Global production

Brazil produced almost 43.7 million 60-kg bags of Arabica, roughly 45% of global supply in 2023.⁷

Global production conceals a fragile workforce: around 60% of coffee is grown by an estimated 12.5 million smallholders working farms of 50 acres or less. Around 44% of coffee farmers live below the World Bank's poverty line, leaving them with limited resources to adapt to a changing climate.⁸

Climate change is expected to reduce the suitability and productivity of coffee-growing regions at lower altitudes, especially below 1,000 metres, across major producing regions such as Brazil, Nicaragua, Colombia, and Southeast Asia. Brazil faces a particular challenge: with little high elevation land available, its extensive low-lying Arabica regions are projected to face substantial losses by 2050.⁹ In Ethiopia, the impacts are already visible, with 80% of coffee farmers reporting significant production irregularities in recent years.¹⁰

Consumption

The world consumed 10.3 million tonnes of coffee in 2023. Global coffee consumption grows steadily by around 1-2% each year by volume, while the value of the coffee market is rising faster at over 5% each year as consumers spend more on premium and convenience coffee products.¹¹

While the United States and the European Union together account for the bulk of global coffee consumption, countries across Asia, Africa, and the Middle East are home to large and growing populations of coffee drinkers.

Nowhere is the shift more striking than in China: long associated almost exclusively with tea, it has become one of the fastest-growing coffee markets, driven by a new generation of urban consumers, app-native consumer behaviour, and rapid expansion of domestic chains. Rising consumption is also evident across India, Indonesia, the Gulf, and sub-Saharan Africa. China is reshaping global coffee trade from both ends: as a major coffee buyer and a growing exporter of affordable café equipment to emerging markets.^12,13

A rising demand curve and a contracting supply curve has one predictable outcome: upward pressure on prices and greater volatility along the way.

Current Climate

Anomalous sea surface temperatures

An El Niño has returned

A strong or very strong El Niño is emerging,¹⁴ with ocean surface temperatures in the central to eastern tropical Pacific Ocean at least 2°C above average. After years of La Niña priming the system, this event could rank among the most intense on record; only five comparably strong El Niños have occurred since 1950, the most recent in 2015/16.

El Niño is a climate pattern in which unusual trade winds trigger slow internal ocean waves that pile up vast amounts of heat (roughly 10¹⁵ kJ, equivalent to several million Hiroshima bombs) in the eastern Pacific. It can persist for over a year and is the most predictable large-scale seasonal climate pattern; when an event is very strong, its influence tends to dominate other patterns. Its effects typically lag the peak of the event (November to December) by about three months.

El Niño events can trigger climate regime shifts: abrupt transitions in sea surface temperature, air temperature, and soil moisture. In each historical case, the event created a new normal that lasted for years after the El Niño itself had passed. Strong El Niño events increase the likelihood of such shifts by roughly 20% compared to other years.¹⁵

El Niño Reshapes Global Coffee Risk

No two El Niños are identical, but the broad pattern involves hot, dry conditions across Southeast Asia. Robusta-producing origins, Vietnam and Indonesia, typically bear the heaviest burden. El Niño arrives during Indonesia's blooming phase and Vietnam's cherry-growth period. Minimum temperatures rise across the region, nights stay warmer through flowering, and rainfall is reduced during key monsoon periods.¹⁶ Historically, El Niño events have tended to raise Robusta prices while lowering Arabica prices.^17,18

Warmer, drier conditions pull moisture from coffee leaves and soil aggressively, measured as vapour pressure deficit or the atmosphere's "thirst". Where rising temperatures are not offset by rainfall, vapour pressure deficit quickly crosses yield-limiting thresholds. Arabica yields, for instance, fall sharply above a vapor pressure deficit of roughly 0.82 kPa during fruit development, a threshold that Kenya, Mexico and Tanzania are already exceeding on average. Global heating beyond 2°C is expected to frequently exceed vapour pressure deficit thresholds. If global temperatures rise from 2°C to 3°C, countries such as Peru, Honduras, Venezuela, Ethiopia, Nicaragua, Colombia, and Brazil will be pushed beyond viable coffee-growing conditions.^19,20

Vietnam and Indonesia's warm, moisture-stressed, rain-dependent growing regions are also highly sensitive to drought. Vietnam's irrigation provides some protection; Indonesia's smallholders have far less. Vietnam's exposure also depends on the Indian Ocean Dipole: a climate pattern in which the western Indian Ocean warms relative to the east, when positive, it brings cooler, wetter conditions that can partially offset El Niño drying.¹⁶

Central and South America experience variable weather. Much of Colombia and Guatemala experience drier conditions during an El Niño, with reduced soil moisture across growing regions. Up to a point, this can be beneficial: increased sunlight and moderate drying can stimulate flowering and intensify flavour. But when El Niño pushes conditions past a threshold into drought, plant growth suffers. La Niña brings the opposite problem: yields decline through excess cloud cover, reduced sunlight, and higher disease pressure. Colombia's mountainous geography means local outcomes vary considerably by elevation and slope.²¹

Wetter conditions bring different risks. Lower vapour pressure deficit reduces atmospheric water stress, but creates the warm, humid conditions in which fungal diseases thrive. Coffee leaf rust (Hemileia vastatrix) is a major cause of coffee yield and quality losses in Brazil, and requires fungicide spray applications every season.²² Its spread tracks daily minimum temperatures, relative humidity, and how long leaves stay wet.²³ Unlike drought, which shows up clearly in harvest numbers, rust quietly erodes quality and adds significant cost.

Because El Niño influences Brazil, Vietnam, Colombia, Peru, Mexico, and Indonesia through the same interconnected atmosphere, a strong event can stress several regions at once, leaving few places to turn for compensating supply.

Brazil's relationship with El Niño is complex

El Niño years raise temperatures and water deficits across most of Brazil's major coffee states.²⁴ Reduced rainfall lets more buds reach dormancy together, so that when rain finally returns it triggers a single, concentrated bloom. Too much rain or severe drought can produce premature 'star flowers' that open without proper rehydration.²⁵

Geographically, the picture is not uniform: southern, eastern, and coastal Brazil often receive above-average rain in some El Niño years, and the south reacts less strongly to El Niño-driven climatic changes than other origins.¹⁶ So Brazil can sometimes act as a partial buffer for global supply during strong El Niño events.

The 2015/16 and 2023/24 events offer useful benchmarks. Each struck a system in a different state of health. The 2015/16 El Niño compounded the effects of the 2014/15 drought, particularly in Espírito Santo, where roughly 90% of irrigation-dependent plantations in the northern, northeastern, and northwestern parts faced sudden hydraulic deficit. Elevated temperatures did more damage than the rainfall decline alone. Dry months through the flowering and fruit-development window cut conilon robusta output by nearly 40%.²⁶

The 2023/24 El Niño struck a system under even more prolonged stress. Since 2014, rising temperatures and recurring rainfall deficits had left much of Brazil in a near-permanent hydraulic deficit. Successive shocks: the 2020 drought, the July 2021 black frost, and droughts in 2016/17, 2019/20, and 2023, disrupted the normal biennial cycle as farmers replanted large areas off-schedule. Many farms then went 120–130 days without meaningful rain during the 2024 growing phase. Combined with extreme heat during the flowering window, these conditions pushed much of the Arabica belt into stress flowering: trees prioritising leaf regrowth rather than fruit, and cannibalising the next crop.²⁵

The July 2021 'black frost': pre-dawn temperatures of −1 to −2 °C across Minas Gerais and São Paulo, the worst since 1994, killed unhardened young plants and extended recovery periods by several years.²⁷ Pruning after frost costs one to two years of regrowth, forcing an artificial, extended 'off' cycle. A severe hailstorm in 2025 damaged around tens of thousands of hectares of coffee farms in Minas Gerais. While national production remained largely unaffected, the event highlighted the growing frequency of climate-related extremes.

An El Niño in late 2026 and extending into early 2027 may lower frost risk in Brazil but increase temperatures during flowering and grain filling, while also altering rainfall patterns. Together, these conditions could place further pressure on the size and quality of the 2027 harvest.

The infographic below illustrates the phenological cycle of Arabica coffee in Brazil and shows whether deviations in key climate variables from their optimal ranges during the 2015/16 and 2023/24 El Niño events were favourable or unfavourable to crop development in Minas Gerais, Brazil's largest Arabica-producing state.

What Makes 2026 Different

Background warming is raising baseline temperature and vapour pressure deficit stress across coffee growing regions, so that even a moderate El Niño can now push conditions past thresholds once reached only by the most extreme events. The 2026 El Niño is forming in an atmosphere hotter and more carbon dioxide-laden than any in the history of coffee cultivation. Climate change is also increasing the swings of impact, with rainfall variability of El Niño and La Niña events projected to increase.^28,29

The nature of drought is changing. Its main driver is shifting from too little rain to too much atmospheric evaporative demand: record land-surface temperatures evaporate water so fast that severe soil-moisture drought can develop even when rainfall is only slightly below average. Drought can inflict up to 80% productivity loss in the driest years, especially in marginal, non-irrigated coffee growing regions.³⁰

In a warmer world, extremes are faster and more damaging. For coffee, adaptation will require redesigning farm microclimates through shade architecture, improved soil and water practices, agroforestry and more heat-tolerant coffee varieties.³¹

Protecting existing forests in coffee landscapes secures the local water cycles and biodiversity that the crop depends on.^32,33

Encouragingly, weather conditions this year were favorable and have supported production recovery, with strong pre-El Niño harvests expected in Brazil and Vietnam. No significant cold spells have been reported so far. While this relatively strong starting point provides a useful buffer going into a warmer and more variable climate, risk remains as El Niño unfolds.

