Climate change and coffee

Arabica is a fussy plant grown in a narrow band of the tropics, and the band is moving. That is the whole story in one line. The detail below is about how much, where, and — just as important for a wiki — which of the frightening numbers in circulation actually have a paper behind them.

Two things make arabica unusually exposed. It wants cool nights and high ground, so warming pushes its comfortable zone uphill, and there is only so much uphill. And it has almost no natural defence against Coffee leaf rust, a disease that spreads better in warm, wet conditions. Its tougher cousin Robusta has neither problem — see Arabica vs Robusta for why the two species diverge on exactly the axes climate is moving.

What the research actually says

The most useful single document is a 2022 systematic review in the journal Plants, which is a study of studies: the authors gathered the published literature on coffee and climate and counted what it found. 148 records made the final synthesis.¹

Of the 42 papers that looked directly at climate effects on yield or production, 35 reported negative impacts, 4 were mixed, and only 3 found positive effects.¹ That lopsidedness is the honest headline. The literature is not split.

The review also quantifies coffee science’s own blind spot: 75 documents — 63% of the total — studied Coffea arabica, against just 9 on robusta.¹ Arabica gets the attention because it is judged “more sensitive to climate factors than robusta coffee and will be more affected by climate change.”¹

The single most striking number in it is a comparison of what future coffee land will suit:

At least 83% of the total future coffee-growing area meets the requirements for robusta cultivation, but only 17% (±6%) meets requirements for Arabica.¹

Read that carefully — it is a statement about where each species could grow, not about what the world will plant. But it is the clearest published version of the argument this wiki keeps running into on the Fine Robusta and Robusta pages: the species that specialty coffee has never taken seriously is the one the century suits.

The 2050 map: who loses, who holds

The reference study for country-level projections is Ovalle-Rivera and colleagues, published in PLoS ONE in 2015. It modelled arabica suitability for the 2050s under the IPCC SRES A2a scenario — a high-emissions storyline — averaged across 21 global climate models.²

Region / country	Projected change in arabica suitability, 2050s
Mesoamerica (regional average)	−0.29
Mexico	−29%
Honduras	−27%
India	−0.28
Brazil	−25%
Vietnam	−0.25
Kenya	−0.12
Ethiopia	−0.11
Rwanda	−0.09
Papua New Guinea	−0.09
Global average	−0.19, across 374,080 km² of current potential coffee area

The pattern is a roughly threefold gap between the worst-hit regions and the best-off ones. Brazil — the world’s largest producer — sits near the bad end at −25%, while East Africa holds up best: Ethiopia, Kenya and Rwanda cluster around −0.1.² Papua New Guinea, in the Pacific, does equally well.

The study’s own spread of outcomes matters as much as the averages. Under its pessimistic model runs, a third of the current coffee area loses more than 40% of its climate suitability. Under the median, 37% of current coffee area loses 20–40% and another 27% loses 10–20%. Even the optimistic case has 34% losing 10–40%, with 52% unchanged.² There is no model run in which nothing happens.

To verify — the units in that table are not consistent in the source

The figures above are reported by the paper in two different-looking forms: bare decimals (−0.19, −0.29) that read as changes in a suitability index, and percentages (−29%, −25%) that read as changes in suitable area. The summary consulted here does not state whether they are the same measurement expressed two ways. Treat the ranking as reliable and the exact quantities as approximate, and read the paper itself before quoting a single number out of context.

To verify — "half the world's coffee land gone by 2050"

The most-repeated statistic about coffee and climate — that climate change could cut the global area suitable for coffee by about 50% — traces to Bunn and colleagues, “A bitter cup,” Climatic Change, 2015.³ That paper is paywalled and could not be read here; every attempt redirected into a publisher login loop. The research report behind this note also attributes both the 50% figure and the regional winners-and-losers findings to it, when the regional findings verifiably belong to the Ovalle-Rivera paper above.² Until the abstract is read at source, cite the Ovalle-Rivera numbers, not the 50%.

