Article
Safe and just Earth system boundaries
https://doi.org/10.1038/s41586-023-06083-8 Johan Rockström1,2,3 ✉, Joyeeta Gupta4,5, Dahe Qin6,7,8, Steven J. Lade3,9,10 ✉, Jesse F. Abrams11, 
Lauren S. Andersen1, David I. Armstrong McKay3,11,12Received: 23 June 2022 , Xuemei Bai
10, Govindasamy Bala13, 
Stuart E. Bunn14, Daniel Ciobanu3, Fabrice DeClerck15,16, Kristie Ebi17, Lauren Gifford18, 
Accepted: 14 April 2023 Christopher Gordon19, Syezlin Hasan14, Norichika Kanie20, Timothy M. Lenton11,  
Published online: xx xx xxxx Sina Loriani
1, Diana M. Liverman18, Awaz Mohamed21, Nebojsa Nakicenovic22,  
David Obura23, Daniel Ospina9, Klaudia Prodani4, Crelis Rammelt4, Boris Sakschewski1, 
Open access Joeri Scholtens4, Ben Stewart-Koster14, Thejna Tharammal24, Detlef van Vuuren25,26, 
Peter H. Verburg27,28, Ricarda Winkelmann1,29 Check for updates , Caroline Zimm
22, Elena M. Bennett30,31, 
Stefan Bringezu32, Wendy Broadgate9, Pamela A. Green33, Lei Huang34, Lisa Jacobson9, 
Christopher Ndehedehe14,35, Simona Pedde9,36, Juan Rocha3,9, Marten Scheffer37,  
Lena Schulte-Uebbing25,38, Wim de Vries38, Cunde Xiao6,39, Chi Xu40, Xinwu Xu7,8, 
Noelia Zafra-Calvo41 & Xin Zhang42
The stability and resilience of the Earth system and human well-being are inseparably 
linked1–3, yet their interdependencies are generally under-recognized; consequently, 
they are often treated independently4,5. Here, we use modelling and literature 
assessment to quantify safe and just Earth system boundaries (ESBs) for climate, the 
biosphere, water and nutrient cycles, and aerosols at global and subglobal scales.  
We propose ESBs for maintaining the resilience and stability of the Earth system (safe 
ESBs) and minimizing exposure to significant harm to humans from Earth system 
change (a necessary but not sufficient condition for justice)4. The stricter of the safe  
or just boundaries sets the integrated safe and just ESB. Our findings show that justice 
considerations constrain the integrated ESBs more than safety considerations for 
climate and atmospheric aerosol loading. Seven of eight globally quantified safe and 
just ESBs and at least two regional safe and just ESBs in over half of global land area  
are already exceeded. We propose that our assessment provides a quantitative 
foundation for safeguarding the global commons for all people now and into the 
future.
Humanity is well into the Anthropocene6, the proposed new geologi- could lead to triggering tipping points that irreversibly destabilize 
cal epoch where human pressures have put the Earth system on a tra- the Earth system7,11,12. These changes are mostly driven by social and 
jectory moving rapidly away from the stable Holocene state of the economic systems run on unsustainable resource extraction and con-
past 12,000 years, which is the only state of the Earth system we have sumption. Contributions to Earth system change and the consequences 
evidence of being able to support the world as we know it7,8. These of its impacts vary greatly among social groups and countries. Given 
rapid changes to the Earth system undermine critical life-support these interdependencies between inclusive human development and 
systems1,9,10, with significant societal impacts already felt1,3, and they a stable and resilient Earth system1–3,13, an assessment of safe and just 
1Potsdam Institute for Climate Impact Research (PIK), Member of the Leibniz Association, Potsdam, Germany. 2Institute of Environmental Science and Geography, University of Potsdam, 
Potsdam, Germany. 3Stockholm Resilience Centre, Stockholm University, Stockholm, Sweden. 4Amsterdam Institute for Social Science Research, University of Amsterdam, Amsterdam, The 
Netherlands. 5IHE Delft Institute for Water Education, Delft, The Netherlands. 6State Key Laboratory of Cryospheric Science, Northwest Institute of Eco-Environment and Resources, Chinese 
Academy of Sciences, Lanzhou, China. 7China Meteorological Administration, Beijing, China. 8University of Chinese Academy of Sciences, Beijing, China. 9Future Earth Secretariat, Stockholm, 
Sweden. 10Fenner School of Environment & Society, Australian National University, Canberra, Australia. 11Global Systems Institute, University of Exeter, Exeter, UK. 12Georesilience Analytics, 
Leatherhead, UK. 13Center for Atmospheric and Oceanic Sciences, Indian Institute of Science, Bengaluru, India. 14Australian Rivers Institute, Griffith University, Brisbane, Australia. 15EAT, Oslo, 
Norway. 16Alliance of Bioversity International and CIAT of the CGIAR, Montpellier, France. 17Center for Health & the Global Environment, University of Washington, Seattle, WA, USA. 18School of 
Geography, Development and Environment, University of Arizona, Tucson, AZ, USA. 19Institute for Environment and Sanitation Studies, University of Ghana, Legon, Ghana. 20Graduate School of 
Media and Governance, Keio University, Fujisawa, Japan. 21Functional Forest Ecology, Universität Hamburg, Barsbüttel, Germany. 22International Institute for Applied Systems Analysis, 
Laxenburg, Austria. 23CORDIO East Africa, Mombasa, Kenya. 24Interdisciplinary Center for Water Research, Indian Institute of Science, Bengaluru, India. 25Copernicus Institute of Sustainable 
Development, Utrecht University, Utrecht, The Netherlands. 26PBL Netherlands Environmental Assessment Agency, The Hague, The Netherlands. 27Swiss Federal Institute for Forest, Snow and 
Landscape Research, Birmensdorf, Switzerland. 28Institute for Environmental Studies, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands. 29Institute of Physics and Astronomy, 
University of Potsdam, Potsdam, Germany. 30Bieler School of Environment, McGill University, Montreal, Canada. 31Department of Natural Resource Sciences, McGill University, Montreal, 
Canada. 32Center for Environmental Systems Research, Kassel University, Kassel, Germany. 33Environmental Sciences Initiative, Advanced Science Research Center at the Graduate Center, City 
University of New York, New York, NY, USA. 34National Climate Center, Beijing, China. 35School of Environment & Science, Griffith University, Nathan, Australia. 36Soil Geography and Landscape 
Group, Wageningen University & Research, Wageningen, The Netherlands. 37Department of Environmental Sciences, Wageningen University & Research, Wageningen, The Netherlands. 