Permalink

]]>El Niño Effect on Coffee YieldUncharted WatersTue, 31 Mar 2026 04:39:26 +0000https://www.unchartedwaters.com/insights/tracking-weather-impacts-on-global-coffee-markets65c60693ed8ae44c17e6713c:65c615c663bc2439a420d65b:69cdf2fe6352dc7767885b6b

Uncharted Waters has partnered with Single O, a Sydney-based independent coffee roaster, to monitor how weather extremes affect coffee production and trade worldwide.

Single O sources single-origin coffee from twenty-six countries, with roasting operations in Australia and Japan supplying over 500 independent cafes. Like much of the coffee industry, the company has faced sharp price volatility in recent years. After reaching multi-year highs in 2024 and 2025, coffee prices remain elevated — a direct consequence of weather disruptions in key producing regions including Brazil and Vietnam. Climate-related production shocks are becoming harder to anticipate using historical patterns alone, as events now frequently coincide across multiple regions, compounding market impacts.

Current forecasts place the probability of El Niño conditions at 58–61% from May through the end of 2026 — a pattern historically associated with drought stress in parts of Asia and Latin America. This makes 2026 an important year to track weather impacts on coffee production closely.

In this pilot, Uncharted Waters is developing and testing a tracking tool that draws on its digital twin framework — integrating real-time climate data with crop and trade modeling — to assess weather impacts across the global coffee supply chain. Working with Single O, we are exploring how this kind of analysis can support sourcing decisions in practice.

This project is supported by the Australian Government's Industry Growth Program.

When the 2025 monsoon arrived in central India, farmers across Maharashtra and Madhya Pradesh prepared for what they hoped would be a productive growing season. The early rains in Maharashtra created opportunities for timely sowing, but the pattern that followed—heavy downpours alternating with prolonged dry spells—disrupted field preparation and crop establishment.

"This year, the rains were very untimely, so the harvest was not as good as expected," explained a farmer from Selu Amba village in Beed district, Maharashtra. His concern wasn't about whether rain fell, but when and how—the difference between a successful season and one requiring costly interventions just to reach normal yields.

When Rain Comes, But Not as Expected

In our pilot districts of Ahilyanagar, Beed, and Sambhajinagar in Maharashtra, farmers reported the monsoon beginning earlier than usual. But the early start didn't translate into the benefits farmers had hoped for. Rainfall arrived in intense bursts separated by extended dry periods, creating a stop-start pattern that disrupted field preparation and forced many farmers to re-sow their crops multiple times.

The experience in Madhya Pradesh's Seoni district differed. Most farmers there reported the monsoon beginning on time or slightly late, with more consistent rainfall distribution overall. Yet even with better timing, farmers faced their own set of challenges. The inconsistent rainfall created conditions favorable for pest and disease outbreaks, which became the primary concern across Seoni district.

The Hidden Costs of Managing Through Variability

What emerged most clearly from a quick pilot survey among 545 farmers across four districts in both states wasn't primarily about rainfall amounts—it was about the cascade of operational challenges and economic pressures that followed.

Pest and disease outbreaks dominated farmer concerns. Across all surveyed districts, pesticide application was the most common coping strategy, with farmers frequently mentioning the need for guidance on spraying and pesticide application. The fluctuating humidity and temperature created by alternating wet and dry periods provided ideal conditions for pest proliferation.

Managing these threats came at a price. When asked about their primary concerns, 60-70% of farmers across both states pointed to rising input costs—seeds, fertilizers, and especially pesticides—as their greatest challenge. The variable rainfall patterns meant farmers had to invest in multiple re-sowings, increased pest control measures, and additional inputs just to maintain crop growth.

Limited access to credit, difficulties navigating crop insurance claims, and uncertain market prices compounded the challenges of an already demanding season. This reveals a crucial insight: farmers can achieve reasonable crop production while still facing significant pressure on their incomes. For smallholder farmers operating on thin margins, the effort and expense required to manage through variable conditions can significantly erode profitability even when final yields appear normal.

Figure 1 Farmer’s perception of this year’s monsoon onset, rainfall amount and pattern, and the main climate hazard

Triangulating Knowledge: Climate Data, Models, and Farmer Experience

This pilot survey, while modest in scale, provided valuable opportunities to triangulate different sources of knowledge about the 2025 monsoon season. By combining farmer experiences with climate data and crop modeling, a more complete picture emerged.

Our digital twin of the global food system—which integrates real-time climate data with crop modeling—confirmed the regional patterns farmers described. Maharashtra experienced particularly wet conditions in May and October, with the early monsoon onset that farmers reported. August brought extreme rainfall to some areas. Madhya Pradesh saw concentrated July rainfall but more stable overall distribution through the season.

Both regions showed above-average seasonal precipitation totals compared to a 15-year baseline—Maharashtra notably so. By the time crops were nearing harvest, this October, our model projected a +9% production anomaly for the monsoon season across India overall. This is quite a significant anomaly on a national scale for India's agriculture, partly reflecting the favorable weather conditions and partly continuing trends in area expansion and yield improvements needed to keep pace with India's growing population. The projection suggests that despite the management difficulties farmers faced, adequate seasonal moisture ultimately benefitted crop development. The upcoming harvest data will reveal whether this production outlook materializes—and crucially, at what cost to farmers.

What Farmers Said They Need

The pilot survey revealed significant gaps between the challenges farmers face and the support available to them. Awareness of, uptake, and usage of weather forecasts differed between cropping systems and states. Among Maharashtra farmers, 78% had heard of IMD's five-day forecasts, and 43% reported actively using them. Confidence in forecasts was high: 95% expressed at least some confidence. In Madhya Pradesh, only 45% of farmers had heard of IMD forecasts, and just 7% reported using them. Those who  used forecasts we more likely to apply coping measures. However, fewer than 10% of surveyed farmers reported receiving formal training on coping strategies or improved crop varieties, despite recurring monsoon-related challenges.

Farmers emphasized several priorities. Many requested training on pesticide application and integrated pest management—practical knowledge for responding to ongoing threats. They also mentioned needing advice on improved seeds, fertilizers, and disaster-related risk management methods.

Economic and financial support appeared just as frequently as agricultural concerns. Farmers emphasized needing help with financial assistance, market information and crop insurance, and many requested training not just on farming practices but on insurance schemes, credit access, and financial planning—reinforcing that they think in terms of net returns, not just production volumes.

What Uncharted Waters Does

Uncharted Waters monitors climate volatility and its impacts in real-time, tracking how weather events affect crop production and how water resources buffer these impacts, while analyzing how disruptions cascade through global food supply chains. Our digital twin of the global food system integrates process-based crop modeling with machine learning to provide monthly updates on weather events, water availability, recent harvests, and expected production.

We don't just produce data—we work to ensure it connects to decision-making at all levels. In India, we partner with WOTR to integrate climate and crop intelligence into advisory services. By monitoring supply, demand, stocks and trade flows in real-time, we deliver market insights before official statistics become available—helping organizations from local NGOs supporting farmers to businesses managing agricultural commodity risk make informed decisions faster.

This pilot project in Maharashtra and Madhya Pradesh, supported by the Kyeema Foundation and Palladium, represents our commitment to evidence-based approaches that triangulate multiple sources of knowledge. While the survey's geographic scope was limited—four districts across two states—it provided valuable insight into how climate data, model projections, and farmer experiences relate to each other. We're validating our model projections against actual harvest outcomes, learning how climate information translates into on-the-ground realities, and refining how different types of knowledge work together to strengthen decision-making.

As extreme weather events occur more frequently and their impacts spread wider, timely insight on current production, expected supply, and farmer conditions becomes crucial for anticipating risks and managing uncertainty effectively. We'll continue monitoring the 2025-26 agricultural year as harvest data becomes available, working with WOTR to understand what this season ultimately meant for farmers—and what it teaches us about building more resilient food systems through integrated, evidence-based approaches.

Uncharted Waters works with Wageningen Environmental Research and WOTR to assess climate impacts on water and food systems. This analysis was conducted in partnership with WOTR's farmer network in Maharashtra and Madhya Pradesh, with support from the Kyeema Foundation and Palladium.

This analysis represents our current understanding and predictions may change as new information becomes available.

Maps in this blog follow the Indian government guidelines. The designations employed and the presentation of these materials do not imply the expression of any opinion whatsoever on the part of Uncharted Waters concerning the legal status of any country, territory or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.

