SSP1-2.6
low-emission
17 / 100
FAO · WMO joint report · 2026 · An interactive reading
Heat is rewriting
what farmers grow
and where they grow it.
For 175 years the planet was a stable place to farm. In a single human lifetime, it stopped being one. This is an interactive reading of the 2026 FAO–WMO joint report on how extreme heat is reshaping crops, livestock, fisheries, forests — and the livelihoods of 1.23 billion people.
Global mean temperature has climbed +1.48 °C above pre-industrial.
Annual global surface temperature anomaly relative to the 1850–1900 baseline. Shaded band: spread across the eight major datasets WMO uses for its annual benchmark.
+1.55 °C2024 PEAK
The past eleven years (2015–2025) are the eleven warmest in the 176-year observational record.
SOURCE · WMO 2025 State of the Climate, multi-dataset mean
↓ SCROLL TO READ
CH 01·definitions
What the report calls “extreme”.
Extreme heat is contextual: it's the level of heat that pushes a reference organism — a wheat plant, a cow, a worker — past moderate physiological stress. The benchmarks are biological, the trends are statistical, and they all point one way.
The report uses precise terms: a heatwaveProlonged abnormal heat, day and night, lasting days to months. is a temporal extreme; a marine heatwaveWater above the 90th-percentile temperature for at least five days — on the surface or at depth. is its underwater counterpart. A flash droughtDrought ignited and accelerated by heat — days, not seasons. is what happens when heat parches the soil before a season is over. When two or more of these arrive together, the report calls it a compound eventTwo or more hazards coinciding, with damage greater than their sum.. And the biological response — from acclimation through metabolic failure to death — is heat stressThe biological response of an organism to high temperatures, ranging from acclimation through metabolic failure to death.. Hover any of those terms for the precise definition.
Concurrent heatwaves now cover 1.6× more ground.
Spatial extent of simultaneous Northern-Hemisphere mid-latitude heatwaves, May–Sept. Indexed to 1979 = 100.
+58 %1979 → 2019
SOURCE · Rogers et al. 2022, in FAO–WMO 2026
And they happen six times as often.
Frequency of concurrent regional heatwaves, indexed to 1979 = 100.
× 61979 → 2019
SOURCE · Rogers et al. 2022, in FAO–WMO 2026
CH 02·physical drivers
A heatwave is built from three things at once.
Long-term greenhouse forcing raises the baseline. Interannual ocean–atmosphere cycles modulate where heat lands. Short-term atmospheric mechanics decide whether that heat becomes a recorded event. Add the three together and you get the observed heat record on the bottom strip.
Three ingredients, added on the same time axis, give the heat record we actually live with.
Time runs left to right, 1850 → 2025. The top three strips are the contributors; the bottom strip is their sum.
SOURCE · Authors' own elaboration; FAO–WMO 2026 §2.1, Fig 1
Human-induced warming reached +1.24 °C in 2015–2024 — accelerating at +0.27 °C per decade, unprecedented in the instrumental record.
CH 03·oceans
Nine-tenths of the planet's extra heat ended up in the ocean.
Of all the energy trapped by greenhouse gases since 1960, only a sliver warmed the atmosphere. The rest went into seawater — and it's still going there.
90 % of the climate system's added heat is stored in the upper 2 km of ocean.
Share of excess Earth-system energy retained since 1960, by reservoir.
SOURCE · von Schuckmann et al. 2023, in FAO–WMO 2026 §2.3
The ocean's heatwave days have nearly five-folded since 1980.
Days per year the sea surface exceeded the 90th-percentile threshold for at least five consecutive days, global mean.
× 51980 → 2024
Every additional 1 °C of sea-surface warming adds +3.7 marine heatwave events per year, lasting 7.5 days longer.
SOURCE · Cheng et al. 2023, in FAO–WMO 2026 §2.3
CH 04·crops
Every staple crop has a temperature it can't survive past.
Yields of the four grains that supply 60 % of global calories — maize, wheat, rice, soy — climb gently to an optimum near 30 °C, then drop off a cliff. Cool-season crops turn over earlier. Every year the average growing season sits closer to that cliff.