38Environmental Systems Analysis Group, Wageningen University & Research, Wageningen, The Netherlands. 39State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing 
Normal University, Beijing, China. 40School of Life Sciences, Nanjing University, Nanjing, China. 41Basque Centre for Climate Change bc3, Scientific Campus of the University of the Basque 
Country, Biscay, Spain. 42Appalachian Laboratory, University of Maryland Center for Environmental Science, Frostburg, MD, USA. ✉e-mail: johan.rockstrom@pik-potsdam.de; steven.lade@
futureearth.org
Nature | www.nature.com | 1
Article
boundaries is required that accounts for Earth system resilience and 
human well-being in an integrated framework4,5. Box 1
We propose a set of safe and just Earth system boundaries (ESBs) 
for climate, the biosphere, fresh water, nutrients and air pollution at 
global and subglobal scales. These domains were chosen for the fol- The ‘3I’ justice criteria used to 
lowing reasons. They span the major components of the Earth system 
(atmosphere, hydrosphere, geosphere, biosphere and cryosphere) analyse safe ESBs
and their interlinked processes (carbon, water and nutrient cycles), 
the ‘global commons’14 that underpin the planet’s life-support sys- Further explanation is in Gupta et al.22. Discussion of the caveats 
tems and, thereby, human well-being on Earth; they have impacts on related to the justice approach applied in this paper is in Methods 
policy-relevant timescales; they are threatened by human activities; and Supplementary Information.
and they could affect Earth system stability and future development Interspecies justice and Earth system stability (I1)
globally. Our proposed ESBs are based on existing scholarship, expert Interspecies justice aims to protect humans, other species and 
judgement and widely shared norms, such as Agenda 2030. They are ecosystems, rejecting human exceptionalism. In many domains, 
meant as a transparent proposal for further debate and refinement by interspecies justice could be achieved by maintaining Earth system 
scholars and wider society. stability within safe ESBs.
First, we identify ‘safe’ boundaries at subglobal and global scales Intergenerational justice (I2a and I2b)
for “maintain[ing] and enhanc[ing] the stability and resilience of the Intergenerational justice examines relationships and obligations 
Earth system over time, thereby safeguarding its functions and abil- between generations, such as the legacy of greenhouse gas 
ity to support humans and all other living organisms”4. To determine emissions or ecosystem destruction for youth and future people. 
safe boundaries, we use assessments of tipping point risks among Achieving intergenerational justice requires recognizing the 
local and regional tipping elements, evidence on declines in Earth potential long-term consequences of short-term actions and 
system functions, analyses of historical variability and expert judge- associated trade-offs and synergies across time. We define two 
ment. We assess the uncertainty in and confidence of these ESBs. types of intergenerational justice: (between past and present; I2a) 
Tipping elements are those components or processes that regulate whether actions of past generations have minimized significant 
the functioning and state of the planet and that show evidence of harm to current generations and (between present and future; I2b) 
having thresholds at which small additional perturbations can trigger the responsibility of current generations to minimize significant 
self-reinforcing changes that undermine Earth system resilience15,16. harm to future generations.
We do not exclusively rely on tipping points for setting safe ESBs, how- Intragenerational justice: between countries, communities and 
ever, and the ESBs should not be interpreted as representing tipping individuals (I3)
points. As a reference state for human life support on Earth, we use Intragenerational justice includes relationships between present 
an interglacial Holocene-like Earth system functioning dominated individuals, between states (international), among people of 
by balancing feedbacks that cope with, buffer and dampen distur- different states (global) and between community members or 
bances.  Methods and Supplementary Information have details on citizens (communitarian or nationalist). Intersectional justice 
how safe boundaries are determined. considers multiple and overlapping social identities and categories 
Second, we use three criteria to assess whether adhering to the safe (for example, gender, race, age, class and health) that underpin 
ESBs could protect people from significant harm (Box 1): ‘interspecies inequality, vulnerability and the capacity to respond. Achieving 
justice and Earth system stability’ (I1)17; ‘intergenerational justice’18 intragenerational justice means minimizing significant harm 
between past and present generations (I2a) and present and future caused by one country to another, one community to another and 
generations (I2b); and ‘intragenerational justice’ (I3) between coun- one individual to another.
tries19, communities and individuals through an intersectional lens20. 
These criteria sit within a wider Earth system justice framework that 
goes beyond planetary and issue-related justice to take a multi-level 
transformative justice approach focusing on ends (boundaries and widespread exposure to significant harm, which will lead to greater 
access levels) and means21,22.  Methods and Supplementary Information impacts when vulnerable populations are exposed3. Methods and 
have more detailed discussions of the justice approach applied in this Supplementary Information have details on how just boundaries 
paper. We define harm as negative impacts on humans, communities are determined. The just (NSH) boundaries described here are nec-
and countries from Earth system change in addition to background essary but not sufficient conditions for Earth system justice, which 
rates. The most recent Intergovernmental Panel on Climate Change must also enable access to resources for all26 and distributional and 
(IPCC) report identifies ‘severe’ risks and ‘high’ reasons for concern procedural fairness22. A foundation that enables minimum access to 
when tens to hundreds of millions of people are exposed to changes water, food, energy and infrastructure for all humans alongside a safe 
in climate, such as increases in temperature and extreme events23. In and just (NSH) ESB ceiling of maximum allowed human pressure on 
this paper, we define significant harm as widespread severe existen- biophysical domains could constitute a safe and just ‘corridor’ over 
tial or irreversible negative impacts on countries, communities and time4,22 (Fig. 1).