Farmers in Maharashtra and Madhya Pradesh, two major food producing states in central India, perceive the monsoon as more unpredictable than in earlier years. A recent questionnaire-based survey completed by 545 farmers across four districts in central India reveals an interesting relationship between forecast use and adaptive behavior.

When Patterns Become Less Reliable

Farmers who depend on predictable seasonal patterns for decisions about planting, crop selection, and resource allocation face mounting challenges when those patterns shift. The 2025 monsoon illustrated this clearly. In Maharashtra, early rains created opportunities for timely sowing, but the pattern that followed—heavy downpours alternating with prolonged dry spells—disrupted farm management and crop establishment. Rainfall arrived in intense bursts separated by extended dry periods, forcing many farmers to re-sow their crops multiple times.

The experience in Madhya Pradesh's Seoni district differed. Most farmers reported the monsoon beginning on time or slightly late, with more consistent rainfall distribution overall. Yet even with better timing, inconsistent rainfall created conditions favorable for pest and disease outbreaks, which became the primary concern across the district.

The growing uncertainty makes timely weather information potentially more valuable, yet farmers' awareness of, trust in, and use of weather forecasts varies considerably—variation that is strongly related to the cropping systems they manage.

Awareness, Trust, and Use of Weather Forecasts

Farmers were asked about their familiarity with India Meteorological Department's five-day weather forecasts, their confidence in weather forecasts generally, and whether they actively use IMD forecasts in farming decisions.

The responses revealed substantial regional differences. In Maharashtra, most farmers had heard of IMD's five-day forecasts and nearly all expressed confidence in weather information, though actual usage stood at 43%. In Madhya Pradesh, awareness and usage remained much lower, with 38% responding "don't know" when asked about forecast reliability—suggesting limited exposure to weather information services rather than active skepticism.

Figure 1 Crop systems (top), trust in forecasts (mid-left), and usage (mid-right) by crop system and perceived monsoon unpredictability change (bottom-left) and adaptive measure adaption per state (bottom-middle). For Maharashtra, users of forecast showed higher measure adoption, both general drought- and flood-related mitigation measures as well as measures related to this year’s weather anomalies - mainly seed selection and crop rotation adjustments (bottom-right). (Madhya Pradesh had too few users to distinguish).

These regional patterns align closely with cropping system differences. Maharashtra's agricultural landscape is more diverse, with farmers managing maize-based mixed systems, soybean, cotton, pigeon pea, and various crop combinations. Madhya Pradesh's farmers, especially in the east, predominantly grow paddy, i.e. rice, most as a monoculture. A significant part of the state lies in the core monsoon zone and receives abundant rainfall. Paddy fields are usually flooded with a shallow layer of water, meaning farmers depend less on daily rainfall fluctuations to maintain sufficient water supply to the plants.  

These waterlogged conditions create a more stable, humid microclimate that also affects pest and disease dynamics differently. While paddy faces its share of pressure, disease management becomes more predictable, aligning with crop growth stages rather than day-to-day weather variability. This contrasts with rainfed crops where moisture conditions fluctuate more dramatically, making short-term weather information more valuable for timing interventions.

Among farmers practicing mixed cropping systems, forecast usage reached considerably higher levels than among monoculture farmers. This pattern held even within the same crop: paddy farmers who also grew maize, wheat, or pulses showed three times higher forecast usage than those growing only paddy.

Managing multiple crops with different timing requirements, water needs, and climate sensitivities corresponds with higher perceived value of weather information. Short-term forecasts become more valuable when farmers must coordinate decisions across crops with different requirements and vulnerabilities, compared to managing a single crop with relatively standardized practices.

Forecasts and Protective Measures

Farmers who actively use weather forecasts are also those taking measures to protect their crops and livelihoods. In Maharashtra, where forecast adoption is more established, 95% of forecast users reported taking measures against heavy rainfall, compared to 65% of non-users. The pattern held for drought measures as well, though both groups showed high rates of drought preparedness.

Crop insurance adoption followed a similar pattern, with substantially higher rates among farmers in regions with greater forecast engagement.

The relationship becomes more direct when examining adjustments made in response to recent weather conditions. When asked whether weather changes had prompted them to consider adjustments in crop varieties, seed selection, or farming practices, 40% of forecast users in Maharashtra answered yes, compared to 26% of non-users. Farmers mentioned condition-based seed selection and crop rotation decisions as common responses.

These general protective measures like insurance differ from specific adjustments to recent weather—the latter showing a more immediate connection between observed conditions, information use, and farming decisions. Farmers using forecasts appear more likely to both maintain ongoing risk management strategies and adjust practices based on changing conditions.

Information Needs Across Timescales

Weather information serves different decision-making needs across different timescales. Five-day forecasts help farmers time immediate operations: when to spray, when to irrigate, when to harvest. These short-term decisions matter enormously for protecting investments and optimizing yields, particularly for farmers managing complex cropping systems with multiple crops that have different sensitivities to temperature, soil moisture, and pest risks at different growth stages.

But farmers also make decisions that require longer time horizons. Which crops to plant? How much area to allocate? When to expect market pressure from production elsewhere? These questions require understanding not just next week's weather, but how the current season compares to normal patterns, what's happening in other growing regions, and how weather events are affecting production and trade flows.

The relationship between forecast engagement and adaptive behavior observed here suggests that when farmers have access to timely, relevant information, they use it to manage risk more effectively. As weather variability continues to increase, building information infrastructure across multiple timescales becomes increasingly important.

What Uncharted Waters Does

Uncharted Waters monitors weather volatility and its impacts in real-time, tracking how climate events affect crop production and how water resources buffer these impacts, while analyzing how disruptions cascade through global food supply chains. Our digital twin of the global food system integrates process-based crop modeling with machine learning to provide monthly updates on weather events, water availability, recent harvests, and expected production.

In India, we partner with WOTR to integrate climate and crop intelligence into advisory services. One such service is WOTR’s five-day weather forecasts, created in cooperation with the Indian Meteorological Department, to support immediate operational decisions. By monitoring supply, demand, stocks and trade flows in real-time, we deliver additional market insights before official statistics become available—helping organizations from local NGOs supporting farmers to businesses managing agricultural commodity risk make informed decisions faster.

This project in Maharashtra and Madhya Pradesh, supported by the Kyeema Foundation and Palladium, represents our commitment to evidence-based approaches that triangulate multiple sources of knowledge. The questionnaires provided insight into how forecast engagement relates to adaptive behavior, revealing patterns across different cropping systems and regions. We're validating our model projections against actual harvest outcomes, learning how climate information translates into on-the-ground realities, and refining how different types of knowledge work together to strengthen decision-making.

As weather variability continues to increase, timely insight on current production, expected supply, and farmer conditions becomes crucial for anticipating risks and managing uncertainty effectively. Understanding who uses forecasts and how that relates to protective actions provides guidance for building more resilient food systems through integrated, evidence-based approaches.

Uncharted Waters works with Wageningen Environmental Research and WOTR to assess climate impacts on water and food systems. This analysis was conducted in partnership with WOTR's farmer network in Maharashtra and Madhya Pradesh, with support from Kyeema Foundation and Palladium.

Central Asia's mountain snowpack reached critically low levels in 2025, with smaller tributaries losing over 50% of their minimum snow cover compared to two decades ago. While major rivers maintain flows through increased glacier melt, this temporary stability masks a looming crisis for millions who depend on seasonal snowpack for water security.

In our latest analysis published in Climate Impacts Tracker Asia, Uncharted Waters and leading water experts from Uzbekistan's Tashkent Institute of Irrigation and Agricultural Mechanization Engineers (TIIAME-NRU) reveal an ongoing decline in snowpack across Central Asian watersheds reducing river flows and water availability for irrigation and other uses.

Dramatic declines in smaller tributaries

The smaller, lower-elevation tributaries are experiencing the most severe impacts. The Surhandarya basin, which supports agriculture in southern Uzbekistan, has seen its minimum snow cover plummet by 51% compared to the 2001-2007 period[1]. The Kashkadarya shows a 54% reduction, with 2025 data showing the basin retaining only 0.7 square kilometers of snow cover—essentially bare. The Chirchik River, flowing near Tashkent and supporting the capital region's agriculture, has lost 31% of its minimum snow cover. The 2025 measurement of 47.4 square kilometers represents less than half of what was typical just a decade ago.

The region's water dynamics present a paradox that masks the severity of the crisis. The major rivers—the Amu Darya and Syr Darya—maintain relatively stable flows, benefiting from increased glacier melt in their upper reaches. This year, the Amu Darya basin still retains about 19,500 square kilometers of minimum snow cover, despite representing a 29% decline from the 2001-2007 average. However, he smaller tributaries, which depend more heavily on seasonal snow accumulation rather than glacier melt, are already in severe decline. Even basins that showed slight increases in 2008-2015, like the Chatkal, have collapsed to minimal coverage—just 3.1 square kilometers in 2025.

Figure 1 Minimum snow and glacier cover in summer, by elevation, for 2025 and three previous periods

2025 drought

Precipitation from January through spring ran 30-70% below average every month, with only June showing near-normal rainfall. Spring temperatures exceeded the 15-year average, accelerating snowmelt when water should have been stored as snow for gradual summer release. With the arrival of autumn in the high mountains, snow cover remains far below normal—a pattern that breaks from typical seasonal accumulation.