Each major staple has its own cliff temperature.
Yield response to mean growing-season temperature (% of optimum). Tap a crop above to swap the focal curve.
The dotted guide marks the focal crop's reproductive heat-damage threshold; gray curves show the others.
SOURCE · FAO–WMO 2026 §3.2.1; Lobell et al. 2011; Schlenker & Roberts 2009
Across the six staples the report tracks, sensitivity sits in a 3-point band — 5 to 8 % per +1 °C.
Observed yield decline per 1 °C of growing-season warming. Each row is one crop; the shaded vertical band is the 5–8 % range.
- PotatoC3 tuber−8.0 %
- MaizeC4 cereal−7.5 %
- BarleyC3 cereal−7.0 %
- SoybeanC3 legume−6.5 %
- WheatC3 cereal−6.0 %
- RiceC3 cereal−5.0 %
3 %5 %8 %10 %
SOURCE · FAO–WMO 2026 §3.2.2; Lobell et al. 2011
The world filled the gap by ploughing 88 million hectares of new land across 110 countries between 1992 and 2020 — emitting more carbon to grow the same food.
CH 05·livestock
Modern farm animals all live inside a 3 °C heat-stress window.
Cattle, goats and sheep share an upper thermoneutral threshold near 25 °C; chickens and pigs, which cannot sweat, top out around 24 °C. Modern high-yield genetics produce up to 20 % more metabolic heat than their predecessors — and are more sensitive to it.
Every major farm animal hits its heat-stress ceiling between 22 and 25 °C.
Upper thermoneutral threshold by species. Above the mark, the animal stops cooling itself and starts losing weight, milk and offspring.
- Dairy cowshigh-yield22 °C
- Pigscannot sweat24 °C
- Chickenscannot sweat24 °C
- Beef cattleunacclimated25 °C
- Goats25 °C
- Sheep25 °C
20 °C22 °C25 °C28 °C
The whole industry — dairy, beef, poultry, pork — operates inside a single 3 °C window.
SOURCE · FAO–WMO 2026 §3.3.1
In 2100, how many of every 100 cows will live in dangerous heat?
Each square is one percent of the world's cattle population, shaded if exposed to dangerous heat by 2100 under three IPCC pathways.
SSP2-4.5
middle of the road
31 / 100
SSP5-8.5
high-emission
48 / 100
A low-emission pathway leaves roughly two thirds of the herd in safe conditions; a high-emission one barely half.
SOURCE · FAO–WMO 2026 Executive Summary
Dairy yields fall ~2 % for each unit of THI heat stress; a single hour above WBGT 26 °C cuts daily milk 0.5 %, with effects persisting for ten days.
CH 06·fisheries
Five stocks have moved hundreds of kilometres poleward.
When water passes a stock's thermal threshold, the fish move, die, or starve while their prey vanishes. Across species the report tracks, more than 80 % show measurable poleward shifts — averaging 343 km, some moving more than 1 000 km from where fleets used to find them.
Five stocks, five poleward migrations.
Centre-of-distribution shift from the historical baseline (open dot) to recent observations (red dot). Arrowheads show the direction of travel.
343 kmMEAN POLEWARD SHIFT
More than 80 % of marine species the report tracks now show measurable poleward shifts.
SOURCE · FAO–WMO 2026 §3.4.2 (compiled from Pinsky et al. 2020 and references)
BOX 4 · BERING SEA SNOW-CRAB COLLAPSE
After the 2018 marine heatwave, more than 90 % of the population was gone.
Researchers calculate the transformation is roughly 200 times more likely under human-induced climate change. There is a 94 % chance the low-ice conditions of 2018 become the climatological norm by 2040.
SOURCE · LITZOW ET AL. 2024, IN FAO–WMO 2026
CH 07·forests
The world's forests have flipped from sink to source.
Wildfire-driven forest cover loss surged across every major forested region in the past two decades. Each lost hectare is carbon the forest used to hold; each new fire is carbon it now releases.
Six headline fires the report cites — every one of them since 2017, four since 2020.