individuals from Earth system change, such as loss of lives, livelihoods Our assessment builds upon and advances beyond previous research 
or incomes; displacement; loss of food, water or nutritional security; and science-based political consensus, such as the Planetary Bounda-
and chronic disease, injury or malnutrition (a glossary is in the Sup- ries (PBs) framework27, doughnut economics28 and the Sustainable 
plementary Methods). Development Goals29 in the following ways. (1) We define just ESBs for 
Third, we combine these justice criteria with historical analy- avoiding significant harm using the same units as the safe ESBs for the 
ses, international health standards, Earth system modelling and same domains and propose that actors use the stricter of the safe and 
expert judgement to quantify safe and just ESBs that minimize just boundaries to inform target setting. The PBs identify only safe bio-
human exposure to significant harm (no significant harm (NSH)) physical boundaries. The social goals related to access to or harm from 
from Earth system change. Minimizing significant harm is a corner- natural resources adopted in Agenda 2030, doughnut economics and 
stone of national and international law and corrective justice24,25. other approaches28,30–32 are not quantified in comparable units or exam-
We focus on assessing the levels of Earth system change leading to ine only the consequences of human activities on the Earth system, not 
2 | Nature | www.nature.com
Current Safe Just Safe and just align
Climate
Aerosols Functional
(subglobal) integrity
Natural
Phosphorus ecosystem
area
Acc ness fou
o
S t
c
afe nd jusa
ou
n
b
Ea mrth syste
Nitrogen Surface
water
Groundwater
Fig. 1 | Proposed safe and just (NSH) ESBs. Visualization of safe ESBs (dark red), boundary quantification for harm from nitrate in groundwater while noting that 
just (NSH) ESBs (blue), cases where safe and just (NSH) boundaries align (green) the just boundary must also incorporate safe considerations via eutrophication, 
and current global states (Earth icons). Radial axes are normalized to safe ESBs. leading to a more stringent safe and just boundary. Minimum access to water, 
Headline or central estimate global boundaries (Table 1) are plotted to support food, energy and infrastructure for all humans (dotted green line) could 
comparison with the current global state, but we emphasize that we have also constitute the foundation of a safe and just ‘corridor’ (green filled area), but we 
defined subglobal boundaries and multiple likelihood levels for many domains do not quantify this foundation here. Alternative visualizations are presented 
(Table 1). For aerosols, however, we display the subglobal boundaries to compare in Extended Data Fig. 1.
safe and just boundaries. For nitrogen, we plot with a dashed blue line the 
harm to humans from Earth system change. Articulating sociopolitical 
notions, such as Earth system justice, and converting their implications Climate
into biophysical units can enable a better understanding of the space We identify safe ESBs for warming (Fig. 1 and Table 1) based on minimiz-
within which humans can function. (2) We define global and subglobal ing likelihoods of triggering climate tipping elements; maintaining 
ESBs in most domains. The PBs’ emphasis on the global scale can be biosphere and cryosphere functions; and accounting for Holocene 
inappropriate for the assessment and management of domains such as (<0.5–1.0 °C) and previous interglacial (<1.5–2 °C) climate variability 
the biosphere33 and fresh water34–37. (3) We set boundaries at multiple (Supplementary Methods). Some climate tipping points, such as circu-
likelihood levels for Earth system states. (4) Tipping element assess- lation collapse or Amazon dieback, have high uncertainty or low con-
ments in climate, biosphere and other Earth system domains are key, fidence in their dynamics and potential warming thresholds16, but the 
although not exclusive, evidence for our ESBs. Recent PB assessments complementary palaeoclimate and biosphere analyses independently 
instead emphasize risks related to the departure from Holocene ranges support the safe climate ESB assessment. Cryosphere function includes 
of Earth system variability38. maintaining permafrost in the northern high latitudes, permanent polar 
ice sheets and mountain glaciers and minimizing sea ice loss. We find 
that global warming beyond 1.0 °C above pre-industrial levels, which 
Quantifying ESBs has already been exceeded9, carries a moderate likelihood of trigger-
For each Earth system domain, we first quantify safe boundaries for ing tipping elements, such as the collapse of the Greenland ice sheet 
maintaining Earth system resilience, with multiple levels of likelihood or localized abrupt thawing of the boreal permafrost16. One-degree 
reflecting uncertainty or variability in the exact position of the bound- Celsius global warming is consistent with the safe limit proposed in 
ary. Adhering to these safe boundaries implements our ‘interspecies 199039 and the PB of 350 ppm CO2(ref. 27). Above 1.5 °C or 2.0 °C warm-
justice and Earth system stability’ criterion (I1 in Box 1) and will safe- ing, the likelihood of triggering tipping points increases to high or 
guard future generations against significant harm from Earth system very high, respectively (high confidence in Extended Data Table 1). 
change (intergenerational justice; I2b in Box 1), but it may not avoid Biosphere damage and the risk of global carbon sinks becoming carbon 
significant harm to current generations, particularly vulnerable popula- sources, potentially triggering further climate feedbacks, increase 
tions (I2a and I3 in Box 1). Hence, (1) we propose that some boundaries substantially40. We conclude that stabilizing at or below a safe ESB of 
be made more stringent to protect present generations and ecosys- 1.5 °C warming avoids the most severe climate impacts on humans and 
tems; (2) we complement safe boundaries with local-level standards to other species, reinforcing the 1.5 °C guardrail set in the Paris Agreement 
protect present generations and ecosystems; and (3) if the boundary on Climate Change.
