Module 50: Ecosystems, Climate & Sustainability

How living systems are organised, how Earth's climate works, how humans have destabilised both, and what a sustainable future actually requires.

Part A · ecology — how living systems are organised
What this module covers — 18 parts, one connected story
Ecology is the study of how organisms interact with each other and with the non-living environment. Everything in this module connects: the carbon in your breakfast cereal passed through the atmosphere, a soil fungus, a wheat plant, and a supply chain before it reached you. Understanding ecology means understanding that connection at every scale.
The levels of ecological organisation — from individual to biosphere
Population growth — carrying capacity and the logistic curve
Populations grow exponentially when resources are unlimited, but growth always decelerates as it approaches carrying capacity (K) — the maximum population an environment can sustain. The logistic growth model captures this S-shaped curve. Real populations overshoot and crash: locust plagues, algal blooms, and human civilisations exceeding their resource base all demonstrate the same pattern.
Carrying capacity pressure slider
Adjust population as a percentage of carrying capacity (K) and see what happens to growth rate and ecosystem stress.
0% K 150% K
Community interactions — click a type to explore
+/+
Mutualism
+/0
Commensalism
+/−
Parasitism
+/−
Predation
−/−
Competition
−/0
Amensalism
Part B · food webs and energy flow
Trophic levels — the 10% rule and why it matters
On average, only 10% of energy stored at one trophic level is available to the next. The rest is lost as heat through metabolism, movement, and cellular respiration. This is why a savanna can support millions of wildebeest but only a few hundred lions. The ecological efficiency principle means large carnivores are always energetically expensive to support, and why a planet of 8 billion humans eating high on the food chain is unsustainable.
Apex predators
~0.1% energy
Carnivores — ~1%
Herbivores — ~10%
Producers — 100% NPP captured
Net Primary Production of Earth's biosphere: ~120 billion tonnes of carbon per year. Humans appropriate roughly 25–40% of terrestrial NPP through agriculture, grazing, and land clearing.
Keystone species — ecosystem impact disproportionate to biomass
Robert Paine coined "keystone species" in 1969 after he removed sea stars (Pisaster ochraceus) from a tidal pool and watched mussels monopolise the entire zone. These species hold ecosystems together through predation, nutrient cycling, or habitat engineering.
Scale = relative ecosystem impact per unit biomass. 100 = most disproportionate effect documented.
Part C · biogeochemical cycles — how matter moves through the Earth system
The carbon cycle — the cycle that drives modern climate change
Carbon moves between four main reservoirs: the atmosphere (~870 GtC), the ocean (~38,000 GtC), land vegetation and soils (~2,600 GtC), and geological stores including fossil fuels (~10,000+ GtC). Natural fluxes are roughly balanced; the problem is that burning fossil fuels extracts carbon from the geological reservoir — which took hundreds of millions of years to accumulate — and releases it to the atmosphere in decades. The atmosphere is now at ~422 ppm CO2, higher than at any point in the last 3 million years.
ATMOSPHERE 870 GtC · 422 ppm CO₂ OCEAN 38,000 GtC LAND BIOSPHERE 2,600 GtC FOSSIL FUELS (geological) 10,000+ GtC photosynthesis respiration ocean uptake outgassing burning (+10 GtC/yr)
The five major cycles — click to explore
Part D · biomes — the major ecosystem types
12 major biomes — click any to explore
A biome is a large-scale ecosystem type defined primarily by climate (temperature and precipitation) and characterised by a distinctive plant community. The same biome type can appear on different continents with entirely different species — convergent evolution produces similar adaptations to similar climates. Biome boundaries are shifting as climate changes: the boreal forest's southern edge is retreating northward at 35–50 km per decade.
🌿
Tropical Rainforest
🦁
Savanna
🏜
Desert
🫒
Mediterranean
🍂
Temperate Forest
🌲
Boreal / Taiga
🧊
Tundra
🦢
Wetlands
💧
Freshwater
🐟
Marine
🪸
Coral Reefs
🐧
Polar
Part E · biodiversity — what it is and why it matters
Three levels of biodiversity
Genetic diversity
Variation within a species. The Irish potato famine (1845–52) killed ~1 million people because virtually all Irish potatoes were clones of a single variety — Lumper — with no genetic resistance to Phytophthora infestans. Genetic diversity is ecological insurance.