Figure 2 Seasonal snow cover pattern. While winter snow cover holds up reasonably well, this snow melts earlier and faster, resulting in a much longer period without significant meltwater contribution in the lower tributaries at a time when water demand is high.

For Uzbekistan and its neighbors, the message is clear: traditional water management approaches based on historical abundance must give way to strategies built for scarcity. Every drop counts when the mountains no longer hold their water in reserve.

Read the full analysis: ‘Central Asia's Vanishing Snowpack: A Real-Time Crisis Unfolding in 2025’ in Climate Impacts Tracker Asia.

This analysis continues our ongoing collaboration with Prof. Abdulkhakim Salokhiddinov and Andrej Savitskiy at TIIAME-NRU, combining real-time satellite monitoring with deep regional expertise to provide decision-relevant insights on Central Asia's evolving water crisis.

[1] 2016-2024 versus 2001-2007 average

When the Rains Don't Stop

August 2025 brought extreme rainfall to Punjab, India's premier Basmati rice region, triggering widespread flooding across districts that produce the aromatic rice variety prized in international markets. The monsoon season started early with heavy May rains and remained persistently wet throughout the growing period. August precipitation reached almost double the 15-year average in some districts, with the most intense rainfall concentrated in the final week—disrupting the critical mid-season growth phase.

The blue shading in satellite data shows how exceptional August precipitation was compared to historical patterns, with extreme rainfall focused in key rice-growing areas like Amritsar district. ERA5 climate data reveals 3 to 6 times more intense rainfall events, measured by days exceeding 30mm, 50mm, or 80mm daily precipitation thresholds.

Assessing the Impact

It's too early to determine the exact impact on India's overall rice production and prices. While Punjab experienced severe flooding, other major rice-producing regions like Madhya Pradesh in central India and Karnataka and Andhra Pradesh in the south also saw above-average rainfall but with less severe flooding. When disaster strikes one region, neighboring areas with better water availability sometimes benefit from improved growing conditions. Rice, due to its distinct cultivation practices in flooded paddies, can tolerate intense rainfall better than other cereals—though even rice has limits when fields remain underwater for extended periods.

Punjab remains the state with the highest rice yields in India, approaching 7 tons per hectare. However, other states have been catching up rapidly. Just fifteen years ago, Madhya Pradesh and Uttar Pradesh yields remained below 2 tons per hectare. They have since doubled. Uttar Pradesh and Telangana have now overtaken Punjab in total production volume, reducing Punjab's relative importance to national supply.

With harvest about to begin, the full impact on domestic prices and export volumes remains to be assessed. Better yields in other regions might partially compensate for Punjab's losses, maintaining overall production levels despite the localized flooding.

India's Evolving Role in Global Rice Markets

The Punjab floods matter particularly for global rice trade. Punjab is a major Basmati rice exporting region, and damage to this premium crop variety has outsized impacts on international markets even when India's overall domestic rice supply remains adequate. India has transformed from a primarily self-sufficient producer—occasionally importing small quantities during deficit years—to a net exporter in recent years through substantial yield and production increases.

However, production must keep pace with population growth. Below-average harvests have repeatedly led to export restrictions that ripple through nations increasingly reliant on Indian rice. India's position as a "pivot" in global rice markets means its domestic weather events can trigger food security concerns thousands of kilometers away, particularly in the Middle East and West Africa where Indian rice represents a significant share of imports.

This volatility reflects a broader challenge: as yields increase and India becomes a more important exporter, the country's exposure to weather variability creates larger disruptions in global markets when extreme events occur. The same progress that enables export growth also raises the stakes when monsoons fail—or when they don't stop.

Figure 3 Major rice producing countries, as expressed in energy content (kCal per year). For food use only (corrected for feed and other uses). Average over 2010-2025. Source: Uncharted Waters, using FAOSTAT data up to 2022 for calibration

Figure 4 Imports and exports of rice for food use. Blue regions are net exporters—countries producing more food than they consume. Red regions show nations depending on imports. Average over 2010-2025. Source: Uncharted Waters, using FAOSTAT data up to 2022 for calibration

Figure 5 India’s import and export. Blue lines and shading show major destinations for Indian rice, red minor imports. Average over 2010-2025. Source: Uncharted Waters, using FAOSTAT data up to 2022 for calibration

Climate Volatility: The New Normal

Recent years have seen multiple rice crises across Asia. Japan faced shortages requiring emergency imports from South Korea for the first time in over 25 years, with ministers being sacked over the crisis. Indonesia battled drought-driven production declines. India itself has oscillated between export bans during poor harvests (2022-2024) and resuming exports when conditions improve.

The common thread: increasingly erratic weather patterns making it harder to rely on historical climate data and lived experience to predict growing conditions. This is precisely why real-time monitoring and analysis has become essential for food security planning.

What we're observing in Punjab—extreme precipitation events concentrated in short periods—fits the broader pattern of weather intensification. These aren't simply "bad years" to be weathered until conditions return to normal. The variability itself is the new normal, requiring different approaches to agricultural planning, water management, and trade policy.

Supporting Timely Decisions

Uncharted Waters monitors these impacts in real-time, tracking how weather events affect crop production and analyzing how disruptions in one region cascade through global food supply chains. We partner with Wageningen Environmental Research (www.wur.nl) and, in India, with WOTR (www.wotr.org) to assess climate impacts on water and food systems.

Our digital twin of the global food system integrates process-based crop modeling with machine learning to provide monthly updates on weather events, recent harvests, and expected production. By monitoring stocks and trade flows in real-time, we deliver market insights before observation data becomes available—helping organizations from local NGOs supporting farmers to businesses exposed to agricultural commodity volatility make informed decisions faster.

In India, WOTR is integrating our data into advisory services reaching 1.2 million farmers through their network. Early feedback from 500 pilot users shows positive reception of enhanced climate information for farming decisions.

As extreme events occur more frequently and their impacts spread wider, timely insight on current production and expected supply becomes crucial for anticipating tipping points and managing uncertainty effectively.

Uncharted Waters analyzes how climate events and water availability affect crop production, food supply, and ultimately food security. We connect food shortages in regions impacted by extreme weather to the export capacity of countries with surplus production.

This analysis represents our current understanding and predictions may change as new information becomes available.

This work was supported by the Kyeema Foundation.

Q2 2025

Uncharted Waters continues to develop and refine our Food Security Indicator (FSI), providing near real-time insights into global food security patterns at the country level. Our Q2 2025 analysis reveals persistent challenges driven primarily by trade and production factors, with conflict continuing to shape food security across multiple regions. While our methodology improvements have refined how we assess multi-actor violence, the fundamental drivers of food insecurity remain consistent: inadequate food supply relative to demand, ongoing conflicts, and regional instability that creates cascading effects across borders.

The second quarter of 2025 shows mixed patterns across regions, with some improvements alongside gradual changes in already affected areas. Regional impacts have shifted since Q1 in Central America, Sub-Saharan Africa, and East Asia, while several nations continue to experience reductions in food storage capacity. Our analysis indicates that countries struggling to meet basic food demands face ongoing challenges from both immediate supply constraints and longer-term structural vulnerabilities.

The Global Situation

Figure 1: Food Security Indicator for Q2 2025. Lower scores (darker brown) indicate lower food security (Conflict data from ACLED, accessed 15-05-’25).

Our Q2 2025 analysis reveals a complex global food security landscape where trade and production factors continue to be primary drivers of national-level food insecurity. High FSI values are predominantly attributed to low realized food demand, in either total food availability or specific crops experiencing the greatest supply shortages.

Several regions show persistent vulnerability. In the Caribbean, Haiti, Belize, and Cuba face potentially unmet food demands, with Haiti experiencing particularly severe challenges with significant storage reductions that directly impact our FSI calculations. Venezuela and Colombia in South America continue to struggle with food access, reflecting ongoing economic and political instability that constrains their ability to secure adequate food supplies through trade or domestic production, and to ensure effective distribution and access within these countries.

The East African situation remains critical, with Somalia, Uganda, South Sudan, and Sudan all showing signs of potentially unmet food demands. The Sudanese Civil War and ongoing tensions in South Sudan continue to severely impact food security in these countries, creating a regional crisis that extends beyond national borders. In other parts of Africa, Libya, Botswana, and Equatorial Guinea face similar challenges, with Equatorial Guinea among the few countries where large storage reductions are directly reflected in FSI scores.

Asia presents a mixed picture, with Bhutan facing food demand challenges while North Korea experiences both potentially unmet demands and significant storage reductions that impact its FSI rating. The Pacific Island Nations, particularly the Solomon Islands and Vanuatu, continue to face structural vulnerabilities due to their geographic isolation and limited domestic production capacity.

Production anomalies have been observed in West Africa and parts of the Middle East, including Iran, Iraq, and Turkmenistan. While these production issues don't directly impact FSI calculations, their effects cascade through our FOODDivider digital twin, creating secondary impacts on global food trade patterns and regional food security.

