The food you love is starting to change. The texture of your rice. The way your everyday bread turns out. The price of olive oil. This isn't a distant story about climate change — it's a live problem on the factory floor and at the dinner table.
Around the world, rice, wheat, olives, and corn appear to be failing for different, unrelated reasons. But look at what's happening inside the plants themselves, and the same underlying process is at work. "Poor harvest due to extreme heat" tells only half the story. Inside a crop, distinct physiological processes — photosynthesis, pollination, grain filling — each start to break down at different temperature thresholds. This dysfunction cascades across the entire growing season, from spring flowering to autumn grain-filling — and in recent years, especially since the record heat of 2023 onward, this structure has become visible simultaneously in staple crops across the world.
The lead actor in photosynthesis is an enzyme called Rubisco, which fixes CO2. But the first thing to fail under heat isn't Rubisco itself — it's its switch, a helper enzyme called Rubisco activase. In many C3 crops, activation begins to decline around a leaf temperature of roughly 35°C (this varies widely by variety; some show effects from around 30°C, others tolerate temperatures close to 40°C). The engine (Rubisco) is fine, but the ignition switch (activase) fails first.
As temperatures climb further, Rubisco increasingly reacts with oxygen instead of CO2 by mistake — a process called photorespiration, which wastes energy by breaking down sugar the plant just made. Between these two mechanisms, photosynthetic efficiency in many C3 plants falls off in two compounding ways under heat (heat tolerance varies by crop and variety).
Today's high temperature may have been a "danger zone" not just for people, but for crops too. When temperatures at flowering exceed a critical threshold for seed fertility, pollen viability, germination, and pollen-tube elongation are impaired, and fertilization itself fails. Thresholds differ by crop, and wheat tends to fail at comparatively lower temperatures (the threshold shifts with variety, humidity, and duration of exposure). In rice, reports show that even 10 minutes of heat stress at 35°C can cause pollen sterility — meaning high heat during the flowering window can be lethal even if brief.
Even after successful pollination, the enzymes that assemble starch inside the seed — soluble starch synthase in wheat, granule-bound starch synthase in rice — are heat-sensitive. Wheat's optimal grain-filling temperature is 15–22°C; above 25°C, starch synthesis is gradually impaired through several mechanisms including reduced enzyme activity and altered gene expression, lowering yield and quality (it isn't a sudden cutoff at one temperature, but a stepwise decline as temperatures rise) — some studies report wheat yield falling 3–4% for every 1°C above the optimal grain-filling temperature. In rice, high-temperature ripening similarly reduces this enzyme's activity and contributes to chalky, immature grains.
Starch made by photosynthesis during the day is consumed as energy through respiration at night. As temperatures rise, this nighttime respiration increases, worsening the starch balance — the cause isn't reduced photosynthesis, but increased consumption. In field observations by Peng et al. (2004, PNAS) at the International Rice Research Institute (IRRI) in the Philippines, rice yield fell by about 10% for every 1°C increase in minimum night temperature (a single-field, single-variety result — a representative observation rather than a universal law). In regions where tropical nights (minimum temperature above 25°C) persist, the starch made during the day is routinely "used up" overnight.
It isn't only about what happens inside the plant. Climate change is also reshaping farming itself, through shifting pest ranges and the spread of disease.
Phases 1–3 described crops breaking down biochemically while stuck in place. But not all food production is fixed to one spot. Mobile species migrate when their "niche" shifts; season-dependent production sees its window narrow when its "timing" shifts. A spatial shift and a temporal shift — these two axes reveal a shared pattern of collapse across cacao, Pacific saury, nori, and kelp (in practice, both often occur together).
"More CO2 causes warming" — most people know this. But few can explain why. Understanding the physical mechanism makes what happened in 2023 click into place.
Climate change comes with several arguments that are "frequently cited but misleading." They share the same structure — the fact being cited is correct, but the conclusion drawn from it is not. Building on the physics from the previous section, here are four representative examples.
*Compared with the natural rate of CO2 rise at the end of the last glacial period (roughly 11,000–17,000 years ago) (NOAA Climate.gov, based on ice-core data from Lüthi et al., 2008)
Climate change doesn't unfold in a straight line. Natural variability between ocean and atmosphere layers additively overlaps on top of the long-term CO2 warming trend. This superposition is why record heat clusters in some years and appears to ease off temporarily in others. Natural variability isn't the cause of CO2 warming — it needs to be understood as a structure that amplifies or damps its swings.