Species diversity
The variety of species in a community, measured by species richness (count) and evenness (distribution). The Amazon basin holds ~10% of all species on Earth despite covering 0.5% of its surface. The tropics concentrate diversity because of stable climate and high energy input.
Ecosystem diversity
The variety of habitat types in a region. Coastal zones where mangrove, reef, and seagrass habitats interdigitate support more species than any single habitat alone. Edge effects and habitat heterogeneity drive local species richness.
Current extinction rate vs background rate
Background rate: ~0.1–1 extinctions per million species-years. Current estimated rate: 100–1,000x background. The "sixth mass extinction" is the first caused by a single species.
Biodiversity hotspots — species richness vs threat level
Norman Myers defined hotspots in 1988: regions with at least 1,500 endemic vascular plant species AND having lost at least 70% of original habitat. 36 hotspots cover 2.5% of Earth's land surface but hold 43% of bird, mammal, reptile, and amphibian species.
Low threatExtreme threat
Lower endemic richnessHigher endemic richness
Part F · the Earth system — planetary boundaries and tipping points
Nine planetary boundaries — the safe operating space for humanity
Johan Rockström and colleagues published the planetary boundaries framework in 2009, updated in 2023. The framework identifies nine Earth system processes where human pressure risks triggering non-linear, potentially catastrophic change. As of 2023, six of the nine boundaries have been transgressed. The most alarming: biosphere integrity and novel entities (synthetic chemicals, plastics, radioactive materials) are both in high-risk zones.
Status of nine planetary boundaries (2023)
Safe
Zone of uncertainty
Transgressed
Safe (2): Ocean acidification, atmospheric aerosol loading
Uncertain (3): Climate change, stratospheric ozone, freshwater (blue water)
Transgressed (4+): Biosphere integrity, nitrogen/phosphorus flows, land-system change, novel entities, green water
Climate tipping points — small push, large irreversible shift
Part G · climate science — how Earth's climate works
The greenhouse effect — natural and enhanced
Without any greenhouse effect, Earth's average surface temperature would be around -18°C. The natural greenhouse effect raises it to +15°C — a difference of 33°C that makes liquid water, life, and human civilisation possible. The problem is not the greenhouse effect itself but its enhancement by human emissions, which has added roughly 2.3 W/m² of additional radiative forcing since 1750.
Greenhouse gas contributions to radiative forcing (W/m²)
CO2 contributes ~66% of enhanced forcing. Methane has 80x the warming power of CO2 over 20 years but a shorter atmospheric lifetime (~12 years vs CO2's centuries).
Feedbacks — amplifiers and dampeners of warming
Positive feedbacks (amplify)
Ice-albedo: melting white ice exposes dark ocean/land, absorbing more heat. Water vapour: warmer air holds more water vapour, the most potent greenhouse gas. Permafrost: thawing permafrost releases stored methane and CO2. Amazon dieback: drought-stressed forest releases rather than absorbs carbon.
Negative feedbacks (dampen)
Planck response: warmer objects radiate more energy, eventually restoring balance. Cloud feedbacks (partially): some cloud types reflect incoming solar radiation. Weathering: higher CO2 accelerates silicate weathering, drawing down atmospheric CO2 over geological time. The Planck response is the reason Earth doesn't experience runaway warming on human timescales.
Thermohaline circulation — the global ocean conveyor
Warm surface current (poleward) Cold deep return current N. Atlantic sinking Pacific upwelling THERMOHALINE CIRCULATION (AMOC) 1,000–2,000 year cycle · moves ~20 million m³/s
The AMOC moves roughly 20 million cubic metres of water per second — 100 times all Earth's rivers combined. It keeps northern Europe ~5–10°C warmer than it would otherwise be. Freshwater from melting Greenland is weakening it; some models show collapse risk above 1.5°C.
Part H · global warming — the human fingerprint
The multiple lines of evidence — why there is no scientific doubt
The evidence for human-caused warming is not a single temperature record but a convergence of independent datasets: satellite measurements, surface weather stations, ocean heat content, sea level measurements, ice cores, isotopic signatures of atmospheric CO2, and the "fingerprint" pattern of warming (lower stratosphere cooling while troposphere warms) which only occurs with greenhouse gas forcing, not solar variation.