is likely to cause considerable difficulties for present generations, we Assessment of significant harm from climate change suggests the 
propose that it is complemented with policies that account for distribu- need for a stricter just (NSH) boundary. At 1.0 °C global warming, tens 
tive justice. We also assess the current state of the Earth system with of millions of people were exposed to wet bulb temperature extremes 
respect to each safe and just ESB. (Fig. 2), raising concerns of inter- and intragenerational justice. At 1.5 °C 
Nature | www.nature.com | 3
dation
rrid
d oraries
Article
Table 1 | Proposed safe and just (NSH) ESBs (visualized in Fig. 1)
Domain: state Relevant Earth Safe ESB subglobal Safe ESB globally Just (NSH) ESB Safe and just ESB Current global state
variable system change (local/regional) aggregated
Climate: global Climate tipping Global climate Likelihood of passing Exposure to additional 1.0 °C at high 1.2 °C
mean surface points; exceed boundary set to tipping points: low, significant harm: moderate, exposure to 
temperature interglacial avoid regional 0.5–1.0 °C; moderate, 0.5–1 °C; high, 1–1.5 °C; very significant harm
change since range; biosphere tipping points and >1.0 °C; high, >1.5 °C; high, >1.5 °C
pre-industrial functioning biome degradation very high, >2.0 °C
(1850–1900)
Biosphere: Loss of climate, Critical natural >50–60% natural Align with safe boundary >50–60% (upper 45–50% natural 
natural water, biodiversity ecosystems need ecosystem area plus ensure distributional end) depending on ecosystem area
ecosystem area NCP to be preserved or (depending on spatial justice distribution
restored distribution)
Biosphere: Loss of multiple >20–25% of each 100% of land area Align with safe boundary >20–25% of each One third (31–36%) of 
functional local NCP 1 km2 under (semi-) satisfies local boundary 1 km2 under (semi-) human-dominated land 
integrity natural vegetation; natural vegetation area satisfies ESB
>50% in vulnerable 
landscapes; at 
<10%, few NCP 
remain
Water: surface Collapse of <20% magnitude 100% of land area Align with safe plus World Regional and 66% of global land 
water flows freshwater monthly surface satisfies local boundary Health Organization and global safe ESBs area satisfies ESB 
ecosystems flow alteration (sums to 7,630 km3 United Nations Environment annually (3,553 km3 
per year global flow Programme quality per year global 
alteration budget) standards alterations)
Water: Collapse of Annual drawdown 100% of land area Align with safe plus World Safe ESB (and 53% of global land 
groundwater groundwater- does not exceed satisfies local boundary Health Organization and ensure recovery) area satisfies ESB 
levels dependent average annual (sums to 15,800 km3 per United Nations Environment (15,700 km3 per year 
ecosystems recharge year global drawdown) Programme quality annual drawdown)
standards
Green water38 Not assessed Monthly root-zone <10% of ice-free land Not assessed Not assessed 18%
(previous soil moisture area exceeds boundary
assessment) deviates from 
Holocene variability
Nutrient cycles: Surface water <2.5 (1–4) mg N l−1 Surplus, <61  Align with local safe plus Local ESBs; and Surplus, 119 Tg N per 
nitrogen and terrestrial in surface water; (35–84) Tg N per year; drinking water (<11.3 global surplus,  year; total input, 
ecosystem <5–20 kg N ha−1 per total input,  (10–11.3) mg NO3–N l−1; 57 (34–74) Tg N  232 Tg N per year
eutrophication year in terrestrial <143 (87–189) Tg N  globally, <117 (111–117) Tg N per year
ecosystems (biome per year per year) and any available 
dependent) air pollution (for example, 
NH3) standards
Nutrient cycles: Surface water <50–100 mg P per m3 Surplus, <4.5–9 Tg P per Align with local safe Local and global Surplus, ~10 Tg P per 
phosphorus eutrophication year; mined input,  boundary to avoid safe ESBs year; mined input, 
<16 (8–17) Tg P per year eutrophication ~17 Tg P per year
Atmosphere: Monsoon systems <0.25–0.50 AOD Annual mean Align with safe plus <15 μg per m3 PM2.5 0.05 annual mean 
aerosol loading interhemispheric AOD <15 μg per m3 mean annual plus regional and interhemispheric AOD 
difference: <0.15 PM2.5; other levels of global safe ESBs difference
exposure to significant harm 
in Supplementary Table 11
warming, more than 200 million people, disproportionately those vulnerability will be necessary. During the 2022 United Nations Climate 
already vulnerable, poor and marginalized (intragenerational injus- Change Conference (COP-27), developing countries indeed focused 
tice), could be exposed to unprecedented mean annual temperatures41, actively on issues of adaptation, loss and damage.
and more than 500 million could be exposed to long-term sea-level rise 
(Fig. 2 and Methods). These numbers of people harmed vastly exceed Biosphere
the widely accepted ‘leave no one behind’ principle29 and undermine For the biosphere, we identify safe ESBs for two complementary meas-
most of the Sustainable Development Goals. Moreover, past emissions ures of biodiversity: (1) the area of largely intact natural ecosystems 
have already led to significant harm, including extreme weather events, and (2) the functional integrity of all ecosystems, including urban and 
loss of habitat by Indigenous communities in the Arctic, loss of land agricultural ecosystems (Table 1). Maintaining areas of largely intact 
area by low-lying states and sea-level rise or reduced groundwater natural ecosystems is necessary for securing the Earth system functions 
recharge from changing glacial melt systems3. Irreversible impacts on which all humans, other species (I1 in Box 1) and Earth system stabil-
from cryosphere and biosphere tipping elements that are committed ity depend, including stocks and flows of carbon, water and nutrients 
by anthropogenic greenhouse gas emissions in the coming decades and halting species extinction (Earth system nature’s contribution 
but which unfold over centuries or millennia also threaten intergenera- to people (NCP) via Earth system functions). Based on climate, water 
tional justice (Supplementary Methods). We conclude that if exposure and species conservation model outcomes, we propose a safe ESB 
of tens of millions of people to significant harm is to be avoided, the of 50–60% (medium confidence in Extended Data Table 1) of global 
just (NSH) boundary should be set at or below 1.0 °C. Since returning land surface covered by largely intact natural areas to maintain Earth 
within this boundary may not be achievable in the foreseeable future, system NCP (Table 1 and Supplementary Methods). This range uses the 
adaptations and compensations to reduce sensitivity to harm and current area of natural land cover as a minimum value while indicating 
4 | Nature | www.nature.com
175
2,000
150
125
100
1,500 75
50
1.00 1.25 1.50 1.75 2.00
1,000
500
0
0 1 2 3
Global mean surface temperature change (°C)
Mean annual temperature Wet bulb temperature 
Sea-level rise (2100) Sea-level rise (multicentury) 
Fig. 2 | Exposure to significant harm from climate change at different  calculated for 2100 (blue solid) and multi-centennial (blue dashed; linear 
levels of warming. We examine the exposure of the 2010 global population to interpolation) responses to a given temperature stabilization by 2100, 
mean annual temperatures above 29 °C (purple; linear fit, P < 0.01), wet bulb representing near-term impacts and long-term equilibria, respectively. The 
temperatures of 35 °C for an average of at least 1 day per year (orange; quadratic inset shows the magnification of wet bulb temperature in the range 1–2 °C. 