Selected major wildfire events by year. The two paired years are the visual story: 2020 and 2023 each had two record-class events.
- 2017
- Portugal540 000 ha116 LIVES LOST
- 2018
- California850 000 haCAMP FIRE ET AL.
- 2020× 2
- SE Australia24.0 M haBLACK SUMMER
- Brazilian Pantanal4.5 M ha30 % OF WETLAND BURNED
- 2023× 2
- Canada18.5 M haLARGEST IN N. AMERICA
- Greece175 000 haLARGEST IN EU HISTORY
SOURCE · FAO–WMO 2026 §3.5; national fire agencies
Fire-driven forest loss rates climbed fastest in the Americas.
Change in annual fire-driven forest cover loss rate, vs. the prior period.
+270 %
North America
CHANGE IN ANNUAL LOSS RATE
+240 %
Latin America
CHANGE IN ANNUAL LOSS RATE
+140 %
Africa
CHANGE IN ANNUAL LOSS RATE
SOURCE · Potapov et al. 2025, in FAO–WMO 2026 §3.5.2
And global wildfire CO₂ emissions are up 60 % across the satellite era.
Indexed (2001 = 100). The trend is accelerating.
+60 %
Global wildfire CO₂ emissions
2001 → 2023 · INDEXED
SOURCE · Global Fire Emissions Database, in FAO–WMO 2026 §3.5.2
CH 08·agricultural workers
In one tropical belt, half the year is already too hot to work safely.
Agriculture is field work, and field work is open to the sky. Above a wet-bulb globe temperature of 32 °C, sustained outdoor labour becomes physiologically unsafe. The world is already losing hundreds of billions of working hours a year — concentrated in the regions least able to absorb the loss.
Global labour hours lost to extreme heat have climbed by half since 1990.
Billions of working hours lost per year, all sectors.
+47 %1990 → 2021
SOURCE · The Lancet Countdown 2022, in FAO–WMO 2026 §3.6.2
By mid-century, half the year is too hot to work safely across the tropical belt.
Projected days per year exceeding the WBGT stop-work threshold by 2050 under a high-emission scenario.
~ 250 d/yrPEAK BELT, BY 2050
- South AsiaIndia, Pakistan, Bangladesh~ 250 days/yr
- Tropical Sub-Saharan AfricaSahel and equatorial belt~ 240 days/yr
- Central AmericaCoastal Honduras, Guatemala~ 220 days/yr
- Coastal South AmericaColombia, Venezuela~ 200 days/yr
- MediterraneanS. Spain, Italy, N. Africa~ 90 days/yr
- Northern Europe / N. Americareference~ 30 days/yr
The most exposed regions are the same belt where most of the world's smallholder farmers live.
SOURCE · FAO–WMO 2026 §3.6.3
35×
Higher likelihood of dying from occupational heat exposure for an agricultural worker vs. any other sector.
1.9×
Increase in extreme-heat exposure for maize farmers between 1979–2000 and 2001–2019 (rice farmers: 1.8×).
CH 09·compound events
One hazard, many failures — the cascade behind the headlines.
The single most important pattern in the report is that hazards rarely arrive alone. Heat ignites flash droughts; droughts feed wildfires; wildfires destroy forage; heat-stressed workers can't bring in what's left; production losses translate into prices and incomes. Compound heat-drought events are already +200 % more frequent in parts of the world than in 1950.
The cascade behind a compound event.
Solid arrows: the dominant chain. Dashed arrows: feedback and bypass paths the report flags. Each node is anchored to a real case the report cites.
SOURCE · Adapted from FAO–WMO 2026 §1.2, §3.1, and the Brazil case study
CH 10·case study · Brazil, Aug 2023 → Dec 2024
One country, every mechanism in this report, at once.
Brazil's 2023–2024 compound event is the report's centrepiece — a real-world test of every chapter that came before it. A drought made +1 000 % more likely by climate change. A ~10 % cut to the soy harvest. Six different indicators all spiking through one growing season.
Daytime maximum ran 4 °C above climatology.
°C above 1991–2020.