Atmospheric circulation, ocean currents, and rainfall patterns are all driven by temperature differences as their driving force. Unlike a factory engine, this isn't a system designed to do work — the temperature difference itself is what drives the system. In systems like this, even a small temperature change alters the magnitude of the driving force itself.
The occurrence of El Niño, the timing of monsoons, the stability of the polar vortex — all of these rest on this balance of temperature differences. A few degrees doesn't mean "a bit warmer" — it means the force driving the system itself has changed.
The monotonic rise in OHC needs to be reread in this light. It's the cumulative record of surplus energy that human economic activity has kept injecting into environmental cycles — that's what the rise in OHC represents. Warming isn't a phenomenon of "the air getting hotter" — it's a process in which the driving force of environmental cycles itself is transforming.
A phenomenon in which tropical Pacific sea-surface temperature and atmospheric pressure fluctuate on a multi-year cycle.
During El Niño (the warm phase), the central and eastern Pacific warm, pushing global average temperature up by 0.1–0.2°C. La Niña (the cool phase) does the reverse.
The record heat of 2023 resulted from El Niño overlapping with the warming trend, and a similar combination is forecast for summer 2026 (based on seasonal forecasts, not a confirmed fact).
Interaction with warming (an active research question): Some research suggests that rising OHC may be shortening ENSO's cycle and widening its amplitude, but this remains a topic where researchers disagree and is not an established consensus. Trade winds, Kelvin waves, and atmosphere-ocean interaction remain essential to ENSO development, and any claim that it could develop thermodynamically "without trade winds" goes beyond current textbook understanding.
A phenomenon where the sea-surface temperature difference between the western and eastern Indian Ocean becomes anomalous. A positive IOD (warmer west) tends to bring heavy rain to India and East Africa, and drought to Indonesia and Australia. Effects are amplified when it co-occurs with ENSO. Summer 2023 saw a combination of El Niño plus a positive IOD, and a similar combination is forecast for summer 2026 by seasonal outlooks (a forecast, not a confirmed fact). It tends to bring heat and low rainfall to Japan.
Multi-decade sea-surface-temperature swings recurring separately in the Atlantic (AMO) and Pacific (PDO). Unlike ENSO, this is a long-period oscillation that a person can observe only a handful of times in a lifetime. Some researchers point to a possible overlap between the warm phase of the AMO (Atlantic Multidecadal Oscillation — sometimes called the AMV; whether it reflects internal variability or an externally forced signal remains debated among researchers) and the sharp acceleration of warming seen since the 1990s, though this is not an established explanation.
The record heat of 2023 was the result of "CO2 warming + El Niño + a positive IOD + reduced aerosol cooling" all overlapping at once. In some years, an overlapping cool phase of natural variability can make the temperature rise look temporarily gentler — but that doesn't mean warming has stopped. "It was cool this year, so we're fine" is the most dangerous misreading of all.
El Niño is a process that releases heat stored in the ocean into the atmosphere. But OHC doesn't decline even after that release — because the rate of heat flowing in from warming keeps outpacing the rate El Niño can release it. This may create a structure in which "each successive El Niño starts from a higher baseline." The sequence from 2023 to 2026 may suggest an early sign of this, though the number of cases is still too small for a settled conclusion.
Crossing +1.5°C isn't a single-year event. Even once an El Niño subsides and temperatures return to baseline, there's a real possibility that the baseline itself is settling in around the +1.5°C level. The cooling effect of the cool phase (La Niña) is shrinking in relative terms, and the central axis of the swing itself keeps rising.
