The Long Landing

An omega block is not new. It is a high-pressure system that parks over a continent and traps hot air for a week. The new part is the starting point. Europe has warmed about 0.56°C per decade since the mid-1990s, while the world warmed about 0.27°C per decade. That is more than double. So the same old weather pattern now starts from a much hotter base.

That is why summers feel like they jumped. The physics changed smoothly, but the feeling is sharp. A weather pattern that once meant a bad week can now mean days 5 to 12 degrees above the old average, weeks near 42°C, and power grids stressed by air-conditioning that many homes were never built to need.

City leaders, building owners, schools, and employers need to plan for the heat we have now. That means heat alerts with clear triggers, cool roofs, changed work hours, and shade treated like basic infrastructure, not decoration.

Copernicus, European State of the Climate

The measured record has already shifted. Foster and Rahmstorf adjust for El Niño, volcanoes, and the solar cycle. They find a clear break around 2013-14. Since then, warming has run at 0.34-0.42°C per decade across all 5 major datasets. That is about double the 1970-2015 pace. At that speed, the 1.5°C line is crossed for good between 2026 and 2029. A single El Niño year touched it first, but that is the pattern of this century: each limit gets crossed twice, first as a spike, then as the new floor. This scenario assumes the speedup continues, past 0.5 degree per decade by the 2040s. The measured trend is the runway.

Part of the speedup is a harsh tradeoff. Cleaner ship fuel helped public health, but it also removed sulfur haze that reflected sunlight away from the ocean. Hansen and colleagues argue that this haze was hiding more warming than models assumed, and that Earth is more sensitive than the IPCC central view says. This timeline does not need to settle that fight. Even the lower reading makes 1.5°C a floor, not a one-time event.

National governments, city leaders, and businesses need to plan for a world above 1.5 degrees, not just hope to avoid it. Their job is to cut emissions hard while also preparing homes, water, food, and health systems for the hotter floor.

Foster & Rahmstorf 2026, GRL · Hansen et al. 2025, Environment 67(1) (attributed; contested)

Three important systems drink from the same rivers. Nuclear reactors need cooling water under a legal temperature. Data centers use water directly and also use far more through the electricity they need. The International Energy Agency (IEA) says global data-center water use is about 560 billion litres a year now, and could reach about 1,200 billion litres by 2030 in its base case. Farming uses far more than both. July 2026 showed the crash coming: heat shut down 3 French reactors - Golfech, Bugey, and Chooz - and put 7 more on notice to cut output, just when fields needed water most.

The EU already made data centers report water-use effectiveness under Delegated Regulation 2024/1364. Amsterdam has paused hyperscale permits before. Texas wrote shutoff rules into grid contracts. The examples exist. What is missing is the order: who gets the river first when there is not enough water.

Governments and water authorities need to write the order before the drought arrives: drinking water first, food second, machines third. If leaders wait until the river is already short, they are not making policy. They are managing a fight.

IEA, Energy and AI (2025), Box 5.4 · Euronews, 13 July 2026 — EDF reactor shutdowns · Delegated Regulation (EU) 2024/1364 · Texas SB6 (2025) — large-load remote-disconnect rules · JRC, Modeling water resilience in Europe (WEFE nexus)

The promise of offsets sounded clean: emit here, cancel it somewhere cheaper. In real life, the market sold anesthesia. "Neutral" on paper let emissions keep going, paid for by forests that were often not extra, not permanent, or sometimes not real. Investigations in the mid-2020s did not kill the idea. They killed its innocence.

The new accounts finish the job because scorecards that track physics have no place for a promise. A tonne is either in the atmosphere or it is not. Offsets get pushed down to what they honestly are: a small accounting tool for the leftover emissions that truly cannot be cut.

Lawmakers, regulators, and companies need to say the rule plainly: only real cuts count as real decarbonization. Offsets can cover the small leftover part that cannot be cut, but they cannot be used as a costume for business as usual.

The human climate niche is a quiet huge fact. For 6,000 years, people have mostly lived where the average yearly temperature sits in a narrow band. Farming, cities, and states were built for that band. Xu and colleagues put a number on losing it: about 1 billion people pushed outside the niche for each degree of warming. That is their central estimate, with worse outcomes possible.

Near 2 degrees, the first billion-plus people are already outside it, mostly in places that emitted least. The World Bank’s Groundswell planning figure is 216 million internal climate migrants by 2050, and it was modeled on a cooler path than this page uses. On this curve, 216 million is a floor, not a forecast.

National governments, city leaders, schools, and employers need to plan arrivals before people are forced to move. Housing, classrooms, language support, and jobs are policy choices. Waiting turns the same movement into an emergency.