What different warming levels mean — 1.5°C to 4°C
Sea level rise calculator
Current rate: ~3.7 mm/yr and accelerating. Under high-emission scenarios (SSP5-8.5), models project 0.6–1.0 m by 2100, with some ice-sheet instability models projecting up to 2 m. Each centimetre of rise threatens 6 million more people with annual flooding.
Part I · the carbon cycle in depth — emissions, sinks, and budgets
Global CO2 emissions by sector — where it comes from
Source: Global Carbon Project 2023. Total: ~37.4 GtCO2/yr. Land use change adds ~3.9 GtCO2/yr on top. Natural sinks (ocean + land) absorb roughly half; the remainder accumulates in the atmosphere.
Remaining carbon budget for 1.5°C
The IPCC AR6 estimates ~380 GtCO2 remained (as of Jan 2023) for a 50% chance of limiting warming to 1.5°C. At current emission rates (~40 GtCO2/yr), that budget runs out in under 10 years. Adjust the annual emission reduction rate below to see how long the budget lasts.
0% cut/yr 15% cut/yr
Carbon accounting — Scope 1, 2, and 3 emissions
Part J · energy systems — the transition away from fossil fuels
Global primary energy — who uses what (2023)
Oil 27%
Coal 23%
Gas 23%
Hydro 10%
Wind+Solar 8%
Nuclear 5%
Other 4%
Fossil fuels: ~73% of global primary energy. Renewables (all types) + nuclear: ~27%. Wind and solar grew from 0.2% in 2000 to ~8% in 2023 — the fastest energy transition in history by capacity additions, but still dwarfed by fossil fuel consumption.
Solar PV cost collapse — the most dramatic price decline in energy history
Solar module prices fell from ~$76/W in 1977 to ~$0.20/W in 2023 — a 99.7% reduction in 46 years. Each time cumulative installed capacity doubles, costs fall by ~20–30% (Swanson's Law, analogous to Moore's Law in semiconductors). This makes solar the cheapest electricity source in history in most of the world.
Energy sources — click to compare
Part K · climate policy and the politics of decarbonisation
The Paris Agreement architecture — pledges, gaps, and ambition
The 2015 Paris Agreement is the first legally binding international climate framework with universal participation (196 parties). Its core mechanism: each country submits Nationally Determined Contributions (NDCs) — voluntary pledges to cut emissions — which must be revised upward every 5 years. The fatal weakness: NDCs are not legally binding, and there is no enforcement mechanism. Current pledges, even if fully implemented, track to ~2.7°C by 2100. The "ambition gap" between pledges and the 1.5°C target is roughly 50 GtCO2e by 2030.
Carbon prices — where they need to be vs where they are
To achieve Paris targets, carbon prices need to reach $130–200/tCO2 by 2030 (IMF estimate). Most current prices are far below this. The EU ETS reached ~€100/tCO2 in 2023. Canada's federal price: C$65/tonne (2023), rising to C$170 by 2030. The US has no federal carbon price; some state-level schemes apply.
SymbolicEffective ($150+)
Net zero claims — credible vs greenwashing
Part L · biodiversity loss and conservation
Five drivers of biodiversity loss — HIPPO
E.O. Wilson's HIPPO framework. Habitat loss is far the dominant driver — responsible for ~80% of species at risk. Climate change is rising rapidly and is projected to overtake other drivers above 3°C.
Rewilding — what trophic restoration actually achieves
The Kunming-Montreal Global Biodiversity Framework (2022) — Agreed at COP15 in Montreal, 196 nations pledged to protect 30% of land and ocean by 2030 (the "30x30" target), up from ~17% land and ~8% ocean currently protected. Critics note that "protected" status without enforcement is largely paper protection: up to 67% of declared protected areas allow extractive activities inside them.
Part M · food systems — the intersection of everything
Agriculture's environmental footprint — the numbers that matter
Land use
50%
Half of Earth's habitable land is used for agriculture. 77% of that feeds livestock; livestock provides only 18% of global calories.
Freshwater
70%
70% of global freshwater withdrawal. Beef requires ~15,000 litres/kg; wheat ~1,300 litres/kg. Virtual water trade shapes global food security.
GHG emissions
26%
Agriculture, forestry, and land use contribute ~26% of global emissions. Livestock: ~14.5% alone, with beef and dairy accounting for 65% of that.
Food wasted
33%
One-third of all food produced is lost or wasted. In high-income countries, most waste happens at retail and consumer level. In low-income countries, it's mainly at storage and processing stages.