The acceleration of the energy fill rate: What's pushing OHC upward is Earth's Energy Imbalance (EEI), driven by increasing downward long-wave radiation from greenhouse gases. EEI sharply accelerated after 1995 and reached 1.8 W/m² in 2023 — roughly double the IPCC AR6's best estimate (about 0.79 W/m², averaged over 2006–2018). That doesn't simply mean "the IPCC was wrong" — it largely reflects differences in observation period, estimation method, and the time window covered (the IPCC's estimate is a long-term average; 1.8 W/m² is a single-year value for 2023), and this distinction matters (Mauritsen et al., 2025, AGU Advances). More than 90% of this EEI is absorbed by the ocean. OHC isn't a static stockpile — it's better pictured as a capacitor continuously receiving an increasing charge. As this site frames it, ENSO is the oscillator drawing down from that capacitor — and as the charging energy increases, the charge rate, the charge ceiling, and the discharge energy all increase together.
JAMSTEC's latest Seasonal Watch, published May 20, 2026, states clearly that a strong El Niño is expected to develop in the tropical Pacific from June through August this year. In addition, the Indian Ocean Dipole index is expected to exceed 0.5°C in June, and a positive Indian Ocean Dipole event is expected to occur. When the Indian Ocean Dipole turns positive, the Tibetan high tends to extend northeast, bringing heat and low rainfall to Japan.
This simultaneous El Niño plus positive Indian Ocean Dipole is the same combination of signs seen in summer 2023 (note that both ENSO and the IOD vary in strength from event to event — this isn't "identical," just the same sign combination). 2023 saw global average temperature set an all-time record, and Japan experienced record heat and a prolonged hot late summer. The same forecast points to above-normal temperatures across much of the world, and Japan is no exception. The Japan Meteorological Agency has also issued a warm-season outlook stating that "summer 2026 temperatures will be high nationwide" (seasonal forecasts are probabilistic, not a settled outcome).
2023 alone doesn't prove the climate system shifted into a new phase. But because so many signals — observational records, the ocean, sea ice, and the gap from climate models — all appeared at once, this is the year many researchers now point to as the moment that prompted them to reassess the changes still to come.
Since 2023, global average temperature has been rising faster than climate models projected. Two main factors are cited: one is the weakening cooling effect of aerosols (airborne particles) from industrial emissions, the other is a decline in low-level clouds themselves.
What climate science considers especially serious isn't linear change, but the cascade of tipping points — thresholds beyond which a process becomes self-amplifying over the long term and extremely difficult to reverse on a human timescale.
In 2025 (May–September), 100,510 people nationwide were transported by ambulance for heatstroke — the highest since record-keeping began in 2008 (preliminary figures; 117 deaths). The confirmed number of heatstroke deaths in 2024 reached 2,160 for the year — approaching the same year's traffic-accident death toll of 2,663 (Fire and Disaster Management Agency, 2025, "Emergency Transport for Heatstroke, May–September Reiwa 7"; Ministry of Health, Labour and Welfare, "Monthly Report of Vital Statistics"). But a number like "35°C today" alone doesn't accurately convey how dangerous that heat has become — because 35°C at 80% humidity places a completely different load on the human body than 35°C at 30% humidity. The existing WBGT (heat index) is a technical metric that's somewhat complex for the public to grasp intuitively from a daily weather forecast — the public needs an index that shows "how dangerous is today" at a glance. Building on equivalent potential temperature — a meteorological measure combining temperature and humidity into a single figure for the atmosphere's energy content, which reflects the heat a person actually experiences — and converting it into an everyday 5-level scale to visualize rising heatstroke risk: that's the proposal behind the AEI (Atmospheric Energy Index).
*The values for each level are proposed thresholds designed by LFS based on perceived mugginess and heatstroke risk to the human body — they are not an official standard.
"Today's AEI is 4; the 1990s average for this time of year was 2" — this single line conveys warming not as an abstract future problem, but as "how my own body feels today." As the numbers accumulate day after day, the public can come to understand the progress of climate change through their own bodies.
How this differs from the existing heat index (WBGT): WBGT is a technical metric designed for safety management in workplaces and sports settings, whereas the AEI aims to give the public clear, everyday information while visualizing the long-term progression of warming — the two aren't competitors; they serve different roles.
The sections so far have covered impacts on photosynthesis and the harvest itself. But the climate crisis also reaches the infrastructure that supports food after it's harvested — the power grid. Heat waves push up electricity demand (air conditioning) while simultaneously reducing supply-side capacity. Thermal and nuclear power plants are forced to cut output due to rising cooling-water temperatures, falling river flows, cooling-capacity limits, and environmental regulations — squeezing the grid from both the demand and supply sides. The resulting blackouts don't just shut off household refrigerators — they also stop the cold chain (the low-temperature logistics network) for supermarkets and distribution hubs. "Can we harvest it?" is no longer the only question climate change affects — "can we store and transport it after harvest?" is becoming one too.