Regional Highlight

Sub-Saharan Africa: Mixed Patterns in an Ongoing Challenge

Sub-Saharan Africa presents a complex picture of gradual changes within the region's ongoing food security challenges. While the overall regional impact has shifted since Q1 2025, the pattern shows variations rather than uniform trends across countries that were already experiencing food insecurity.

The most significant factor continues to be the cascading effects of conflict, particularly the Sudanese Civil War, which creates spillover impacts affecting neighboring countries beyond Sudan's borders. These regional conflict effects demonstrate how food insecurity can transcend national boundaries, with trade disruptions and population movements affecting broader areas than the immediate conflict zones.

Storage capacity changes vary significantly across the region. While several countries have experienced reductions in food reserves, Equatorial Guinea and South Sudan are among the few where these storage reductions directly impact FSI calculations. This suggests that most countries in the region maintain some buffer capacity, though the countries where storage affects FSI scores represent particularly vulnerable situations where reserve depletion directly threatens food access.

The integration of economic equality measures within our FSI continues to reveal how food insecurity disproportionately affects the poorest citizens within countries, even where national-level indicators might suggest adequate food availability. This internal distribution factor remains a critical component of understanding true food security conditions across the region.

Changes from Previous Quarter

The transition from Q1 to Q2 2025 reveals mixed patterns across regions, with some improvements alongside gradual changes in countries already experiencing food security challenges. Regional impacts have shifted across Central America, Sub-Saharan Africa, and East Asia, representing changes in existing conditions rather than new countries becoming food insecure.

Our refined methodology for processing conflict data has resulted in some notable changes in how we assess food security impacts. The new approach reduces the measured impact of multi-actor low-level violence, leading to decreased assessments of food security impacts from crime in countries like Mexico, Brazil, and Colombia, as well as from multi-actor civil wars in Myanmar and parts of East Africa. While this methodological improvement provides more precise measurements, these regions continue to face food security challenges from conflict, even if the calculated impacts are now more nuanced.

Storage capacity presents a varied picture, with several countries experiencing reductions in food reserves. However, only Haiti, South Sudan, Equatorial Guinea, and North Korea see these storage reductions directly reflected in FSI calculations, suggesting that most countries retain sufficient buffer capacity or alternative mechanisms to maintain food access despite reduced reserves.

Future Developments

We continue to enhance our FSI methodology and expand its analytical capabilities. Our immediate priorities include developing subnational resolution updates that will provide more granular insights into crop production and related food security patterns within countries. This development will be particularly valuable for large countries with diverse regional conditions and will help identify specific areas where interventions might be most effective.

Our analysis does not include Gaza, where famine is confirmed by the UN. We are exploring options to incorporate Gaza and the West Bank, recognizing the need for specialized methodological approaches given the unique circumstances.

Our research team is also examining the long-term sustainability of food production systems, particularly focusing on regions where food production heavily depends on irrigation that exceeds natural recharge rates. This analysis aims to highlight how unsustainable water use might affect food security not only in producing regions but also in countries that import food produced using such water resources. This forward-looking approach will help identify emerging vulnerabilities before they manifest as acute food security crises.

Additionally, we continue to refine our seasonal forecasting capabilities and improve the integration of climate data with crop modeling and trade analysis. These improvements will enhance our ability to provide early warning of developing food security challenges and support more proactive response planning.

Key Insights

Food insecurity in Q2 2025 remains fundamentally driven by the inability of countries to meet food demand through either domestic production or trade. The mixed patterns observed across regions suggest that while some areas show gradual improvements, others continue to face ongoing challenges, particularly in Central America, Sub-Saharan Africa, and East Asia. Countries experiencing storage reductions alongside demand shortfalls represent areas requiring continued attention and support.

Conflict continues to be a major driver of food insecurity, though our refined methodology assumes that the impact of organized, sustained conflicts (such as those in Ukraine, Syria, Yemen, and Sudan) creates more significant food security challenges than diffuse, multi-actor violence.

The cascading effects of production anomalies through global trade networks highlight the interconnected nature of food systems and the importance of maintaining robust trade relationships and supply chain resilience. Countries with limited economic capacity to compete in global markets or offset production issues with imports continue to face the greatest vulnerabilities to food insecurity.

Our FSI complements existing approaches such as the FAO Global Information and Early Warning System (GIEWS) and the Integrated Food Security Phase Classification (IPC), providing valuable insights where traditional data collection faces challenges. For more information about our methodology or to discuss how our data might support your organization's work, please contact us.

Q4 2024

Uncharted Waters are proud to announce the release of the beta version of our Food Security Indicator (FSI). This first release provides the FSI for the 4th quarter of 2024. Utilising observed climate data, the latest generation of crop modelling, and our bespoke digital twin for global food trade, FOODDivider, the FSI provides a near real-time metric for the stability of food access across the globe at the individual country level.

Our FSI incorporates measures such as access to specific food types, the level of inequality and conflict within countries, and regional food instability. Benchmarked against other available food security measures from organisations such as the Food and Agriculture Organization (FAO) and the Integrated Food Security Phase Classification (IPC), and assessed for historical representativeness, we are confident that the UW FSI can provide insights into those countries and regions where people may be at risk of inadequate food security.

The FSI is currently deployed at a country-level resolution using data calculated or aggregated to the same level.

The Global Situation

Our analysis reveals several regions experiencing significant food insecurity, including parts of Central America and the Caribbean, Colombia, Venezuela, Brazil, Ukraine, Myanmar, Syria and other parts of West Asia, sub-Saharan Africa, Libya, and several Pacific Island nations including Papua New Guinea, the Solomon Islands, and Vanuatu.

Figure 1: Food Security Indicator for Q4 2024. Lower scores (darker brown) indicate lower food security. Source: Uncharted Waters. Conflict metrics used data from ACLED (accesses 07/01/2025 from acleddata.com)

The drivers of low FSI vary considerably by region. Severe conflict, including inter-state wars such as in Ukraine, intra-state conflicts in Myanmar and Sudan, and criminal activity and societal unrest in Mexico, significantly affects food security across the globe. Regional insecurity appears to be creating challenges in Eastern Africa and the Indo-Pacific. Economic inequality within countries is intensifying food insecurity risks in Southern Africa and Central and South America. Trade and production factors predominantly impact nations in Africa, particularly Somalia, Eritrea, Equatorial Guinea, and Eswatini.

Changes from Q3 2024

Between Q3 and Q4 2024, our analysis shows no major changes in the global food security landscape. The most notable improvement is seen in Guinea, with North African countries generally experiencing modest gains in food security. These improvements are mostly driven by increased food availability either through increased production or trade flows. In some cases, improvements reflect increased availability of specific food groups. Despite these gains, the modelled situation remains fragile, with the FSI remaining low in many of the region's nations.

A handle of countries, all located within Africa, have seen relevant declines in food security. According to our digital twin, Madagascar and Burundi have seen decreased ability to meet food demand, both in terms of total kilocalorie requirements and for specific food groups. This does not appear to be related to production anomalies but is likely driven by trade dynamics. Similarly, Zambia and Benin show FSI reductions primarily due to declining availability of specific food groups, also attributed to trade influences.

Madagascar's modelled food storage has been depleted, likely reflecting attempts to buffer the trade impacts. The FSI decline in all four countries has been somewhat mitigated by general improvement in regional FSI, assumed to aid in the stability of food access.

Future Developments

We continue to improve and validate our FSI. We are planning a version based on subnational production and demand estimates, which will provide more granular information on where food insecurity is most severe within countries. We are also examining regions where food production heavily depends on irrigation that exceeds natural recharge rates and is unsustainable long-term. Our aim is to highlight how this might affect food security not only locally but also in countries that import food produced using such water resources. Additionally, we continue to enhance our seasonal forecasts.

Our FSI can complement existing approaches such as the FAO Global Information and Early Warning System (GIEWS) and the Integrated Food Security Phase Classification (IPC), which often struggle to access necessary data in conflict areas. It can also serve alongside early warning systems such as the Famine Early Warning Systems Network (FEWS Net), which is currently unavailable and at risk due to a reorientation of US funding and support.

Key Insights

Food insecurity remains a significant threat across multiple global regions. The rise in severe conflict and violence across various categories represents the greatest threat in many areas. Nations lacking the economic capacity to offset production issues with imports are also vulnerable. Such countries are likely to experience further declines in food security if exposed to climate change impacts. Areas already affected by climate extremes, such as the extensive floods and droughts seen in 2024 across sub-Saharan Africa, face greater insecurity risks, especially if unable to compete effectively in global trade.

For more information, please contact us.

In a recent one-year pilot project for the Food and Agriculture Organization (FAO) in Rome, Uncharted Waters* have developed and tested an innovative approach to understanding and predicting climate-related conflict risks. The project specifically focuses on how changes in our food systems, driven by climate variability, might contribute to conflict risk in vulnerable regions.

Why This Matters

The spread of conflict coupled with climate and economic stressors is pushing millions of people to the brink, with acute food insecurity set to increase in both magnitude and severity according to a recent FAO report. Evidence of a direct relationship between climate, food insecurity and conflict, however, is patchy and incomplete. While we know that environmental degradation and climate change can affect resource availability and threaten livelihoods, the link to conflict is far from straightforward. Understanding these complex relationships is crucial for organizations like FAO to develop "conflict-sensitive" climate adaptation strategies in the agriculture sector.