fit, P < 0.01) and future sea-level rise (blue; linear interpolation). Sea-level rise is Shading indicates one s.e.
the need to restore largely intact natural areas. The exact safe bound- requiring conservation attention; it does not imply protection that 
ary depends strongly on the demand for specific ecological functions excludes human habitation and sustainable use.
(which in turn depend, for example, on the remaining carbon emis- Functional integrity is the capacity of urban, agricultural or other 
sions to be sequestered) and on the spatial distribution of the largely human-modified ecosystems to provide ecological functions and their 
intact natural area across ecoregions and ecosystems. Studies gener- contributions to people at landscape scale, complementing the Earth 
ally indicate that up to 60% of the terrestrial earth surface area may be system NCP provided by large-scale intact natural ecosystem areas. We 
needed, with some extending up to 80% (Supplementary Methods). analyse what minimum amount, quality and distance of natural habitat 
Natural ecosystem areas comparable with the 50–60% terrestrial ESB and seminatural habitat are needed to maintain local terrestrial NCP 
are needed in the ocean to maintain carbon sequestration and minimize provision, including pollination, pest and disease control, water quality 
additional marine species extinction42. Biome-scale boundaries may regulation, soil protection, natural hazards mitigation and recreation. 
be more stringent: for example, to protect tropical forest biomes due We identify that at least 20–25% diverse seminatural habitat including 
to their contribution to climate stability and moisture recycling. If native species in each square kilometre in human-modified lands is 
allocation and coordination of restoration efforts are less than opti- needed to support the provisioning of multiple local NCP48. The exact 
mal, the required minimum area will be larger. If these boundaries are amount and quality required differ based on landscape type, climate 
transgressed, tipping points involving loss of biome-scale functional and topography; the amount can range up to 50% in some landscapes 
integrity and associated NCP may be triggered, including increases in vulnerable to natural hazards, such as steep slopes or highly erod-
species extinction rates. ible soils. This boundary applies to fine scales, currently proposed 
Adherence to our proposed safe ESB for the area of largely intact as 1 km2, because NCP are not transferable (for example, erosion or 
natural ecosystems should minimize harm to future generations (I2b in landslide can only be avoided by natural cover on the same slope) and 
Box 1) by securing biosphere contributions to all life support through a are often provided or supported by non-mobile or limited mobility 
stable and resilient Earth system and localized NCP provided by largely species (for example, foraging ranges of pollinating or pest-regulating 
intact nature. However, achieving justice for current generations insects are limited to a few hundred metres). About two thirds of 
(I2a and I3 in Box 1) may require a stricter boundary because the safe ESB human-dominated land area (approximately 40% of total land area) has 
does not account for the current uneven distribution of largely intact insufficient functional integrity (Supplementary Methods), and large 
natural ecosystems needed to support local livelihoods43, especially in areas are showing symptoms of resilience loss49, requiring regenerative 
poor or Indigenous communities44,45. Some people and countries may practices to restore local and Earth system functions.
directly benefit from policies to maintain or increase natural ecosystem The safe boundary for functional integrity reduces future expo-
area46, while others may face opportunity costs47. Hence, to ensure just sure to significant harm (intergenerational justice). Loss of functional 
distribution of largely intact natural ecosystems, a just (NSH) bound- integrity in agricultural ecosystems and cities below the safe bound-
ary may need to be set at the upper end of the 50–60% safe range, as ary would reduce food productivity, ecosystem capacity to mitigate 
allocation will be less than optimal for achieving the functions the lower natural hazards, pollution and nutrient losses and increase reliance on 
boundary was optimized for. We emphasize that natural ecosystem area harmful pesticides and biocides and capacity to choose alternate land 
includes all largely intact natural areas and not only those currently uses (intragenerational justice). The dependence on these services is 
Nature | www.nature.com | 5
Exposed population (millions of people)
Article
often higher in regions with more vulnerable communities. Specific insecurity and (3) water quality. The regional surface and groundwater 
interventions that secure functional integrity are highly local and are ESBs are generally in the long-term interests of surrounding commu-
best implemented under local authority, knowledge and leadership50, nities, as they conserve future fresh water (intergenerational justice: 
with policy interventions often needed to ensure that marginalized I2b in Box 1). Where depleted aquifers have already caused significant 
groups are not further disempowered but are given the space to use environmental impacts66, groundwater extraction should urgently be 
their knowledge and approaches to participate in such processes51. reduced, and recharge areas should be protected to restore aquifers to 
safe levels (NSH to present generations: I2a and I3 in Box 1). Minimizing 
Water significant harm to current generations also requires the following. 