This isn't a sudden, one-off event. Rising temperatures reduce transmission capacity while increasing cooling demand; it's estimated that by 2040–2060, transmission capacity will fall by an average of 1.9–5.8%, while summer peak electricity demand rises by 4.2–15% (source: Bartos et al., 2016, Environmental Research Letters, 11(11), 114008, DOI: 10.1088/1748-9326/11/11/114008). The cases below, which occurred simultaneously across Europe and the US in summer 2026, can be seen as concrete examples of this forecast becoming reality.
Heat's effects reach beyond the power grid, into the logistics networks that move food itself. On the US Northeast Corridor, to avoid the risk of overhead catenary wires expanding, contracting, and being damaged by heat, speed restrictions kick in above 95°F air temperature and 128°F rail temperature — slowing high-speed trains that normally run at 150mph down to 80–100mph, with 18 trains cancelled in a single day on some days. The UK, France, Belgium, and other European countries also saw repeated heat-related rail delays and cancellations. These delay pressures fall not only on passengers but equally on the container freight and food logistics rail carries.
Climate-driven food crises don't happen only in the field. If the power grid supporting post-harvest storage, transport, and sale becomes vulnerable to heat waves, more food fails to reach the table even in a good harvest year. It isn't just agricultural policy anymore — energy infrastructure itself is becoming part of food security.
What these cases show is that large, centralized power grids can become simultaneously vulnerable to a single weather event like a heat wave. Maintaining the cold chain requires not just power plants, but mechanisms that can sustain power at the regional level too. One such approach is the microgrid — a distributed power source that completes generation, consumption, and management within a region, potentially maintaining local power supply even when broader supply-demand strain or equipment failures occur.
Climate change is starting to test not just "the ability to grow food," but "the ability to deliver food without letting it spoil."
The food crisis isn't confined to agriculture. Depleted water resources, large-scale climate displacement, tension between nuclear-armed states — climate change is setting off a geopolitical chain reaction. Some of its effects (glacier melt, drought, rising numbers of refugees, tension over water treaties) are already reality; the rest (nuclear risk, shifts in the global order) are also emerging as major security risks.
Events reported individually as "this year's drought" or "this year's flood" are cross-sections of a single structural shift. The sustained rise in OHC, the breakdown of water treaties, rising numbers of refugees — these aren't separate pieces of bad news, they are a structural deterioration that compounds and interacts. This needs to be treated as a structural risk that single-event disaster response alone can't capture.
Ocean warming isn't only driving changes to the Pacific's ENSO or weakening the Atlantic's AMOC. Another major change emerging from the same root cause — the buildup of surplus heat stored in the ocean (OHC) — is ocean deoxygenation. This long-term decline in dissolved oxygen in seawater is counted, alongside ocean warming and acidification, as one of the three major climate-driven stresses on marine ecosystems. This isn't a coastal, enclosed-bay problem — it's happening simultaneously across broad areas of the open ocean.
The ocean isn't just warming. The very mechanisms of "breathing" that sustain its circulation, oxygen, and ecosystems are starting to change.
Everything the crisis page has described converges into a single physical structure. The climate system has always had a handful of heat-discharge pathways — outlets for shedding excess heat. They are now closing in tandem.
Here, "heat-discharge pathway" is a convenient umbrella term covering not just the mechanism by which Earth radiates surplus energy to space (OLR), but also the mechanisms that temporarily suppress or sequester that buildup (heat transport to the deep ocean), the mechanisms that reduce how much heat is absorbed in the first place (albedo, aerosols), and the carbon-cycle mechanisms that affect the radiation budget via CO2 concentration (the biological pump). Their physical natures differ, but they share one thing: all of them are shrinking or weakening.
The problem isn't that heat is increasing. It's that the mechanisms for shedding that heat keep weakening.
Denmark is already acting. Brazil is already acting.
The problem isn't "no solutions exist" — it's that "solutions haven't spread."