Ethiopia’s wild arabica, which is the backup copy

Ethiopia’s note already makes the point that the country’s forest arabica is the only genetic reserve the crop has. The study that put numbers on the risk is Davis and colleagues, PLoS ONE, 2012 — work out of Kew — modelling indigenous arabica, not farms.⁴

Projecting to 2080, it found:

• Most favourable scenario: a 65% reduction in the number of bioclimatically suitable localities. “Bioclimatically suitable” means the climate at that spot still fits what wild arabica needs.⁴
• Worst-case scenarios: “almost 100% reduction” — only one surviving locality per scenario.⁴
• By area rather than locality count: a 38% reduction at best, and around 90% at worst.⁴

Two specific places are named. Populations on the outer edges of the southwest Ethiopian range, several in the Bale Mountains, were projected to be in unsuitable climate space “even by 2020.” And all populations on the Boma Plateau in South Sudan are projected unsuitable by 2080.⁴

To verify — a 2012 model's 2020 prediction has never been checked here

The Bale Mountains finding is now a projection about a year in the past. No source consulted for this note reports whether those populations were actually observed to decline, and a model saying a place has become climatically unsuitable is not the same as fieldwork saying the plants died. Every figure in this section is a model output, not an observed loss.

Brazil, where irrigation changes the answer

A 2024 study in Science of the Total Environment looked at Brazil alone, under the modern SSP2-4.5 and SSP5-8.5 emissions scenarios (moderate and very high), across 2041–2060, 2061–2080 and 2081–2100.⁵

Its findings split the country rather than condemning it:

• In Aw climates — tropical with a dry winter, the hotter and drier end of Brazil’s coffee ground — losses “may reach 100%” under high-emission scenarios.⁵
• In the cooler C climates, including southern Minas Gerais, São Paulo and northern Paraná, yields are projected to rise, partly through CO₂ fertilisation — more atmospheric carbon dioxide making plants grow faster. These regions are “projected to have the highest yields,” though with more uncertainty attached.⁵
• Irrigation “will play an important role in mitigating yield losses, especially in northern regions.” With it, parts of northern Minas Gerais and Bahia could still exceed 30 bags per hectare.⁵
• The timing of the crop shifts: flowering comes earlier in colder regions and later in warmer ones, while ripening moves earlier everywhere.⁵

That last point is the one producers feel first. A harvest that arrives at a different time than the mill, the pickers and the buyer’s calendar expect is a logistics problem before it is an agronomy problem — and it lands on the same direct-trade relationships and shipping windows described in Green coffee buying.

Altitude buys time, not escape

The standard industry answer is “plant higher.” The literature supports the direction and not the comfort: impacts are consistently strongest at low altitudes and mildest at high ones, which is why the East African highlands hold up better than Mesoamerica in the 2050s modelling.²¹

But a mountain is finite. Moving a farm 300 metres uphill works exactly once, and the land above an existing farm is often forest, park or someone else’s. The pattern the models describe is not migration to safety — it is the suitable band getting narrower while the plant is pushed against the top of it.

To verify — "coffee is now grown higher than it used to be"

The claim that farms are routinely planting above elevations that were formerly normal appears in commercial coffee-education writing⁶⁷ but in none of the peer-reviewed sources read for this note, and no source gives before-and-after elevation figures for any named region. It is a plausible industry observation, not a documented trend. This wiki has hit the same wall on altitude numbers before — see the To-verify callout on species altitude bands in Arabica vs Robusta.

Pests, disease, and the second-order damage

Warming does not only move the map. The systematic review reports expected increases in the distribution of the coffee berry borer and in the severity of coffee rust.¹ The borer is a beetle that tunnels into the bean itself; rust is the leaf fungus documented at length in Coffee leaf rust.

That partly settles an open question already flagged on this wiki. The Coffee leaf rust note carries a To-verify callout saying the climate-change-to-rust link was “widely asserted in the trade press” but unconfirmed against a primary source. A peer-reviewed systematic review now supports the direction of that link.¹ What still has no source here are the specific temperature and leaf-wetness thresholds at which rust takes off — that part of the callout stands.

The review also gives the clearest published sensitivity figure for the other species: each 1°C rise above a 16.2°C minimum temperature cuts robusta production by 350–460 kg per hectare, or 14%.¹ Robusta is hardier than arabica; hardier is not immune.

What is being done about it

The organised response runs mostly through breeding. World Coffee Research exists for it, describing its own mission as driving “the development of high quality, climate-resilient coffee varieties,” and arguing on its homepage that “improved varieties are a powerful tool to address the climate crisis” — with better farm productivity capable of reducing greenhouse-gas emissions by up to 32%.⁸ Its Innovea breeding network and multi-location variety trials are described on the World Coffee Research page.

The genetic material for that work comes from two places, both already on this wiki. One is Robusta, via the Timor Hybrid and the Catimor and Sarchimor families it produced — the same borrowed resistance genes that answered rust. The other is Ethiopian forest arabica, which is precisely the population the Davis modelling says is at risk.⁴ The backup copy and the thing it backs up are exposed to the same weather.