For fresh water, we propose two spatially defined safe ESBs based on (1) Accounting for multi-level distribution indicates the allocation of 
subglobal boundaries that can be aggregated to the global scale: (1) allowed alterations between communities, sectors or nations shar-
a flow alteration ESB for surface water and (2) a drawdown ESB for ing the water body, whether directly or indirectly via virtual water. 
groundwater (Table 1). Flow alteration in rivers is one of the key driv- This allocation is particularly challenging where the safe ESB requires 
ers of freshwater biodiversity loss52, leading to declines in freshwater drastic reductions in water use. (2) Minimizing exposure to significant 
biodiversity that outpace those of terrestrial and marine systems53 harm should account for water insecurity in different regions of the 
and in large-scale NCP, such as coastal and inland fisheries, on which world. For example, harm associated with poor water sanitation and 
millions of people depend54,55. Local-scale flow-ecology analyses are hygiene conditions disproportionately impacts the health of young 
often used to establish environmental flow needs to define safe levels children in low-income countries67, particularly in Sub-Saharan Africa 
of flow alteration for individual watersheds56. These local-scale assess- and South Asia68. (3) Minimizing exposure to significant harm implies 
ments could provide the basis for spatially explicit safe boundaries addressing surface water quality guidelines for human use69, not just 
but are absent across most of the world57. In their absence, we propose an allocation of water quantity. At a minimum, water needs to be safe 
that a presumptive subglobal safe ESB of 20% alteration (increase or for consumption and irrigation, meaning that acceptable standards 
decrease) of monthly surface water flows compared with the prevailing for faecal coliforms and salinity must be met. We align our just (NSH) 
natural flow regime be met in all rivers globally (medium confidence in ESBs for water with the safe ESBs while noting that adhering to the 
Extended Data Table 1). This ESB leaves 80% of flows unaltered to meet boundaries would considerably restrict current use and will require 
environmental needs58,59, assuming that required water quality stand- policies to ensure distributive justice.
ards are also met. The ESB is supported by empirical studies showing These proposed surface and groundwater ESBs are independent 
that flow alterations within 20% support native fish species and flow of green water stocks. Green water stocks are critical for maintaining 
alteration beyond this level strongly affects biodiversity and ecosystem the atmospheric water cycle, which regulates seasonal precipitation 
structure and function60,61 (Supplementary Methods has additional levels34; can support a significant proportion of global agricultural 
references supporting the use of this threshold). The global ESB for production70 with less impact on aquatic ecosystems than blue water 
surface water is that 100% of all land area meets the subglobal boundary use71; and are closely related to the biosphere ESBs. A recent assess-
by limiting alterations of flows by 20% in all rivers in the world. Meet- ment38 proposed a spatially explicit green water boundary to ensure 
ing the global ESB sums to a global alteration budget of 7,630 km3 per hydrological regulation of terrestrial ecosystems, climate and bio-
year (Supplementary Methods; with high confidence in Extended Data geochemical processes by defining a maximum allowed deviation 
Table 1). Globally aggregated river flow alterations are currently less (drying or wetting) of soil moisture levels from mid-Holocene condi-
than this figure; however, we are outside the global ESB because the tions. The state variable for green water is defined as the percentage 
subglobal safe ESB is only met for 66% of land area (Table 1) and less of ice-free land area that in any month has root-zone soil moisture 
than half of the global population (Supplementary Methods). These levels outside the 95th percentile of the local baseline variability. 
results are consistent with recent analyses of water scarcity, which The boundary value is set at 10%, corresponding to the median depar-
highlight the challenge of meeting environmental flow requirements ture level from mid-Holocene conditions. We include this green water 
to support ecosystem services, such as fisheries production, while boundary in our set of safe ESBs (Table 1), but we limit our inter- and 
ensuring there is sufficient water for human needs57,62. intragenerational justice analysis (I2 and I3 in Box 1) to surface and 
Groundwater aquifers contribute to base flows in many river systems ground blue water.
and directly sustain wetlands and terrestrial vegetation. Unsafe levels of 
groundwater extraction occur when drawdown exceeds replenishment Nutrients
rates, impacting groundwater-dependent ecosystems and in some We set safe ESBs for agricultural nitrogen (N) and phosphorus (P) 
instances, leading to land subsidence and irreversible aquifer loss12,63,64. surpluses for minimizing eutrophication of surface water and terres-
Given the temporal nature of groundwater recharge and discharge and trial ecosystems due to runoff, leaching and atmospheric N deposition 
a lack of widespread consistent data on historical aquifer levels, we via ammonia and nitrogen oxide emissions (Table 1). We propose safe 
propose that the safe ESB for annual groundwater drawdown for all global-scale ESBs of 61 (35–84) Tg N per year for agricultural nitrogen 
aquifers be the average annual recharge, with groundwater considered surplus72 and 4.5–9.0 Tg P per year for cropland soil phosphorus sur-
safe if drawdown is less than recharge. The subglobal safe ESB is met for plus73,74 (medium confidence in Extended Data Table 1). These ESBs are 
a given aquifer when local drawdown does not exceed average annual based on recent papers72,74 calculating subglobal and global agricultural 
recharge. The global ESB for groundwater is that the subglobal ESB is nutrient losses, surpluses and inputs from critical N and P concentra-
met for all aquifers around the world. For the 2003–2016 period, the tions in water and air beyond which eutrophication occurs (Methods, 
global sum of average annual recharge is approximately 16,000 km3 Table 1 and Supplementary Methods). These ESBs primarily relate to 
per year (Table 1 and Supplementary Methods; with high confidence agriculture, which accounts for approximately 90% of anthropogenic 
in Extended Data Table 1). The groundwater extraction that may safely N/P inputs to the Earth system72,75. Our ESBs are based on agricultural 
occur within this boundary naturally varies across the planet and, where surpluses and losses72,74, although for comparison with previous PB 
possible, should be defined based on local-scale monitoring, although quantifications (Supplementary Methods), we also provide corre-
broad trends can also be determined via satellite remote sensing65. We sponding global inputs assuming current N/P use efficiency. These 
estimate that we are currently outside the global ESB because ground- recent studies also account for non-agricultural sources, assuming 
water levels in 47% of basins are currently in decline (Table 1). they remain at current levels, and the redistribution of nutrients 
Our justice analysis of the safe ESBs for surface and groundwater from over-fertilized to under-fertilized regions (Supplementary 
highlights the challenges of (1) multi-level distribution, (2) water Methods).