Enter CoPro: A New Tool for Risk Assessment

At the heart of our pilot project is CoPro, a machine learning model designed to predict conflict probability. What makes CoPro unique is its ability to process diverse spatial data alongside conflict records to forecast the likelihood of future conflicts. The model doesn't just look at traditional conflict indicators – it also incorporates climate and food system data, making it particularly valuable for understanding how climate change might affect conflict risk through its impact on agriculture and food security.

The Data Challenge

A critical aspect of this project was the selection and integration of indicators. We identified 38 unique indicators across five categories: conflict, community, governance, climate & water, and food. Ten of these indicators specifically relate to food systems, covering different dimensions of food security from availability to stability.

We faced an interesting challenge when working with agricultural data. We compared three different sources of data: national crop production statistics from FAOSTAT, remote sensing data on vegetation health from FAO-ASIS, and crop yield model simulations from the LPJmL model. Each source has its strengths and limitations. For instance, while FAOSTAT provides comprehensive national-level production data, affected by both climate and socio-economic impacts, it is made available with a two-year lag. Remote sensing offers near real-time data but currently only covers cereal crops. In large parts of Africa root crops like cassava and fruits and vegetables crops like beans and plantains form an important part of the diet. The LPJmL model has no lag and can be used for future risk projections, but needs improvement in capturing year-to-year variations in yield in Sub-Saharan Africa. Interestingly, only in a few regions across Africa did these data sources agree on the years with the lowest yields.

What We Learned 

Our pilot project revealed several crucial insights about predicting conflict risk:

 1. Definition Matters: How we define conflict significantly affects the model's performance. When looking at conflicts with 50 or more fatalities, the model struggled due to limited data points. However, it performed much better when including smaller-scale conflicts (those with one or more fatalities), partly because there was more data to learn from.

 2. History Matters: For smaller-scale conflicts, previous conflict in an area emerged as a dominant predictor. This aligns with existing research suggesting that conflict often breeds more conflict.

 3. Food Systems Matter: When we removed agricultural indicators from the model, it relied more heavily on existing conflict to predict future conflict. While this increased the model's recall (ability to identify actual conflicts), it reduced precision (more false positives). This suggests that food system indicators help refine our understanding of conflict risk.

 4. Contex Matters: The relationship between climate, food systems, and conflict varies significantly by region and context. What might trigger conflict in one area might not have the same effect in another.

Looking Forward

Our pilot project demonstrates both the potential and challenges of using machine learning to understand climate-related conflict risks. While CoPro shows promise in identifying areas of elevated conflict probability, its success depends heavily on how it's configured and what data is available.

The findings highlight the importance of having reliable, up-to-date agricultural data. We found that crop simulations could be improved by better accounting for local farming practices, such as planting dates and growing periods. This would be particularly valuable in Sub-Saharan Africa, where agricultural data is often limited.

Practical Implications

For organizations like FAO, this work provides a foundation for developing more sophisticated early warning systems that consider both climate and food system impacts. However, we emphasize that the model should be seen as one tool among many for understanding conflict risk. Its predictions should be interpreted alongside other analyses and local knowledge.

Our pilot project also underscores the need for better data collection and monitoring systems, particularly in vulnerable regions. Improving our ability to track and understand changes in food systems could significantly enhance our capacity to anticipate and prevent conflicts.

Conclusion

While this one-year pilot project has demonstrated the feasibility of incorporating food system indicators into conflict risk assessment, it also highlights the complexity of the challenge. As climate change continues to affect food systems globally, tools like CoPro could become increasingly valuable for organizations working to prevent and mitigate conflicts. The next steps will involve refining the model, improving data quality, and testing its applications in different contexts.

Our work for FAO represents an important step forward in understanding how climate change, food systems, and conflict intersect. It provides a foundation for future research and practical applications in this critical area of study.

* Uncharted Waters teamed up with Sophie de Bruin and Jannis Hoch, developers of the CoPro model.

The Conversation Africa covered Uncharted Waters’ work on food security and the impact improved water management can have in improving food production in Eastern Africa.

The application of our foodshed model confirms the potential of Africa to supply much of the increased demand for food within the continent. We looked at all food crops, including regionally important ones such as cassava, beans and millet. Countries in Eastern Africa play a pivotal role. Improved productivity due to local agricultural water management measures would reduce the need for more land elsewhere to grow crops.

What we propose requires large investments. Exploring these costs against benefits in the Rufiji basin in Tanzania we found that most water management measures would be cost effective, but only when considering the overall impact of water conservation on agriculture, hydropower production, and the riverine ecosystem.

Now also published in the French edition of The Conversation.

A growing and largely urban population in Sub-Saharan Africa will require more food, putting pressure on land and water resources. Mixed progress in making agriculture more efficient combined with climate change threatens the region’s food security. Local water management measures, such as better irrigation or water harvesting, can boost food production.

We found that ambitious yet realistic adoption of such measures increases food supply to cities and make the region as a whole self-sufficient. Better use of water would limit the expansion of land to grow food which helps to preserve biodiversity, combat climate change, and reduce land use-related conflicts.

The analysis was based on a combination of our global crop production computer simulations, converted into food energy supply, with the latest data on crop food demand. Applying an innovative foodshed ‘search and allocate’ algorithm using travel time open source data, we then delineated foodsheds of all major cities in Africa.

Heat waves

India is recovering from yet another summer marked by extreme heat waves. This follows a worrying trend; there has been a noticeable increase in heat wave days over India in the past decade, with severe Heat waves more extreme, longer and hotter in El-Niño years[1]. This year, a receding El Niño amplified temperatures already rising due to climate change.

The summer of 2024 began on a relatively normal note in the country, contradicting the trend of an early onset of summer heat in recent years. Active western disturbances and wet spells moderated temperatures in the country during February and March, with good growing conditions especially in the wheat belt of Northern India. April, however, brought severe heat to the southern and eastern parts of India, with record-breaking temperatures and heatwave conditions. A temperature of 47.2°C was recorded in Kalaikunda, West Bengal, on 30th April, some 10°C above normal. At the end of May, extreme heat conditions were widespread. The capital, Delhi, with approximately 30 million inhabitants, sweltered under temperatures of close to 50°C.

Chili production is impacted

In the village of Subhanpur, in the Jalna district in Maharashtra, a major state in central India, Pandurang Sheshrav Rajale, a local farmer, tries to keep his green chili plants well-watered in the early morning. The automatic weather station close to his field already shows a temperature of 35oC just an hour after sunrise. During the night, temperatures haven’t dropped below 30oC which made sleeping almost impossible. These conditions have now lasted for weeks with only some reprieve during occasional pre-monsoon rainstorms. The region where Mr. Rajale farms is famous for its chili production because of its large plantations and high output. The fertile soil and generally favorable climate help chili plants grow well, making it an important place for chili farming and supply. Also here, though, farmers are struggling.

Maharashtra has always experienced extreme heat and dry condition during the summer months. Its inland Marathwada and Vidarbha regions are considered heat wave hotspots. Unlike previous years, this years’ summer heat initially affected mainly the coastal regions rather than the region where Mr. Rajale farms. Mumbai, known for its high humidity, recorded its highest temperature at 39.7°C. Then, slowly, the heat shifted inland. Temperatures of around 40°C were observed in during the whole of April and May – high, but not unusual this time of the year. Only in the last week of May the region entered official heatwave conditions, according to the Indian Meteorological Department.

Beat the heat with water

Mr. Rajale is one of 25 farmers that grow chili that we interviewed at the end of May. Farmers were keen to share their experiences and reported many detrimental effects on chili crops due to heat stress. Yields start to get affected when temperatures exceed 35 oC to 40 oC, with reduced fruit quality and increased susceptibility to pests and diseases. With high air temperatures, the crop often does not germinate, soil temperatures increase above critical levels and soil borne diseases fester. Symptoms observed this year include wilting, flower drop, growth retardation, burning of leaves due to high temperature, delaying to flowering, reducing the percentage of flower to fruit ratio, drying of seedlings, yellowing of leaves, size reduction of chili fruits. These manifestations collectively impair crop growth and exacerbate the vulnerability of chili plants. Consequently, pest pressure from pests such as thrips, aphids, and whitefly and the chili leaf curl virus intensify, further compromising the health and vigor of the crops. In many cases this year, farmers had to replant seedlings repeatedly, affecting up to 20-25% of fields. Because of reduced growth and poor flowering, the overall yield of chili crops is expected to drop in the region.

In the drylands of central Maharashtra, where summer vegetable cultivation is a cornerstone of agricultural practice, heat stress poses a significant threat to agricultural productivity. We found a unanimous sentiment regarding the increase in heat stress, with the pervasiveness of heat stress over the past two to three years considered unprecedented. It was felt that not only the intensity of heatwaves and extreme temperatures, but also their duration has heightened the risks associated with crop production, making it more difficult for farmers to keep healthy yields and income.

