith Magal

Founder, Syntropy Earth. Nine years across two global pharmaceutical multinationals in patient access and commercial roles, with health-equity work alongside the WHO-Foundation and UNICEF. He works on climate-access: where climate disruption breaks the link between patients and care.

Last updated: 3 October 2026