6 | Nature | www.nature.com
Number of
boundaries
transgressed
0
1
2
3
4
5
6
7
Fig. 3 | Hotspots of current ESB transgressions. The number of subglobal elevation coastal zones (<5 m) exposed to sea-level rise as proxies for local 
climate (two local exposure boundaries), functional integrity, surface water, climate transgression while acknowledging that the impacts of climate change 
groundwater, nitrogen, phosphorus and aerosol safe and just ESBs currently are far more diverse. We also emphasize that exposure of a location does not 
transgressed by location. No more than seven of these eight metrics have their necessarily imply responsibility for causing or addressing these environmental 
ESBs transgressed in any one pixel. Since climate is a globally defined ESB,  impacts. We invite the reader to investigate the consequences of different 
we use wet bulb temperatures of over 35 °C for at least 1 day per year and low- boundary values using the code in the code availability information.
Elevated N and P concentrations cause harm through the conse- triggered by differences in sulfate AOD between the Northern and 
quences of eutrophication on ecosystems and their services, such as Southern Hemispheres81. Observational studies on the impacts of inter-
fishery collapse, toxic compounds released by algal blooms72,76 and the hemispheric AOD difference on the Indian monsoon are lacking, but 
health impacts of air pollution from ammonia-derived aerosols77. Harm observations based on past volcanic eruptions and climate modelling 
can also occur from drinking surface or groundwater with elevated studies show that an increased concentration of reflecting aerosols in 
nitrate concentrations78 but at a higher level than the safe N concen- one hemisphere leads to precipitation decreasing in the same hemi-
tration for surface water eutrophication. We therefore align the just sphere’s tropical monsoon regions while increasing in the opposite 
(NSH) ESBs for subglobal N and subglobal and global P with their safe hemisphere80,82,83. Observed changes in the South Asian monsoon 
boundaries, as human harm from nutrient cycle disruption is primarily have well-understood mechanisms (Supplementary Information) that 
driven by environmental degradation. Accounting for significant harm are consistent with the effects of interhemispheric AOD difference84. 
from groundwater nitrate tightens the global N boundary slightly to 57 The volcanic eruptions of El Chichon in the 1980s (AOD difference of 
(34–74) Tg N per year (Supplementary Methods). These ESBs should 0.07) and Katmai (AOD difference of 0.08) provide empirical exam-
be complemented by standards for local air and water pollution for N ples80, while model-simulated AOD differences of 0.1 and approxi-
and water pollution for P. Additional justice considerations include mately 0.2 lead to declining precipitation in tropical monsoon regions85. 
lack of access to N and P fertilizers, which can threaten food security Interhemispheric AOD difference and its impact on shifts in tropical 
especially for low-income communities and countries76, and extraction precipitation are sensitive to the aerosol particle size and the latitu-
of phosphate rock, which is a limited resource currently underpinning dinal and altitudinal distribution of reflecting aerosols86. Consider-
food production but exposes poor and marginalized communities to ing this and the range of these studies (approximately 0.05–0.20 of 
mining waste, destroyed land and human rights abuses76,79. additional AOD difference), we assess that these shifts may become 
disruptive if the interhemispheric AOD difference, currently approxi-
Aerosol pollution mately 0.0587 on average and approximately 0.1 in the boreal spring and 
For aerosols, we propose a safe ESB defined by the interhemispheric dif- summer87, exceeds 0.15 (low confidence in Extended Data Table 1) due 
ference in aerosol optical depth (AOD) (Table 1) based on evidence that to air pollution85 or geoengineering-related aerosol asymmetries81,85  
a rising North/South Hemisphere difference can trigger regional-scale (Supplementary Methods).
tipping points and cause substantial adverse effects on regional hydro- Significant harm to human health from exposure to aerosols, such 
logical cycles, in addition to the existing PB of 0.25–0.50 AOD based as particulate matter (PM), suggests a more stringent just (NSH) 
on regional considerations27. We consider AOD differences and their boundary based on local air pollution standards88. PM and other aero-
potential impacts arising from natural emissions, anthropogenic sols are associated with respiratory illnesses and premature deaths 
emissions and stratospheric aerosol injection (solar geoengineer- as well as heart problems and debilitating asthma89. We select a just 
ing). Observational data for the West African monsoon rainfall80 and (NSH) boundary of 15 μg per m3 mean annual exposure to PM2.5 to 
climate modelling studies for the Indian monsoon81 have identified avoid a high likelihood of significant harm from aerosols (Table 1 
potential shifts in the location of the Intertropical Convergence Zone and Supporting Information) based on World Health Organization 
Nature | www.nature.com | 7
Article
202188 guidelines (Table 1) and European Union and US Environ- on translation of the safe and just ESBs to actor fair shares99. Climate 
mental Protection Agency air quality standards90,91. Such local and is the only ESB that has a relatively well-established and implemented 
regional guidance is needed because PM2.5 characteristics, such as methodology
100,101, with methodologies for other domains under devel-
toxicity, are highly place and source specific. Eighty-five percent of opment101,102. We emphasize that our ESBs complement, not over-ride, 
the world population is currently exposed to PM2.5 concentrations environmental restrictions for specific local settings: for example, 
beyond this boundary92, and exposure to ambient PM2.5 is estimated stricter biosphere boundaries for carbon-dense ecosystems or targeted 
to cause 4.2 million deaths annually89, with vulnerable groups being conservation efforts for protecting endangered or emblematic spe-
affected disproportionately more while polluting less93. Air pollution cies. We also acknowledge that other actors may choose to implement 
scenarios based on globally successful stringent mitigation and pol- targets based on other likelihood levels than those we have highlighted 
lution control show reductions in affected populations, but areas of (Fig. 1 and Table 1): for example, a lower risk tolerance than the high 
high air pollution might remain94. A 15 μg per m3 PM2.5 concentration risk of passing tipping points associated with a 1.5 °C safe boundary.
translates95,96 to an AOD of approximately 0.17, indicating that the just We offer our ESBs as an integration of social and natural sciences for 
(NSH) boundary for aerosols is more stringent than the safe regional further refinement, in the spirit that the PBs were proposed over a decade 
boundary (0.25–0.50) (Table 1). ago103. Seven of the eight globally quantified ESBs have been crossed 
and at least two local ESBs in much of the world have been crossed, 
Novel entities and other pollutants putting human livelihoods for current and future generations at risk. 