Xu et al. 2020, PNAS — the human climate niche · World Bank, Groundswell (2021) · Lenton et al. 2023, Nature Sustainability — 2 billion outside the niche at 2.7°C total warming

Politics often talks about the middle climate scenario. Actuaries cannot safely do that. Insurance exists because the ugly tail is where ruin lives. By the mid-2020s, actuaries were saying this in their own language. The Institute and Faculty of Actuaries’ Planetary Solvency report says Earth-system stability should be treated like a balance sheet: test it against reasonable worst cases, and say the risk appetite out loud.

The report is careful. Its risk matrix is a severity scale, not a prediction. It defines what "catastrophic" would mean — including for deaths. But defining a band and assessing where we stand on it are two different acts: the report does the first, and declines the second because there is too little research. That discipline is the point: name the bands, place the current path on them, and notice that society is quietly accepting risks no company board would sign off.

In this scenario, these dashboards spread from insurers to pension funds to treasuries. Money that must survive to 2060 cannot use 2020 discount rates for 2050 physics. Reasonable worst cases go into every plan. On this page’s curve, the 2040s cross the Catastrophic band, and by mid-century the curve sits on the Extreme row: breakdown of critical Earth systems, state failure, and frequent large-scale mortality events, by definition.

IFoA & University of Exeter, Planetary Solvency (January 2025)

The Beauce is where French agriculture ran an experiment nobody wanted: thirsty maize on a falling aquifer under more and more drought orders. What worked there was not a miracle. It was a toolkit: soil-moisture probes so irrigation follows measurement, hard withdrawal caps with legal force, and a quieter change underneath it all: crops that need less water.

Sorghum, lentils, and chickpeas move north with the climate. Farmers do the math on varieties, not sentiment. The menu changes a little. The map changes more.

Cap withdrawals in law, then let the crop list follow the water. Do not do it backward. Crops can change in seasons. Aquifers refill in decades. That order is the policy.

SAGE Nappe de Beauce — reference volumes allocated per farm; arrêtés cadre with defined triggers

Adaptation money has a public-relations problem. It prevents disasters, and prevention photographs badly. A dike that holds is a non-event. A retrofit that prevents a heat death has no headline. Growth-era budgets starved this work because it does not look like it "creates" something new.

The useful frame is insurance. Nobody asks fire insurance for a profit. They ask it to be there when the fire comes. Civita-style dedicated funds show the premium can be paid in normal years, which are the only years when it can be paid calmly.

City councils, national treasuries, and public agencies need to set aside money for adaptation now. The premium only rises. Every hotter year makes the same protection more expensive.

The climate niche keeps moving, and many people are not allowed to follow it. Past 2.5 degrees, the rough math of 1 billion people per degree puts about 2 billion people outside the climate band their societies were built for. They are concentrated in South Asia, the Sahel, the Middle East, and the tropics - places that did the least to cause it. That is the central estimate, not the ceiling.

This does not create 1 migration. It creates a permanent condition: places that evacuate, partly return, and evacuate again; borders as infrastructure; citizenship as climate shelter. This becomes central politics everywhere, not because anyone chose it, but because it walks through the door.

National governments, city leaders, schools, and employers need to choose what they fund: arrival or walls. Both cost money. Arrival can buy workers, taxpayers, and allies. Walls mostly buy delay.

Xu et al. 2020, PNAS — ≈1 billion people per degree outside the niche · Lenton et al. 2023, Nature Sustainability — 2 billion outside the niche at 2.7°C total warming

Global food security rests on a quiet assumption: the handful of regions that grow most traded grain — the American Midwest, the Black Sea belt, the North China Plain, the Ganges basin — do not all have a bad year at once. Research began pricing that assumption in the 2010s: warming raises the odds of simultaneous breadbasket failure, and a wavier jet stream can lock heat domes over several of them in the same season.

In this scenario, the first synchronized failure lands in the 2040s. The sequence is fast and familiar from smaller rehearsals: exporters ban exports to protect home prices, importers bid against each other for what remains, and countries that buy most of their calories — much of North Africa and the Middle East — face bread prices their politics cannot survive. The 2010s already showed the miniature version.

Governments, grain boards, food companies, and aid agencies need real food reserves, varied suppliers, and crops chosen for the climate arriving now. Contracts are promises. Silos are food.

Gaupp et al. 2020, Nature Climate Change — rising risk of simultaneous breadbasket failure · Kornhuber et al. 2020 — amplified jet-stream waves and concurrent heat extremes in crop regions

The Seine is on no company balance sheet, even though many factories along it would stop without its cooling water. Roads, aquifers, pollinators, and wetlands are the same. They are physical preconditions for business, but accounting treats them like ghosts: used by everyone, owned by no ledger, invisible until they fail.