At the other end of the chain, Carbon-negative coffee and the certification schemes around it — Rainforest Alliance, Sustainable Agriculture Network’s Climate Module, B Corporation certification — are the marketing-facing response. As those notes document, the verification behind most of these claims is thinner than the labels suggest.

To verify — what warming does to quality, as opposed to yield

The research report behind this note asserts that climate change is expected to reduce cup quality as well as yield and to raise pest pressure, citing a global review.² The quality half of that claim was not confirmed in any source read here — the papers verified report suitability, area, yield and phenology, not cupping scores. Given how much of this wiki turns on the 80-point line (see Coffee scoring), a documented link between temperature and score would matter a great deal; nothing in this research established one.

Why it matters here

Almost everything else on this wiki — Best of Panama records, Cup of Excellence lots, Geisha at altitude on Hacienda La Esmeralda and Elida Estate — depends on high-grown arabica remaining possible in the specific places that grow it. The competitions are, in effect, a price-discovery machine bolted onto a narrow climatic niche.

The projections above do not say that niche disappears. They say it shrinks, shifts uphill and moves east, that Mesoamerica takes the worst of it while East Africa holds, and that the species built for the resulting climate is the one specialty coffee has no auction for. Whether that is a crisis or a reshuffling depends on questions nothing in this research answers.

Related

• Arabica vs Robusta — the species comparison this note is the climate half of
• Robusta · Fine Robusta — the hardier species and the only standard that lets it be sold as specialty
• Coffee leaf rust — the disease that warming is expected to worsen
• World Coffee Research — the breeding programme built as the response
• Timor Hybrid · Catimor · Sarchimor — where borrowed climate resistance came from last time
• Ethiopia — the wild genetic reserve, and the population Davis modelled
• Brazil · Cerrado Mineiro · Colombia · Kenya — origins with country-level projections above
• Carbon-negative coffee · Rainforest Alliance · Sustainable Agriculture Network · B Corporation certification — the claims and certifications on the consumer side
• Green coffee buying · Direct trade · C-market — the trade the harvest-timing shifts land on
• Liberica — the third species, floated as a heat-tolerant option on evidence this wiki could not verify
• Specialty coffee · Coffee scoring — the quality framework that assumes high-grown arabica

Official links

A scientific topic has no homepage. These are the industry-facing bodies whose own published work covers it:

• Official site — World Coffee Research — the coffee industry’s climate-resilience breeding programme
• Official site — World Coffee Research variety catalogue — per-variety climate and disease ratings

Footnotes

1. Plants (Basel) — Bilen, El Chami, Mereu, Trabucco, Marras & Spano, A Systematic Review on the Impacts of Climate Change on Coffee Agrosystems (2022) — confirmed via WebFetch; source of the 148 records, the 35-of-42 negative finding, the 63%-arabica research split, the 83%/17% future-area comparison and the robusta temperature sensitivity ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹

2. PLoS ONE — Ovalle-Rivera et al., Projected Shifts in Coffea arabica Suitability among Major Global Producing Regions Due to Climate Change (2015) — confirmed via WebFetch; source of the 2050s country projections, the SRES A2a / 21-GCM method and the pessimistic/median/optimistic spread ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

3. Climatic Change — Bunn et al., A bitter cup: climate change profile of global production of Arabica and Robusta coffee (2015) — paywalled; every fetch attempt redirected into a publisher login loop, so the abstract was not read ↩

4. PLoS ONE — Davis, Gole, Baena & Moat, The Impact of Climate Change on Indigenous Arabica Coffee (2012) — confirmed via WebFetch; source of the 65%/100% locality reductions, the 38–90% area range, and the Bale Mountains and Boma Plateau findings ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

5. Science of the Total Environment — de Freitas, Coelho, Costa & Sentelhas, A bitter cup of coffee? Assessing the impact of climate change on Arabica coffee production in Brazil (2024) — confirmed via WebFetch; source of the SSP scenarios, the Aw-climate 100% loss, the irrigation finding and the phenology shifts ↩ ↩² ↩³ ↩⁴ ↩⁵

6. Clive Coffee — How climate change is affecting coffee — retailer education blog; cited via the research report, not fetched ↩

7. CoffeeGeek — Coffee in crisis: climate change’s impact — coffee blog; cited via the research report, not fetched ↩

8. World Coffee Research — confirmed via WebFetch; source of the “climate-resilient coffee varieties” mission line and the up-to-32% emissions figure ↩