A tea bush does not tolerate much. Below about 13 to 14 degrees Celsius it stops putting out new shoots; above roughly 30 degrees, growth is restricted again, and the crop's whole economics sit inside that narrow band. A synthesis of 126 studies published in the journal Agronomy in 2021, led by researchers S.L. Jayasinghe and L. Kumar, found that pushing an average month above about 28 degrees Celsius in Assam cuts yield by roughly 3.8 percent for every additional degree. Drought does more damage faster: in Kenya's Rift Valley in 2009 it cut yields by about 30 percent, and in Sri Lanka in 1992 by about 26 percent. Warming and drought are not a future risk for the tea trade. They are already a line item.

The three numbers that matter

Three figures carry the weight of the whole picture, and each comes from a different kind of evidence.

The first is a temperature threshold, from that same Agronomy review: in Assam, once the average monthly temperature climbs past 26.6 degrees Celsius, further warming starts costing yield, at a rate near 3.8 percent per additional degree at the higher end of the range studied. That is a plant physiology finding, drawn from decades of agronomic data, not a model of the future.

The second is a drought shock, and it is already history rather than projection. Kenya's Agricultural and Livestock Research Organization, in a 2020 paper in Frontiers in Plant Science, traces a 58-year climate record at its own Tea Research Institute in Kericho: average temperature there has risen 0.016 degrees Celsius a year and rainfall has fallen 4.82 millimetres a year over that span. The same paper cites the 2009 Rift Valley drought's roughly 30 percent yield cut as the kind of acute loss that a drying long-term trend makes more frequent, not as a one-off.

The third is a suitability projection, and it is the most contested of the three because it depends on which emissions path the modellers assume. The Ethical Tea Partnership's figures, reported by Mongabay in 2024, put Kenya's loss of climatically suitable tea-growing area at 26.2 percent by 2050 to 2070, Sri Lanka's at 14 percent by 2050 (rising toward 30 percent by 2070), and China's at a comparatively modest 4.7 percent. KALRO's own modelling for Kenya, using a longer 2075 horizon, lands in the same range: roughly 22.5 percent of currently suitable area lost if temperatures rise as expected and growers do nothing to adapt.

Why a hotter Kenya is the sharpest case

Kenya is the clearest example because the numbers line up from three independent directions: farmer surveys, research-station records, and formal modelling, and because so much of the world's tea trade runs through it. The country supplies close to half the tea consumed in Britain and the crop employs around three million Kenyans, mostly as smallholders, according to Christian Aid's reporting via the World Economic Forum. A United Nations survey of 700 growers across Kenya's seven tea-growing regions, cited in the same report, found more than 40 percent had already noticed a shift in the timing of the rainy and dry seasons, and 35 percent named drought specifically as a concern. That is not a model's output. It is what the people who grow the crop are telling surveyors, in the present tense.

The mechanism KALRO's researchers describe is straightforward: as average temperatures climb, the elevation band where tea grows best gets pushed uphill, and land does not expand to follow it. Kenya's national research body models Kenya's average air temperature rising by roughly 11 percent by 2075 on an unmitigated path, with tea's currently suitable growing area shrinking correspondingly. Its response has been to breed and distribute drought-tolerant, higher-yielding cultivars through the Kenya Tea Development Agency, working with the Tea Research Foundation of Kenya, alongside irrigation investment, according to KTDA's own public reporting on its adaptation programme.

Assam and Sri Lanka: same physics, different shape of the loss

Assam's exposure runs through a different mechanism: a shrinking harvest window rather than a shrinking map. Because the region's most valuable leaf, the first and second flush, is picked early in a compressed season, any warming that shortens the cool window before the hot months cuts directly into the priciest part of the crop, even before whole-year yield is affected. The Jayasinghe and Kumar review's temperature-threshold finding for Assam, the 3.8 percent yield loss per degree above the 28-degree mark, sits on top of that timing problem rather than replacing it.

Sri Lanka's case is closer to Kenya's suitability story than to Assam's. The same review reports Sri Lankan yield already falling once mean temperatures pass roughly 25 to 26 degrees Celsius, and the Ethical Tea Partnership figures a further 14 percent loss of suitable growing area by 2050, widening toward 30 percent by 2070 on a medium-emissions path. Two different countries, two different physical pressures, drought and heat interacting with elevation in Kenya, a compressed picking calendar in Assam, and both converging on the same conclusion: less tea, or lower-value tea, from the same land, unless the growing zone itself moves.

The one place the figures point the other way

Not every producer loses. The Jayasinghe and Kumar review's suitability modelling shows China gaining rather than losing growing area, projected up 20 percent by mid-century on the pathways it surveyed, alongside smaller gains for Vietnam, Turkey, and, on some projections, India, whose optimal zone is modelled to expand by 15 percent by 2050 and 25 percent by 2070. The review also flags a genuinely two-sided effect within the crop itself: elevated atmospheric carbon dioxide raises soluble sugar, catechins, and polyphenols in the leaf while lowering caffeine, a change in the tea's chemistry, not just its volume, that the literature has not settled into a simple "better" or "worse."

That unevenness is itself a market fact worth sitting with plainly. A shock that shrinks Kenyan or Sri Lankan supply does not remove tea from the world; it shifts where the crop is grown and who captures the margin on it, and the shift favours producers already gaining land at higher latitudes over the smallholders on marginal, low-altitude plots who have the least capital to relocate or irrigate their way out of a bad decade.

What growers are doing about it, and what it costs

None of the adaptation on offer is free, which is why it is arriving unevenly. KALRO's cultivar-breeding programme and KTDA's irrigation investment target the roughly 600,000 smallholders the agency represents, but a drought-tolerant seedling and a drip line both cost money a grower whose income is already set by a flat auction price does not necessarily have spare. Mongabay's 2024 survey of the wider adaptation landscape names shade-tree agroforestry, integrated pest management, and alternative processing fuels (rice husks, sawdust, biochar) as lower-capital options researchers at Brunel University, Uganda's National Agricultural Research Organization, and the Rainforest Alliance are testing across the same producing countries. The common thread is that every option is a cost incurred now against a loss that would otherwise land later, which is a bet only some growers can afford to place.

The bottom line

The figures agree on direction, even where they disagree on magnitude: Kenya and Sri Lanka are already measurably worse off and modelled to lose more growing area by mid-century; Assam is losing its priciest picking window to a compressed season even before whole-crop yield falls; and China, Vietnam, and possibly India stand to gain suitable land as the same warming that hurts the others pushes the optimal zone toward higher latitudes and altitudes. None of this is a forecast this publication is making. It is a synthesis of what growers, researchers, and their own weather stations have already measured. For the structural side of who grows the crop today, on what terms, see Who Grows the World's Tea; for how a shock like a bad harvest reaches the auction floor and the price a grower actually receives, see How the Trade Works.