Diet carbon footprint calculator
Adjust how many times per week you eat beef to see the annual carbon footprint impact. Average beef portion: 200g, emitting ~5 kgCO2e.
0 times/week 14 times/week
Alternative proteins — where they stand
Part N · oceans
Ocean depth zones — life in the water column
0m 200m EPIPELAGIC — sunlit zone · most marine life 1,000m MESOPELAGIC — twilight zone · daily vertical migrators 4,000m BATHYPELAGIC — midnight zone · no light · giant squid 6,000m ABYSSAL — sediment plains · 60% of Earth's surface HADAL — trenches to 11,034m (Mariana) · amphipods, snailfish
Ocean threats — the interconnected crises
Part O · cities and the built environment
Cities — disproportionate importance in the climate challenge
Share of global population
57%
By 2050, projected to reach 68%. Urbanisation is accelerating fastest in sub-Saharan Africa and South/Southeast Asia.
Share of global CO2 emissions
70%
Cities consume 78% of global energy. The density of cities is actually an advantage: urban residents typically have lower per-capita footprints than suburban or rural residents in wealthy nations.
The urban heat island effect raises city temperatures 1–3°C above surrounding areas due to dark surfaces, waste heat, and reduced vegetation. Cities in tropical regions face compounding heat stress: by 2050, 1 billion urban residents could experience deadly heat conditions (>35°C wet-bulb) annually. Green roofs, urban trees, and reflective pavements can reduce heat island effects by up to 2°C.
Urban emissions breakdown
Buildings 40%
Transport 30%
Industry 20%
Waste 10%
Buildings are the largest urban emitter — mostly from heating, cooling, and hot water. Deep retrofit (insulation + heat pumps + electrification) can cut building emissions by 70–90%.
Part P · sustainable living — what individuals can actually do
High-impact vs low-impact actions — the evidence hierarchy
A 2017 study by Wynes & Nicholas in Environmental Research Letters found that the four highest-impact actions — having one fewer child, living car-free, avoiding transatlantic flights, and eating a plant-based diet — together save roughly 58.6 tCO2e/year. By contrast, recycling saves ~0.21 tCO2e/year, changing light bulbs ~0.1 tCO2e/year. The mismatch between what people think matters and what actually does is real and significant.
Annual CO2 savings (tCO2e/yr per action). Source: Wynes & Nicholas 2017, updated with Poore & Nemecek 2018 food data.
Individual action vs systemic change — the genuine debate
The case for individual action
Demand signals markets. Consumer choices have driven plant-based food from niche to mainstream in a decade. Individual action normalises behaviours and creates political permission for policy. Personal choices have direct emissions impact: a vegan diet saves ~1.5 tCO2e/year in wealthy nations. Moral consistency matters: it's hard to advocate for climate policy while flying constantly.
The case for systemic focus
The "personal carbon footprint" concept was popularised by BP in a 2004 advertising campaign to shift responsibility from corporations to individuals. Top 100 companies are responsible for 71% of global emissions. The poor have tiny footprints and bear the greatest harm. Political action — voting, advocacy, strikes — can achieve emissions reductions that dwarf any individual consumption change.
The honest answer: both matter, but in different ways. Individual high-impact choices (diet, flights, heating) are worth making. But treating individual consumption as the primary lever is a category error. The systems-change argument is correct that structural shifts dwarf personal choices. The personal-action argument is correct that individual choices have real effects and aren't meaningless.
Part Q · the circular economy and industrial ecology
Linear vs circular economy — the fundamental distinction
The linear economy extracts materials, makes products, and disposes of them. The circular economy keeps materials at their highest value for as long as possible: repair, reuse, remanufacture, recycling — in that order of preference. The Ellen MacArthur Foundation estimates that a circular economy could reduce global CO2 emissions by 45% by 2050, largely through reducing demand for new materials. The real barrier is not technical but economic: virgin materials are often cheaper than recycled ones because their full environmental costs are not internalised.
The recycling reality — what actually gets recycled
Global averages. Plastic recycling rate particularly misleading: much "recycled" plastic was exported to developing countries where it was often burned or landfilled. After China's 2018 National Sword policy banned plastic waste imports, Western recycling rates collapsed.
Degrowth vs green growth — the contested macroeconomics
Part R · environmental ethics and the bigger picture
Three ethical frameworks — who or what has moral standing?
Anthropocentrism
Nature has value only insofar as it serves human interests. Mainstream economics is anthropocentric: ecosystem services are valuable because humans benefit from clean water, stable climate, and pollination. Defended by many environmental economists who argue that "natural capital" accounting gives nature more practical protection than ethics ever could.
Biocentrism
All living organisms have intrinsic value regardless of their utility to humans. Peter Singer's extension of utilitarian ethics to animals, and Paul Taylor's "respect for nature" philosophy. Practically demanding: implies eating animals and experimenting on them without consent is ethically equivalent to doing so to humans. Most people hold inconsistent positions on this.
Ecocentrism
The whole ecosystem — soils, rivers, species, processes — has intrinsic value. Aldo Leopold's "Land Ethic" (1949): "A thing is right when it tends to preserve the integrity, stability, and beauty of the biotic community." Legal personhood for nature (Ecuador's Pachamama rights, New Zealand's Whanganui River) operationalises ecocentric ethics in law. The practical problem: rivers and ecosystems cannot speak for themselves in court.
The Anthropocene — when did humanity become a geological force?
The term was popularised by atmospheric chemist Paul Crutzen in 2000. The International Commission on Stratigraphy's Anthropocene Working Group voted in 2023 to define its start at 1952 — marking the global dispersal of nuclear fallout detectable in sediment layers worldwide. The "Great Acceleration" from 1950 onward shows near-vertical curves for every metric of human impact: CO2, population, species loss, concrete production, plastic production.
Hope and realism: The two traps are despair (the problem is too large, nothing matters) and complacency (technology will solve it, we don't need to act now). The evidence supports neither. Renewables costs have collapsed 90%+ in a decade — faster than almost any climate model predicted. Arctic sea ice loss is also faster than almost any model predicted. The direction is clear; the outcome is genuinely contingent on choices made in the next decade. The most evidence-based position is urgent, specific, and unsentimental.
Part S · Q&A
If the 10% rule means carnivores are energetically inefficient, how did large predators like lions and wolves evolve at all?
The 10% rule describes average trophic transfer efficiency across ecosystems, not a hard physical limit. Some systems achieve 15–20% efficiency, particularly aquatic ones. Large carnivores evolved because the prey they hunt is often too large, too fast, or too dangerous for smaller predators — they occupy a niche defined by their ability to handle large prey, not by energy optimisation. They also often eat the entire prey including bones and organs, capturing more energy than the rule implies. The deeper answer: evolution selects for reproductive success, not energy efficiency. If being a lion is a viable reproductive strategy given available prey density, lions evolve. The 10% rule explains why apex predators are always rare, not why they don't exist.
CO2 levels were much higher in Earth's deep past — during the Cambrian, they were 15–20 times today's levels. Why wasn't the planet uninhabitable then?
The sun was also dimmer in the distant past — roughly 4% dimmer per billion years going back. Higher CO2 compensated for lower solar output to maintain temperatures compatible with liquid water. The relevant question is not the absolute CO2 level but the rate of change: life adapts to stable conditions over millions of years, but rapid changes outpace adaptation. Today's CO2 rise from 280 to 422 ppm has occurred in 170 years — geologically instantaneous. The Cambrian transition to high CO2 took millions of years. Ecosystems, ocean chemistry, and species can adapt to a new equilibrium; they cannot adapt to the speed of change now occurring. The Permian mass extinction (~252 Mya) was caused by rapid CO2 release from Siberian Traps volcanism — and it killed 90% of marine species. That's the relevant deep-time analogue, not background CO2 levels.
If nuclear power is low-carbon, why don't climate scientists advocate for it more strongly?
Many climate scientists and economists do advocate for nuclear — the IEA, IPCC, and most energy models include significant nuclear in low-emission scenarios. The resistance comes primarily from cost and construction time. A new nuclear plant in a Western country now costs $8–12 billion and takes 10–20 years to build. Solar and wind are cheaper, faster, and require no waste management. The IPCC AR6 includes nuclear in mitigation pathways but notes its high costs and long lead times compared to renewables. The practical case for nuclear is strongest for existing plants (extending their life is very cheap per tonne of CO2 avoided) and for hard-to-decarbonise sectors like industrial heat and shipping. The case against is primarily economic, not safety-based: nuclear deaths per TWh are actually among the lowest of any energy source.
Is biodiversity loss really a crisis comparable to climate change, or is the "sixth mass extinction" overstated?
The biodiversity crisis is real, severe, and arguably harder to reverse than climate change. Extinct species cannot be recovered; a stable climate, in principle, can be restored over centuries once emissions stop. Current extinction rates are genuinely 100–1,000 times the background rate based on fossil evidence, though some researchers argue these estimates carry high uncertainty. The Living Planet Index (WWF) found a 69% average decline in monitored vertebrate populations since 1970 — not extinctions, but catastrophic abundance losses. The most important distinction from public debate: most attention goes to charismatic megafauna (tigers, rhinos), but the actual crisis is in insects, freshwater species, and soil organisms — less visible but ecologically foundational. A world with fewer tigers is aesthetically poorer; a world with 40% fewer insects, as documented in European entomological surveys, risks collapse of pollination and food chains.
Carbon offsets are widely used by companies claiming net zero. Are they legitimate?
The vast majority of existing carbon offsets do not deliver the claimed emissions reductions. A 2023 investigation by The Guardian, Zeit, and SourceMaterial found that over 90% of Verra's rainforest offset credits — the world's leading offset standard — were "phantom credits" with no real impact on deforestation. The core problems are additionality (would the forest have been saved anyway?), permanence (forests burn down), and measurement (baseline emissions are estimated, not measured). Well-designed offsets can work — the EU ETS allowances and some industrial process offsets have verified reductions. But voluntary market offsets used for "carbon neutral" product claims are almost universally unreliable. The honest use of offsets is as a temporary bridge while reducing emissions, not as a substitute for them. Using offsets to claim "net zero" while maintaining emissions is greenwashing by definition.
If renewables are now the cheapest electricity source, why hasn't the energy transition happened faster?
Cheapest new capacity is not the same as the cheapest to replace existing capacity. Fossil fuel infrastructure — power plants, pipelines, refineries — has been written down over decades and has near-zero marginal cost. Replacing working assets with new ones incurs capital costs even if the new source is cheaper on a levelised cost basis. Grid infrastructure, transmission lines, and storage all require massive investment. Permitting timelines for new renewables are astonishingly slow: in the UK, an offshore wind project averages 4–5 years of permitting before construction begins. Political opposition from fossil fuel industries has suppressed carbon pricing, which would accelerate transition. And roughly 70% of the energy system is not electricity — shipping, aviation, steel, cement, and heat are far harder to electrify and have seen almost no transition yet. The rate is accelerating but is not fast enough: the IEA's Net Zero by 2050 scenario requires tripling renewable capacity by 2030.
Wetlands are often described as crucial carbon stores. How much carbon do they actually hold, and what happens when they're drained?
Peatlands alone store roughly 550–650 GtC — about as much as the entire atmosphere. They cover only 3% of Earth's land surface but have been accumulating carbon for thousands of years under waterlogged, anaerobic conditions that prevent decomposition. When peatlands are drained for agriculture (palm oil, soy, cattle) or fuel extraction (e.g., Scottish and Irish peat burning), this carbon is released as CO2 and methane over years to decades. Indonesia's drained peatlands emit roughly 0.5–1 GtCO2e/year — comparable to Germany's total annual emissions. The UK's upland peat bogs, many degraded by overgrazing and drainage, now emit more carbon than they absorb. Restoration (rewetting) is possible and cost-effective: re-wetting one hectare of drained peat can sequester 2–5 tCO2e/year at a cost of £50–200/tonne — competitive with most mitigation options.
What is the strongest argument against taking dramatic action on climate change right now?
The strongest serious argument — not denial, but a genuine policy debate — is about optimal timing and distribution of costs. William Nordhaus (Nobel laureate) argued that gradual carbon pricing rising over decades is economically optimal because discounting future damages justifies lower near-term action, and because future generations will be wealthier and better equipped to adapt. A more defensible current critique is that rapid decarbonisation of the global energy system may require imposing large costs on developing nations that contributed negligibly to the problem — a climate justice concern. Some economists argue that adaptation investment delivers more near-term welfare per dollar than mitigation in the poorest countries. The counterarguments are compelling (discount rates are ethically contestable, tipping points make gradual approaches catastrophically risky, renewable costs now make deep decarbonisation cheap), but the distributional justice question about who bears transition costs is genuine and unresolved.