OCARA, Carbone 4’s open framework for checking a company’s real dependence on physical systems, moves from voluntary practice to regulation in this scenario. The river that cools the plant and the road that reaches it become line items, with owners, values, and risk assessments.

Regulators, accountants, and company leaders need to put physical dependencies in the books. Roads, rivers, aquifers, pollinators, and wetlands are not free background. If their absence can shut a business down, the risk belongs on the balance sheet.

OCARA — Carbone 4 (open climate-resilience audit framework)

More than 260 river basins cross national borders, and the treaties that share their water were mostly written in a wetter, calmer century. The stress test has already begun: in April 2025, India put the sixty-five-year-old Indus Waters Treaty — survivor of three wars — into abeyance, and a year later it was still suspended. Pakistan calls it the weaponization of water. Both countries have nuclear weapons.

In this scenario, the Indus is the template, not the exception. Upstream states finish dams that downstream states call existential — the Nile has lived this argument for a decade — and as flows shrink, every basin renegotiates from a position of thirst: the Mekong, the Tigris and Euphrates, the rivers of Central Asia. Some treaties adapt. Some simply stop being honored, one dry season at a time.

Foreign ministries, water agencies, and upstream and downstream governments need to agree on sharing rules while the river still flows. Real-time data and dispute channels matter most before the water is gone.

Indus Waters Treaty in abeyance since April 2025; still suspended into 2026 · UN — transboundary basins shared by two or more states (~260+)

South of Greenland sits one of the century’s most watched signs: a patch of ocean that cools while the world warms. The "warming hole" is real. Its meaning is debated. It looks like a sign of a slowing Atlantic overturning circulation, the conveyor that brings tropical heat toward Europe, but there are several possible causes.

Hansen and colleagues argue that polar melt is ahead of the models and that AMOC shutdown is likely within 20-30 years without stronger action. They say this is "in contradiction to conclusions of IPCC" and call it a point of no return because shutdown would lock in metres of sea-level rise. The IPCC central view says collapse this century is unlikely. This timeline does not settle the fight. It is built on the branch where the overturning holds.

Governments, climate agencies, and fossil-fuel regulators need to watch the Atlantic and cut emissions at the same time. Monitoring tells us how close the fork is. Cutting emissions is the only lever that can move the odds.

Hansen et al. 2025 (attributed; contested) · Keil et al. 2020, Nature Climate Change (multiple drivers)

This number comes from people paid to price ruin. The Institute and Faculty of Actuaries’ Planetary Solvency scale defines its Extreme band - 3 degrees or more by 2050 - with a mortality column that says more than 4 billion deaths. The report is careful: the matrix is a severity scale, not a forecast, and it says there is little research on mass death at these levels. Hold both facts together. Nobody has predicted 4 billion deaths. But the professional risk scale for this level of warming had to include a row where 4 billion deaths fit.

The mechanism is not 1 disaster. It is disasters stacking: deadly heat where one-third of humanity lives, harvest failure in several breadbaskets in the same year, water systems collapsing under demand, and conflicts growing in every crack. Each might be survivable alone. The band exists because they stop arriving alone.

This scenario’s curve reaches that band at mid-century. What stands between the definition and the body count is the rest of this page: water hierarchies, care commons, living buffers, and arrivals planned instead of suffered. That is what adaptation is for. It does not bend this curve. It decides what the curve costs in people.

Governments, city leaders, health systems, and adaptation agencies need to treat this as a life-protection job. Water rules, care systems, living buffers, and planned arrivals do not erase the danger. They decide how many people it costs.

IFoA & University of Exeter, Planetary Solvency (2025), Fig. 12 · Mora et al. 2017, Nature Climate Change · Xu et al. 2020, PNAS

Start with scale. The last glacial maximum, when Europe sat under kilometres of ice, was about 4 to 6 degrees below the pre-industrial baseline, depending on the reconstruction. Three degrees of warming is most of that distance in the other direction, moving about 100 times faster. In this scenario, 2 degrees came and went in the mid-2030s. Three arrives by mid-century, the top of the worst-case very likely band, on schedule.

Then the bill. The United States produced about one-fifth of all CO2 emitted since 1850. The EU-27 produced about another one-eighth. Together that is about one-third, and closer to one-half if counting only fossil carbon across all Europe. The rich world built itself on a loan against the world’s climate. At 2 degrees, when the worst impacts hit places with the lowest emissions, effort-sharing payments stop looking like charity and start looking like settlement.