High temperature in and of itself, however, is ultimately not the main problem. The availability of water for irrigation is essential during the dry, hot summer months. Whatever little rain that falls in summer evaporates quickly so regular water application ensures plants have enough moisture to stay healthy. But after an El Niño year with below average rainfall during the monsoon, water levels in wells and canals were low this year. Usually, in the monsoon, water infiltrates into the soil during heavy rainstorms and replenishes groundwater and fills up the wells which should last till the next monsoon season. The 2023 monsoon only brought 80% of the normal precipitation to the region[2] and in August 2023, most areas recorded less than a third of the usual amount. In the midst of the monsoon (also known as kharif) growing season, with plants in full development, farmers were forced to keep the emerging crop alive with irrigation; water that was supposed to be saved for summer. This lack of irrigation water compounded the impact of high temperatures this May. Keeping chili plants well-watered is crucial. During extreme heat, plants lose a lot of water through their leaves and dry soil heats up quickly. Irrigation requires substantial investment in time and resources which is why farmers prefer vegetables which usually fetch a high price. That is, when one has a harvest to sell.

Price volatility

The agricultural sector in semi-arid regions is not only a critical source of food but also a major part of the local economy. The erratic nature of heat stress disrupts traditional sowing patterns, leading to variations in harvest timing. This volatility extends to market dynamics, influencing chili prices due to fluctuations in both quality and quantity of the produce. The chilies that do grow might not be of good quality, with chilies smaller, mis-shaped, or less flavorful. Under stress, the mature green chili turns brown quicker. This affects their market value and can make them less appealing to buyers. Furthermore, burning and delayed fruit set results in a late harvesting, causing the harvested chili to arrive late in the market with farmers exposed to uncertain price fluctuations. The inconsistent supply due to unpredictable harvesting times leads to further price volatility, making it difficult for both producers and buyers to plan accordingly.

Heat also inflates production costs; a diminished crop resilience exacerbates the challenge of pest management, necessitating additional labor and resources. Farmers need to invest more in irrigation, or in heat-tolerant crop varieties, and other adaptive measures. This year, in addition to challenges with crop inputs, farmers met issues such as labor shortages for essential tasks such as spraying, weeding, and input application, and harvesting. Extreme temperatures made it challenging for workers to work efficiently, leading to delays in crucial farming activities. Additionally, the health risks associated with prolonged exposure to high temperatures result in increased absenteeism and higher turnover rates among farm laborers, further exacerbating the labor shortages. Those laborers that did come to work, demanded higher wages.

Reduced crop yields and higher costs translate to lower income for farmers, worsening poverty and food insecurity. Decreased agricultural output can lead to higher food prices, making it difficult for low-income families to afford basic necessities. The impacts of heat stress extend beyond crop yields, affecting the livelihoods of millions of farmers and their communities.

India is the world's largest producer, consumer and exporter of chilies. The chili cultivated in Jalna is primarily a green chili variety, most of which is exported to other countries, especially Bangladesh. This year, Bangladesh is experiencing a surge in chili prices due to heat wave conditions in combination with a seasonal effect of reduce supply at the start of the monsoon,  and is likely to import green chilies to stabilize its market. However, the timing and sources of these imports are still uncertain.

Limits to coping

Many farmers in the area boast decades-long experience cultivating chili crops. They innovate and have adopted a range of management practices to mitigate the impact of heat stress on chili production. These strategies include the deployment of shade nets, sprinkler irrigation, field fencing, soil application near the root zone, and timing agricultural activities to coincide with cooler periods of the day. Such measures have shown promise in safeguarding chili crops amidst escalating heat stress and drought conditions.

There are limits to coping, though, and these are often financial. More than 80% of interviewed farmers had adopted plastic mulching for their crops over the years. This reduces evaporation from soil and weeds, and thus irrigation water demand, and – in the case of white plastic - moderates soil temperatures in summer. With mulch, a drop of water lasts longer and grows more crops. But it is expensive and when wells fall dry, and yields drop below critical levels, the cost of plastic mulch becomes a liability. When farmers cannot apply or adhere to a proper irrigation schedule, the risk of seedling burning and unnecessary wilting due to the elevated temperature of the dry root zone increases again.

Farmers indicated that attempting to grow chili crops under heat stress with only 50% of the required irrigation water is highly risky, as the crops are likely to fail at a later stage. They questioned the necessity of taking such a high risk when they know the crops cannot thrive under these conditions. An estimated 25% of farmers who in previous years sowed chili as a cash crop, this year planted for only home consumption or even kept land fallow. Of those who did sow chili crops as per usual in February and March 2024, 80% reported that the crops are not bearing flowers due to factors such as moisture stress, heat stress, and evaporation loss. With little expectation of fully recapturing investments, a quarter has stopped applying the necessary nutrients, crop protection measures, and agronomic practices during the growing season.

Adapting

There is more that can be done. Drip irrigation delivers water directly to the plant roots, reducing water waste and ensuring consistent moisture levels in the soil. Though more expensive than sprinkler irrigation, an income from vegetables like chilies is likely to pay back the investment. Scheduling irrigation, by watering during the early morning or late evening, reduces evaporation losses and helps conserve soil moisture. As with drip irrigation, it requires farmers to have control over their water resources and/or supply. Short-term location specific weather forecasts, e.g., on the possibility of rain in the coming days, which could save an irrigation application or lead to more targeted pesticide or fertilizer application, can further improve crop-water management.

While plastic mulch has its benefits and has become the standard, applying organic mulch helps keep soil moisture, reduce soil temperature, and prevent weed growth. Incorporating compost and other organic materials into the soil can improve soil structure, water retention, and root health. A healthy soil improves a plant's resilience to heat stress. Sowing tall crops such as maize, sorghum, fodder grass around the main field can provide added shading as do trees and tree lines.

Specific measures such as foliar sprays containing calcium, potassium, or silicon can strengthen plant cells and improve heat tolerance while anti-transpirants can reduce water loss from plant leaves, helping to keep hydration during periods of high heat.

All these measures have been piloted, and form part of a farmers’ portfolio of options. But each measure also implies added investments or operational costs which need to be weighed against the expected benefits. In a highly variable climate, these will vary from year to year and with climate change accelerating, historic evidence and experience offers limited guidance. Farmers don’t have unlimited capacity to experiment. 

A better insight into medium-term weather conditions, weeks or a month ahead can help farmers make better management decisions. Notoriously difficult, with innovations in AI and weather forecasts, accuracy of prediction at these timescales is likely to increase though usefulness in practice still needs to be assessed. Better and timely information on growing conditions elsewhere, affecting future market conditions and prices, might also help farmers decide whether investments are worth it.

Take-away message

The findings underscore the pressing need for proactive measures to address the combined threat of water and heat stress on chili production. By prioritizing adaptation and innovation, stakeholders can mitigate adverse impacts, ensuring the sustainability and viability of chili cultivation in the face of mounting climatic pressures. As climate change continues to worsen environmental conditions, collaborative efforts among farmers, researchers, and policymakers are imperative to develop resilient agricultural systems capable of withstanding the rigors of a changing and variable climate.

Method: Colleagues from WOTR, a large watershed and livelihood development NGO based in Pune, interviewed 25 farmers, one farmer from each chili-growing village in the Bhokardan block, Jalna district, Maharashtra, India, at the end of May. The aim was to ensure that the sample reflects the conditions within this area where WOTR has been operational for over a decade supporting watershed development initiatives. The analysis of weather conditions was based on the observations of WOTRs automatic weather station network and from data from the European Centre for Medium-Range Weather Forecasts (ECMWF), with its ERA5 dataset providing historic and (near) real time data for a range of climatic indicators.

This story has also been published on WOTR’s website as a blog.

[1] Mahavedh AWS Grid, Govt. of Maharashtra

[2] M. Rajeevan, Rohini. P, S.N. Nair, S. Tirkey and T. Goswami (2023), Heat and Cold Waves in India; India Meteorological Department, Ministry of Earth Sciences, New Delhi

Uncharted Waters analysis contributes to Indian newspaper the Business Standard’s ‘Climate crisis cooks up costly meals: Why your 'thaali' could be on fire’ article.

Our analysis of 30 years of data revealed that India’s major wheat growing states often face significant crop loss if a cold snap is followed by a Spring heatwave - a likely scenario this year, given the combined impact of El Niño and climate change, but one that it might have avoided. While India is expecting a good wheat harvest in 2024 our analysis raises questions about India’s continued ability to produce enough wheat to meet the needs of a growing population in a changing climate.

India is the world’s second biggest wheat producer. While the majority of production is for domestic use, India is also an important source of wheat for countries in the region such as Nepal and Bangladesh, and was gradually increasing exports elsewhere - including UAE and Ethiopia - when it introduced an export ban in 2022 after a heatwave damaged harvests.

Uncharted Waters received the 2023 Palladium Challenge Fund 

Uncharted Waters is selected as one of two winners of this years’ Palladium Challenge Fund by the Kyeema foundation. We will pilot the use of localized, custom digital ‘twins’, to monitor the impact of emerging regional climate hazards, water and food security, or geo-political instabilities around the world. The goal is to provide policy makers and society with timely information about risks and to increase support for climate action.