+4.1 °CSEP 2024
SOURCE · FAO–WMO 2026 Fig 13
Two-thirds of cropland slipped into severe drought.
% under SPI severe-drought class.
68 %SEP 2024
SOURCE · FAO–WMO 2026 Fig 15
Soy crossed 30 °C on half of days.
% of days the daytime max exceeded 30 °C.
48 %OCT 2024
SOURCE · FAO–WMO 2026 Fig 16
Swine risk peaked at a third of the month.
% of days at lethal-stress THI for pigs.
36 %OCT 2024
SOURCE · FAO–WMO 2026 Fig 18
Fire-weather days tripled in the dry season.
% of days with FWI > 30.
66 %OCT 2024
SOURCE · FAO–WMO 2026 Fig 20
Outdoor work was unsafe on most days.
% of days with WBGT > 26 °C.
54 %OCT 2024
SOURCE · FAO–WMO 2026 Fig 23
Heat built first, drought intensified in its wake, soy passed its damage threshold, swine risk climbed, fire weather peaked, outdoor labour became unsafe — all inside one growing season.
CH 11·adaptation
Three layered lines of defense.
The report makes a structural argument about adaptation: it has to scale faster than the damages do, and it has a ceiling. Below, the three layers in plain language, with the examples the report repeatedly returns to.
I
Anticipate.
Heat is more predictable than most climate hazards. The first line of defense is forecast-based action: agrometeorological advisories tied to early-warning systems that trigger preventive measures before a heatwave lands.
- Anticipatory cash transfers to vulnerable farming households.
- Pre-positioning irrigation, shade structures and feed reserves.
- WBGT-triggered stop-work rules for outdoor labour.
II
Redesign.
When the climate window for a crop or breed has shifted, no amount of forecasting brings it back. The second line is structural — heat-tolerant varieties, indigenous and composite livestock genetics, mixed-species silviculture, climate-aware aquaculture siting. The horizon is decades.
- Reproductive-heat-tolerant maize, wheat and rice lines.
- Cross-breeding bos taurus with bos indicus heat tolerance.
- Marine spatial planning that buffers stock movement.
III
Mitigate — adaptation has a ceiling.
Every adaptive measure sits on top of one constraint. With the 1.5 °C carbon budget "virtually exhausted" — and with the most extreme outcomes confined to high-emission scenarios — mitigation is what determines whether the first two lines of defense have anywhere to stand.
- Emission reductions in agrifood systems themselves (≈ ⅓ of global emissions).
- Land-use planning that doesn’t push expansion into burned forests.
- International finance for adaptation in the most exposed regions.
CH 12·the choice
The next decade decides the gap between these two futures.
Every adaptive measure in the report sits on top of one constraint: the Global Carbon Budget reports the 1.5 °C budget is "virtually exhausted." Below, the headline numbers the IPCC and the report give for the two pathways the world is actively choosing between.
Same metric, two futures — the gap is the story.
Six headline numbers under the IPCC's low- and high-emission scenarios. The rightmost column reads each row's gap in plain language.
| Metric | LOW-EMISSIONSSP1-2.6 | HIGH-EMISSIONSSP5-8.5 | THE GAP |
|---|---|---|---|
| % of global cattle exposed to dangerous heat by 2100 | 17 % | 48 % | ≈ 3× more cattle in danger |
| Marine heatwave frequency by 2100, vs. 1850–1900 | × 20 | × 50 | 2.5× more events |
| Marine heatwave peak intensity by 2100 | unchanged | × 10 | an entire order of magnitude |
| Global outdoor labour capacity retained by 2100 | 81 % | < 40 % | roughly half is lost |
| Average lake-heatwave duration by 2100 | 27 days | 95 days | 3.5× longer |
| Days per year too hot for safe outdoor work, tropical band | ~ 100 | ~ 250 | 2.5× more lost work days |
SOURCE · FAO–WMO 2026 (executive summary + chapter aggregates)
The only durable solution to protect global agrifood systems from the escalating threat of extreme heat lies in ambitious, multilateral climate change mitigation.