Coffee and climate change

Why this crop is exposed

Sourced

Arabica does best within a narrow band of temperature, and a tree stays in the ground for many years, so a farm cannot follow the climate from season to season. A review of coffee physiology cites fruit ripening faster, often with a loss of quality, above about 23 °C, and flowers aborting in heat at blossoming. The crop also depends on the timing of rain: a dry spell followed by rain is what sets off flowering.

Those are the mechanisms through which a change in climate reaches the plant. They are established. How large the effects will be on farms is a different question.

Modelled projections: where the climate would suit arabica

Sourced

Each of these is a model of climatic suitability. None predicts what farmers will grow.

A global study of the 2050s

A 2015 study took 21 climate models from the fourth international assessment and a single high-emissions scenario, and projected the climate of the 2050s onto a map of where arabica grows now. Its summary reports less suitable land at lower altitudes in Mesoamerica, possible gains at higher elevations in South America near the equator and losses at higher latitudes, and coffee regions in India and Vietnam becoming less suitable. On East Africa the summary says Ethiopia and Kenya become more suitable; the body of the paper says the areas growing arabica there now would change little, with possible gains higher up, and calls East Africa, Uganda apart, among the regions least affected and Mesoamerica the most. Globally it projected losses at low altitudes and high latitudes. The authors attach a proviso: gains at altitude count only if the land is available, accessible, has suitable soil, and its people want to grow coffee.

A second study by overlapping authors

A study published the same year used 19 models from the following assessment and one intermediate emissions pathway, for the years 2040 to 2069. It projects the climates that now suit arabica best moving uphill by about 500 metres. It points out the catch: higher land is often hard to reach, too steep, thin-soiled, forested or of high conservation value, and because mountains narrow as they rise there is less of it.

Wild arabica

A 2012 study modelled where the wild plant could grow in and around Ethiopia. Its results are described in the Ethiopia entry of the atlas. It concerns wild populations, not farms, and used a climate model and scenarios that have since been superseded.

What a suitability model does not tell you

CoffeeHQ explanation

A suitability model takes the climate where coffee grows now, works out where that climate will be later, and reports the difference. It says nothing about soil, water for irrigation, who owns the land, what else could be grown there, or what a farmer with shade trees, a different variety or a different species might manage. It also treats a species as one thing, when varieties differ.

So a projection that an area becomes less suitable is not a forecast that coffee will stop being grown there, and a projection that higher land becomes suitable is not a forecast that coffee will move to it. The models are best read as a map of where pressure will rise.

Observed trends and case studies

Sourced

Two systematic reviews summarise what has been published. One, read here only in its abstract, reports that the effects found in the literature are mostly negative: lower yields, loss of the areas best suited to coffee, and the spread of pests and diseases. The other counted 148 studies and found 90 of them from the Americas, with comparatively little on robusta; it says there is a general lack of large-scale research on the flowering and growth of robusta and that the range of temperatures it does best in needs establishing more precisely.

A documented case is the leaf rust epidemics in Colombia and Central America between 2008 and 2013. The study of them found that both coincided with a narrower gap between day and night temperatures, and its authors describe the epidemics as a warning. They also found that every such epidemic in 37 years coincided with low profitability. That is a case study of weather and economics acting together, not a measurement of climate change.

For the berry borer, a 2011 study says that increased damage and an expanded range had already been reported in East Africa as temperatures rose, and gives an example: on Kilimanjaro the insect was being found 300 metres higher than ten years earlier. A later review of the pest cites work concluding that, before 1984, minimum temperatures around Jimma in Ethiopia would have been too low for it to complete a generation a year. The 2011 study's own model projects the number of generations a year rising across East Africa's arabica areas by 2050. The first is reported observation, at second hand; the second is an inference from temperature records; the third is a projection from one climate model under two emissions scenarios, and the study calls its generation counts hypothetical.

What is proposed, and how well it is supported

Sourced

Shade trees

A review of shaded coffee reports lower air temperatures and frost protection under shade, and also that heavy shade reduces flowering and can reduce yield. It asks whether shade is a strategy for a warmer climate and does not answer with a plain yes: effects depend on site, altitude and shade species. One model it cites suggests shade becoming beneficial at high elevations in future conditions; that is a model result.

Moving uphill

The models project suitable climate at higher elevations. The same authors note that the land there is often unavailable, and moving a perennial crop means years without income.

Irrigation

It is the direct answer to drought where water exists. The atlas entry for Vietnam describes what a crop that depends on irrigation does to the water it draws on.

Breeding and new varieties

Resistant and more vigorous varieties are described on the variety pages. What a variety's catalogue entry records is its performance in trials, not its performance in a future climate.

Other species

A 2021 study reported that stenophylla, a wild West African coffee, was judged similar in flavour to high-quality arabica. A 2025 paper by the same group, read in its introduction, restates the finding and the climate behind it: wild stenophylla grows at a mean annual temperature 6.2 to 6.8 °C higher than arabica, in lowland forest at about 400 metres. It calls the species undomesticated, and neither paper as read says anything about yield or whether it can be farmed commercially. Robusta already tolerates more heat than arabica and less cold.

Five labels for a climate statement

CoffeeHQ explanation

Every statement about coffee and climate is one of five kinds. OBSERVED: recorded as having happened. MODELLED: computed from data about the present, such as where a crop or a disease is found. PROJECTED: a model run forward under an assumed future. EXPERIMENTAL: produced by changing one thing under controlled conditions. INDUSTRY RESPONSE: something being done or promoted, which is evidence of concern and not of effect. The labels are CoffeeHQ's; the statements are the studies' own.

A statement does not become stronger by being repeated under a different label. A projection that a region will lose its coffee is not an observation that it has, and a company planting a new species is not evidence that the species will yield there.

Adaptation pathways: a model of the choices

CoffeeHQ explanation

A ladder of responses, from the least disruptive to the most, assembled by CoffeeHQ from the studies read. Each rung says what kind of evidence stands behind it. It is not advice, and nothing here says which rung any farm should be on.

Quality, not only quantity

Sourced

A systematic review of 73 studies looked at quality instead of yield. Its most consistent findings were that higher altitude is associated with better sensory scores and more light exposure with worse ones. It reports that the literature shows coffee quality to be sensitive to water stress, temperature and carbon dioxide too, but that the direction of those effects is not established and results differed between studies.

That leaves an honest gap. It is reasonable to expect a warmer farm to ripen its fruit faster, and the physiology review links faster ripening to lower quality; nobody has shown how much the cup changes on a given farm for a given rise in temperature.

No date and no percentage

CoffeeHQ explanation

It gives no figure for how much coffee land will be lost and no date by which anything will happen. A widely repeated figure of about half the suitable land by 2050 belongs to a study that was not read. It says nothing about prices, about what any company or certification scheme is doing, or about frost and drought in a particular country; those are in the atlas entries where a source describes them.

What to do next

• Where coffee grows, and why — The temperature and rainfall the projections are measured against.
• Ethiopia in the atlas — The modelled outlook for wild arabica, with its scenarios and limits.
• Shade, spacing and water on a coffee farm — What shade is reported to do, and what it costs.
• Coffee pests and diseases — The rust epidemics and the berry borer in full.
• Coffee sustainability claims — How to read a climate claim printed on a bag.