Wealthy-country governments need to treat climate finance as a bill, not a favor. The emissions were spent long ago. The countries hit hardest now need money for safety, repair, and survival

Carbon Brief — cumulative emissions since 1850 (US ~20%, EU-27 ~12%) · Jean-Marc Jancovici (glacial-scale comparison, attributed); LGM reconstructions span ~4–6°C

Stratospheric aerosol injection is the cheap, fast, frightening option: spray reflective particles high in the atmosphere and the planet cools within a couple of years, the way it does after a large volcano. The price is within reach of a single ambitious country. The governance does not exist — scientists have spent years arguing over whether to even research it.

In this scenario, the decision is not made in a conference. After a heatwave kills at scale in the hot zone, a coalition of affected countries starts spraying and dares the world to object. The cooling is real, and so are the side effects: shifted rain patterns that some countries blame for their droughts, and a new kind of dependency — because stopping suddenly would snap temperatures upward in a few years, faster than anything on this page. The world becomes committed to maintaining a sunshade it never agreed to build.

The United Nations, national governments, scientists, and the public need to decide the rules before anyone grabs the thermostat. The question is not only whether the sunshade works. It is who controls it, whose rain changes, and what happens if it stops.

UNEP, One Atmosphere (2023) — solar radiation modification: risks and governance gap · Termination-shock literature — rapid rebound warming if deployment stops (attributed)

Today’s CO2 level last happened in the Pliocene, about 3 million years ago, when oceans stood 10 to 25 metres higher. That world had different orbits and ice, but it is the plainest clue for what this much carbon eventually asks of ice sheets. The process takes thousands of years. At 2 degrees, the long-term sea-level commitment already reaches metres.

By this era, the slow feedbacks are visible: Greenland’s margins and West Antarctica’s grounding lines. Coastal policy crosses a quiet line. The question changes from whether to retreat to what order to retreat in. Committed sea-level rise is the same in every branch of this timeline. The branches differ in metres and centuries, not direction.

Coastal governments, banks, planners, and port authorities need to decide the retreat order early. Say clearly which streets become future shoreline. Planned retreat is a project. Unplanned retreat is panic.

IPCC AR6 WG1 Ch. 9 — multi-millennial sea-level commitment (~2–6 m at 2°C)

In this scenario, atmospheric CO2 passes 600 parts per million. That is beyond the Pliocene and into conditions last seen tens of millions of years ago. Ice sheets respond to the level, not the trend, and their response takes thousands of years: Greenland’s edges, West Antarctica’s grounding lines, and coastlines moving for a long time.

Humanity spends this era past the analog: an atmosphere with no recent match, lived in with Holocene institutions, on a map that keeps moving too slowly for election cycles but fast enough for every port authority to plan around.

Coastal authorities, port leaders, national governments, and planners need to think in centuries, not election cycles. Like cathedral builders, they are making choices their grandchildren will live with.

CenCO₂PIP Consortium 2023, Science — Cenozoic CO₂ history (600+ ppm last seen tens of millions of years ago)

The wet-bulb limit is not a political number. It is the body’s limit. Past a certain mix of heat and humidity, a healthy person in shade with water can die in hours. Regions that get close to that line stop being places where people can work outdoors in summer. Then they stop being places where people can safely be at all.

At the warming this curve reaches, Mora and colleagues estimate that about three-quarters of humanity faces deadly heat more than 20 days a year by 2100. When this page was written, it was about one-third. That is their central estimate, and bad years can go far past it. This is not discomfort getting worse. It is the habitable map being redrawn, with the tropics drawing the first lines.

City leaders, employers, grid operators, and health systems need to design for heat the way colder countries once designed for winter. Night work, buried infrastructure, and cooling as a right become survival tools, not luxuries.

Mora et al. 2017, Nature Climate Change — global risk of deadly heat

The clearest Pliocene signal in reconstructions is a forested Arctic: trees where tundra used to be, in the same kind of CO2 world we restored. Polar amplification made the Arctic the fastest-warming place on Earth through the century. By the 2090s, forest walks into old tundra, finishing in the far north what the 2020s started.

The symmetry is strange: farmland turns back into forest in Europe while forest moves into tundra above the treeline. The green edges of the map move in 2 directions at once, one from abandonment and one from heat.

Arctic governments, Indigenous communities, scientists, and land managers need to steward the new frontier as it forms. The youngest ecosystems on Earth can become a conservation project or an unmanaged experiment. That choice is policy.

Brigham-Grette et al. 2013, Science — Pliocene warmth, forested Arctic