Every year, Palladium brings in ideas from across its global network to source innovative responses to global challenges through the Palladium Challenge Fund. The theme for the 2023 Challenge was “Technology for Humanitarian Action” and the use of technologyto increase the impact of humanitarian crisis response. 

“Recent disasters have made it clear that the humanitarian community needs to adapt to the myriad pressures exacerbated by climate change and global instability,” explains Palladium co-CEO Sinéad Magill. “There are answers in the use of technology, and particularly in collaboration with local communities.” Globally, experts are working hard to stretch the available funding and explore how technology can enable faster and more impactful humanitarian action. This has included shifting thinking from response to preparation, early warning analysis, and anticipatory response activity.

The grant contribution of AUD 100.000 will help us pilot the design of customized digtial ‘twins’ with our knowledge and implementation partners Wageningen Environmental Research and WOTR, a large NGO in India. The first application will target India and surrounding countries in South Asia and link to our global platform. 

Cold wave

In the midst of an El Niño year¹, usually associated with above normal temperatures, India has been in the grip of a cold wave during most of January. Parts of northern India saw temperatures dropping below 0ᵒC. “Delhi has also been reeling under a cold wave. Some areas have recorded temperatures as low as 7ᵒC. The cold snap has been particularly hard for Delhi's homeless people who mostly sleep on pavements.”²  Extremely dense fog that has accompanied the cold temperatures affects people’s health and wellbeing, especially those with asthma or other respiratory diseases. Road, rail and air traffic has been disrupted and in several areas, schools have been closed.

Critical moment

Impacts of cooler winter temperatures on agriculture are rarely mentioned. Yet, an analysis of recent historic climate events suggests that cold conditions could increase risk to the wheat crop that is currently growing. Wheat provides an important source of income to farmers in the Indo-Gangetic Plain, also known as South Asia’s ‘bread basket’.  It is an important staple in the regional diet. Wheat can withstand low temperatures – it is grown in colder climates than India’s - but below average temperatures slow down development which in turn increases the risk of exposure to more extreme climate conditions later in the growing season.

Spring is near and inevitably the cold wave will end. A critical moment³ will be how quickly spring turns into summer. In India, climate change has already shortened the mild spring season. Summer, characterised by daytime temperatures of 30ᵒC or more, arrives now almost instantly in early March. With El Niño expected to continue for the next several months⁴, an early heat wave this year seems likely. Heat after cold has proved to be a key risk factor for Indian wheat production the past years, and one that farmers have difficulties coping with. 

Grown during the winter months, wheat is harvested by the end of spring, before high temperatures can affect grain filling and suppress yields. Often, it is planted directly after paddy (rice), the other important staple crop. If it is planted late, or grows slower because of below average temperatures, and summer heat arrives early, serious harm can be done to the crop. An analysis of below average yield in India’s northern state shows that such yield-impacting critical temperature swings have occurred several times in the past decades, lately in 2010 and 2022 (see Figure 2). A relatively cold February followed by an above average hot March had a larger impact on yields than years with consistently hotter or colder temperatures. Another heat-related risk that increases the longer the wheat crop is left on the land is hail, which shattered yields across northern India in 2015.

India, the new agri-trade pivot.

If such a swing occurs again this year, the impact will be felt beyond India. For decades, India was largely self-sufficient in its wheat and rice production, with population growth and demand matched by an equal increase in production, propelled by crop improvements and irrigation expansion, and governed by a system of state procurement, storage and price control. What happened in India, stayed in India. With yield increase and land expansion outpacing growth in national demand, India has, in recent years, become a relevant player on the global food commodity market, exporting its wheat within the region and beyond to countries like Bangladesh, the United Arab Emirates, Indonesia, South Korea and Ethiopia. Export volumes are still modest, at about a tenth of the biggest exporters and just below 3% of total global exports in 2021⁵. But its harvest in March and April falls in between those of Europe and North America in the north and Australia and Argentina to the south, which helps smooth global supply through the year.

Split trajectories of growth.

This increased global dependency on India’s agriculture system puts the focus on recent production increases and whether production will continue to grow at the same pace. In 2022, expectations of high exports to compensate the lack of grain coming out of Ukraine had to be revised downward quickly after the heat hit⁶. Wheat is still under export restrictions⁸. With a growing population and more affluent population, a ‘bumper harvest’ each year is merely a necessity. As other countries start to rely more on India’s exports, the world can ill afford anything less. A successful Indian harvest will ease pressure on prices globally. Sub national data reveals a split trajectory; yield growth in the major growing areas in the North west has levelled off, with below average yields in the past 15 years associated with major adverse weather events (Figure 3). Part of the historic increase here was fuelled by rapid groundwater development, especially in the northwestern states of Haryana and Punjab. Massive overexploitation has seen groundwater levels falling as a result, raising the costs of pumping both for farmers and for society with energy heavily subsidised. More towards the east of the country, farmers are catching up, increasing yields and production strongly over the past decade.  While the eastern part of the Ganges has more groundwater resources and a higher rate of replenishment, it is unclear for how long current growth trajectories can be sustained. If Uttar Pradesh, the largest wheat growing state, follows the trend of its smaller neighbours, year-on-year yield increases might soon start to slow down.

What to look out for?

This agricultural year has been eventful from the start in northern India. In mid-2023, high rainfall and localised floods impacted rice transplanting causing delays and replanting. This was followed by a long break monsoon period with below average precipitation. The extent to which farmers have been able to catch up fully is as of yet still unclear. It might have merely delayed rice harvesting, or reduced yields impacting farm income and the capacity of farmers to cope with adverse conditions during the wheat harvest.

Regarding wheat, one should keep a close eye on weather conditions in late February and early March. At present there is no reason to revise wheat production estimates downward. USDA projects a yield just one million ton higher than the record year of 2021/22⁹. An early heat wave, however, can have serious consequences. Extreme hail, such as  in 2015 can also be a late spoiler. India will grow a lot of wheat and diversification of production beyond the traditional main growing regions adds buffer against localized climate hazards. But despite good forecasts, lifting the export ban and large export volumes entering the market is far from guaranteed. 

Uncharted Waters tracks current food production, trade and storage using our bespoke digital twin of the global water & food system. We combine process-based crop production simulations with machine learning powered trade flow modelling to analyse, in real time, the linkages between climate, water, crop production and food availability. We can deliver reliable, real-time information, fast. 

Forward-looking statements are based on the Uncharted Waters’ beliefs, as well as assumptions made by, and information currently available to us. Because such statements are based on expectations as to future events and results and are not statements of fact, actual results may differ materially from those projected. All forward-looking statements reflect the beliefs and assumptions only as of the date such statements are made. We undertake no obligation to update forward-looking statements to reflect future events or circumstances. 

1. NOAA Climate Prediction Center, El Niño Southern Oscillation (ENSO) diagnostic discussion. Accessed 29 January 2024

2. Times of India. Cold wave continues in North India as mercury level dips

   BBC, Delhi temperature: North India shivers under the grip of cold wave

   Hindustan Times. Dense fog, cold wave grip Delhi-NCR; north India sees mercury plunge

   Hindustan Times. Extreme cold in Delhi, IMD says no respite soon; dense fog grips UP, Bihar: Weather updates

3.  Hassnain Shah et al., 2021. Climate risk to agriculture: A synthesis to define different types of critical moments. Climate Risk Management.

4. NOAA Climate Prediction Center, El Niño Southern Oscillation (ENSO) diagnostic discussion. Accessed 29 January 2024

5. United Nations’ Food and Agriculture Organisation: FAOSTAT Food Balance Sheet

6. USDA. India - Extreme Temperatures Scorch Indian Wheat Production. May 25 2022. 

7. Raupach et al. 2021. The effects of climate change on hailstorms. Nature reviews: earth and environment

8. IFPRI. Food and fertilizer export restrictions tracker.

9. USDA country summary & Wheat outlook: January 2024  

Improved understanding of conflicts in agrifood systems under climate change and climate variability

Uncharted Waters and partners will support the Food and Agriculture Organisation of the United Nations in Rome (FAO) in better understanding how climate impacts on agrifood systems and people may fuel conflicts, together with other factors, and how conflicts may exacerbate vulnerability to climate change and variability in rural regions across the world.

It is fairly well known how extreme weather events and long-term climate change can detrimentally affect agriculture and more broadly rural livelihoods and food security. Less well understood is how climate impacts may contribute, interacting with other important factors, to fuelling conflicts. Conflict early warning systems are emerging, but there is still a lack of well-grounded scalable tools to effectively address and forecast the interlinkages between climate, agriculture and conflict risk.

We will help FAO with the design and development of a tool to analyse and forecast conflict hotspots which relate directly or indirectly to climate impacts on agrifood systems, based on our bespoke digital ‘twin’ platform and the machine learning driven CoPro conflict model. Such an operational tool can inform investments and policies, and help FAO and development partners to design, fund, and implement local adaptation actions that take into account climate-security risks.

The initial scoping and design activities will run until November 2024.

For more information send an email to info@unchartedwaters.com