We acknowledge the risks to Earth system stability and human well- Nothing less than a just global transformation across all ESBs is required 
being from other air and water pollutants, for which there are already to ensure human well-being. Such transformations must be systemic 
well-accepted guidelines88, and the emerging threats from novel enti- across energy, food, urban and other sectors, addressing the economic, 
ties, new forms of existing substances and modified life forms that technological, political and other drivers of Earth system change, and 
are geologically or evolutionarily novel and could have large-scale ensure access for the poor through reductions and reallocation of 
unwanted geophysical or biological impacts on the Earth system27,97. resource use. All evidence suggests this will not be a linear journey; it 
Evidence on the diverse risk potentials of novel entities, such as micro- requires a leap in our understanding of how justice, economics, technol-
plastics, ‘forever chemicals’, antibiotics, radioactive waste, heavy met- ogy and global cooperation can be furthered in the service of a safe and 
als or other emerging contaminants, for Earth system function and just future.
human health and food security is increasing, but knowledge gaps 
on the scale and scope of potential impacts remain98. Persson et al.97 
reported that humanity has crossed the PB for novel entities, although Online content
data limitations and quantification are challenging even for the known Any methods, additional references, Nature Portfolio reporting summa-
novel entities. The differentiated impacts of novel entities already ries, source data, extended data, supplementary information, acknowl-
witnessed today across different populations and the long lifetimes of edgements, peer review information; details of author contributions 
these substances raise clear intragenerational and intergenerational and competing interests; and statements of data and code availability 
justice concerns97,98. are available at https://doi.org/10.1038/s41586-023-06083-8.
Current state 1. IPBES. Global assessment report on biodiversity and ecosystem services of the 
Seven of the eight global-scale safe and just ESBs that we quantified Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. 
have already been crossed (Fig. 1 and Table 1). Transgression of ESBs is Zenodo https://doi.org/10.5281/zenodo.5657041 (2019).
2. Folke, C. et al. Our future in the Anthropocene biosphere. Ambio 50, 834–869 (2021).
spatially widespread, with two or more safe and just ESBs transgressed 3. IPCC Climate Change 2022: Impacts, Adaptation, and Vulnerability (eds Pörtner, H.-O. 
for 52% of the world’s land surface, affecting 86% of the global popula- et al.) (Cambridge Univ. Press, 2022).
tion (Fig. 3). Some communities experience many ESB transgressions, 4. Rockström, J. et al. Identifying a safe and just corridor for people and the planet. Earth’s Future 9, e2020EF001866 (2021).
with four or more ESBs transgressed for 28% of global population but 5. Rockström, J. et al. Stockholm to Stockholm: achieving a safe Earth requires goals that 
only 5% of global land surface (Fig. 3). Spatial hotspot transgressions incorporate a just approach. One Earth 4, 1209–1211 (2021).
are therefore concentrated in regions of higher population density, 6. Zalasiewicz, J. et al. The Working Group on the Anthropocene: summary of evidence and interim recommendations. Anthropocene 19, 55–60 (2017).
raising major intragenerational justice concerns. 7. Steffen, W. et al. Trajectories of the Earth system in the Anthropocene. Proc. Natl Acad. 
Sci. USA 115, 8252–8259 (2018).
8. Xu, C., Kohler, T. A., Lenton, T. M., Svenning, J.-C. & Scheffer, M. Future of the human 
Toward a safe and just future climate niche. Proc. Natl Acad. Sci. USA 117, 11350–11355 (2020).9. IPCC Climate Change 2021: The Physical Science Basis (eds Masson-Delmotte, V. et al.) 
We defined and quantified safe and just (NSH) ESBs for sustaining (Cambridge Univ. Press, 2021).
the global commons that regulate the state of the planet, protect 10. UNEP Global Environment Outlook—GEO-6: Healthy Planet, Healthy People (Cambridge Univ. Press, 2019); https://doi.org/10.1017/9781108627146.
other species, generate NCP, reduce significant harm to humans and 11. Lenton, T. M. et al. Climate tipping points—too risky to bet against. Nature 575, 592–595 
support inclusive human development (Fig. 1 and Table 1). Because (2019).
exceeding safe boundaries results in widespread significant harm, our 12. UNEP Global Environment Outlook—GEO-6: Technical Summary (Cambridge Univ. Press, 2021); https://wedocs.unep.org/20.500.11822/32024.
just and safe ESBs align for surface water, groundwater, functional 13. Biermann, F., Dirth, E. & Kalfagianni, A. Planetary justice as a challenge for earth system 
integrity, natural ecosystem area, phosphorus and nitrogen. Meeting governance: editorial. Earth System Governance 6, 100085 (2020).
these boundaries without transformation, however, could signifi- 14. Nakicenovic, N., Rockström, J., Gaffney, O. & Zimm, C. Global Commons in the Anthropocene: World Development on a Stable and Resilient Planet. IIASA Working Paper 
cantly harm current generations. In two cases, aerosols and climate, (IIASA, 2016); http://pure.iiasa.ac.at/14003/.
the just boundaries are more stringent than the safe boundaries, 15. Lenton, T. M. et al. Tipping elements in the Earth’s climate system. Proc. Natl Acad. Sci. 
which indicates that people experience significant harm before that USA 105, 1786–1793 (2008).16. Armstrong McKay, D. I. et al. Exceeding 1.5 °C global warming could trigger multiple 
Earth system domain is destabilized. climate tipping points. Science 377, eabn7950 (2022).
We identified subglobal ESBs, which, in many domains, are the 17. Burke, A. & Fishel, S. in Non-Human Nature in World Politics: Theory and Practice (eds 
Pereira, J. C. & Saramago, A.) 33–52 (Springer International Publishing, 2020).
relevant scale for action to avoid loss of Earth system stability and 18. Meyer, L. Intergenerational justice. In The Stanford Encyclopedia of Philosophy (ed. Zalta, 
minimize exposure to significant harm, and global ESBs, which are refer- E. N.) (Stanford, 2021); https://plato.stanford.edu/archives/sum2021/entries/justice- 
ence points for monitoring human impacts at the Earth system scale. intergenerational/.
19. