Six of Nine Boundaries Breached — What Transgressing Planetary Limits Actually Means

The first planetary boundaries paper, in 2009, found three boundaries transgressed: climate change, biodiversity loss and nitrogen flows. The 2015 update found four. In 2023 Richardson and colleagues quantified all nine for the first time and found six transgressed: climate change, biosphere integrity, land-system change, freshwater change, biogeochemical flows and novel entities. They noted that ocean acidification was close to its boundary and that aerosol loading exceeded it regionally. In 2025 the Planetary Health Check assessed ocean acidification as transgressed for the first time, and in 2026 it confirmed seven of nine, each at its worst recorded level.

Two kinds of change lie behind the rising count. Some are real: atmospheric CO₂ has risen from 351 ppm in 1988, when the global mean passed the 350 ppm boundary, to 426 ppm; ocean surface pH has fallen by about 0.1 units since the industrial era began, a 30–40% increase in acidity. Others reflect better science: novel entities and green water were quantified only recently, and several control variables have been refined. Either way, the assessments agree on direction — every transgressed boundary is moving further from safety.

The framework treats two boundaries as “core”: climate change and biosphere integrity, each of which could on its own push the Earth system into a new state. Both are in the high-risk zone. The 2026 report adds a warning about their interaction: about half of humanity’s CO₂ emissions have so far been absorbed by land and ocean, but the land carbon sink is showing signs of weakening, and if Amazon deforestation rises from about 17–18% to 22–28%, interactions between forest loss, drying and warming could expose around two-thirds of the remaining forest to degradation at only 1.5–1.9 °C of warming.

The wheel follows the convention of the planetary boundaries literature. The green circle is the safe operating space; each wedge is a control variable and extends outward according to how far it lies beyond its boundary, with the dashed red circle marking the start of the high-risk zone. Wedges inside the green circle are safe. The grey wedge is novel entities: transgressed, but unquantified. Two control variables span ranges — the extinction rate and the share of plant production appropriated by humans — and are shown with a lighter extension.

The gauges beneath put every control variable on a common scale: the boundary sits a quarter of the way along, the high-risk line halfway, and the far end marks three times the width of the zone of increasing risk beyond the boundary. Radiative forcing, nitrogen and phosphorus run off the end of the scale; ozone and global aerosols sit in the green; ocean acidification sits on the boundary line itself.

Each selection opens a decision card. It states the value and its distance from safety, explains the variable in plain words, says what is at stake, sizes the gap, names who decides and which levers they hold, lists existing agreements, and gives a judgement of reversibility. The cards are the report’s synthesis of the assessments and policy frameworks cited; they are a starting point for a decision brief, not a substitute for one.

Control variables are measured in incompatible units — parts per million, watts, teragrams, percentages of land. One way to compare them is to ask how much of the safe disturbance has been used: the change from the pre-industrial state divided by the change the boundary allows. A value of one means the boundary has been reached; two means twice the safe disturbance. The measure turns each boundary into something like a budget that has been overspent.

On that measure the extinction rate stands far beyond the others, at 11 to 111 times the safe disturbance, followed by blue-water disturbance (4.7×), human-caused radiative forcing (3.1×), mined phosphorus (2.9×), human appropriation of plant production (2.4–3.5×), nitrogen (2.7×) and CO₂ concentration (2.1×). Forest loss stands at 1.6 times the safe loss, and ocean acidification just above one. Ozone and global aerosols remain below their allowances.

The measure has limits. It depends on the pre-industrial reference, which for some variables is uncertain or has been revised, and it treats the distance from boundary to collapse as if it were linear, which it is not. Green water and novel entities are left out because the 2026 assessment does not state a comparable baseline. The point is the ordering: the pressures that come from food and land — nutrients, plant appropriation, habitat loss — are overspent as badly as the climate.

Budgets. Where a boundary limits a flow that accumulates, it can be expressed as a budget. Climate policy already works this way through carbon budgets, and the same logic applies to nitrogen and phosphorus: the safe levels of 62 teragrams of nitrogen and 6.2 of mined phosphorus a year can be divided among regions and farms, and nutrient surpluses tracked like emissions. The Kunming-Montreal Global Biodiversity Framework commits governments to reduce excess nutrients lost to the environment by at least half by 2030, a step towards such a budget.

Floors and ceilings. Some boundaries define what must be kept: a floor of 78% of potential forest cover, with biome-level floors for tropical, temperate and boreal forests, and an extinction rate low enough to preserve the genetic basis of life. Others define how much disturbance is tolerable: no more than about 13% of land with altered streamflow or soil moisture. These translate into land-use plans, protected and restored areas — the framework’s targets to conserve 30% of land and sea and restore 30% of degraded ecosystems by 2030 — and environmental-flow rules for rivers.

Bans and risk ratings. For novel entities the boundary is set at zero untested release, which translates into testing before release rather than after harm appears. And the zones themselves read like a credit rating: safe, increasing risk, high risk. The Stockholm Resilience Centre notes that the framework is moving from science into practice — Amsterdam’s city portrait links its impacts to planetary boundaries, New Zealand and several European countries have assessed their global footprints against them, companies are setting science-based targets for nature, and the Taskforce on Nature-related Financial Disclosures is pushing investors to report dependencies and impacts on planetary systems.

Because the boundaries share drivers, they also share decision-makers. Fossil-fuel combustion drives climate change, ocean acidification and much aerosol pollution. Food and land systems drive biosphere loss, land-system change, freshwater disturbance and the nitrogen and phosphorus flows. Chemical and material production drives novel entities. Finance, trade and planning cut across all three. A finance ministry that reforms fossil-fuel and fertiliser subsidies acts on climate, ocean acidification, aerosols and nutrient flows at once; an agriculture ministry that raises nutrient efficiency and protects forests acts on nutrients, land, freshwater and biodiversity together.

The matrix sets out, for each boundary, the decision unit it translates into, the actors who hold the main levers and the existing agreement or framework. The lens buttons highlight the boundaries in which a given actor is central. The aim is not completeness but traction: to show any decision-maker which planetary limits run through their own portfolio.

The Planetary Health Check authors stress that the next step is to move from fragmented efforts to integrated action — aligning climate, biodiversity and pollution policy, building accountability so that pledges translate into measurable reductions, and placing equity at the centre, because absolute biophysical limits raise questions about who has used the safe space and who must make room. The matrix is a tool for that alignment.

The pressures behind the transgressed boundaries are still growing. Global mean CO₂ rose from 351 ppm in 1988 to 426 ppm in 2025. Nitrogen fertiliser applied per hectare of cropland rose roughly ninefold between 1961 and 2023, from under 8 kilograms to 68. Forests covered 33.5% of the world’s land in 1990 and 31.9% in 2025. The Living Planet Index, which tracks monitored populations of vertebrate species, fell by an average of 73% between 1970 and 2022. None of these lines has yet turned.

The ozone layer shows that they can. When science linked chlorofluorocarbons to ozone loss, governments agreed the Montreal Protocol in 1987 and tightened it repeatedly. Consumption of ozone-depleting substances, weighted by their potency, fell by 99% from its 1989 peak of about 1.66 million tonnes. Stratospheric ozone is back within its boundary at about 286 Dobson units, and the 2022 scientific assessment projects a return to 1980 values by around 2040 for most of the world, 2045 over the Arctic and 2066 over Antarctica, if current policies are maintained. Aerosol loading has also improved as air-quality rules spread.

The lessons transfer. The ozone success combined a clear, measurable limit; a binding agreement with trade measures and finance for developing countries; available substitutes; and steady tightening as science improved. Each of the transgressed boundaries now has at least part of that combination — the Paris Agreement, the Kunming-Montreal framework, chemicals agreements — but few have all of it. Translating the boundaries into budgets, floors and bans, and assigning them to the actors who control the drivers, is how the remaining pieces get built.

Control-variable values, boundaries and high-risk lines are the best estimates published in the Planetary Health Check 2026 (Potsdam Institute for Climate Impact Research, summary report). Pre-industrial references come from the 2026 report where stated, from its boundary definitions where implied (the ozone boundary of 277.4 DU is a 5% decline, implying a reference of about 292 DU; the aragonite boundary of 2.86 is 80% of pre-industrial, implying about 3.575), and otherwise from Richardson et al. (2023, Table 1). The count of transgressed boundaries by year follows the framework’s published assessments as reported by the Stockholm Resilience Centre and PIK.

The gauge position is (current − boundary) ÷ (high-risk line − boundary). The multiple of safe disturbance is (current − pre-industrial) ÷ (boundary − pre-industrial). Both are summary devices: they assume linear scales that the Earth system does not have, and they depend on reference values, some of which have been revised between assessments. Changes between the 2023 and 2026 assessments mix real trends with redefinitions, so the dashboard does not present them as trends except for CO₂ and radiative forcing, whose definitions are unchanged.

Context series are from Our World in Data: NOAA global CO₂, UNEP ozone-depleting substance consumption, NASA ozone-hole area, the Living Planet Index (ZSL/WWF), FAO nitrogen fertiliser use and forest area. Decision cards and the actor matrix are the author’s synthesis of the assessments and policy frameworks cited, intended to support translation rather than to rank policies. Related reports: the pollution–climate feedback system and S-curves against the clock.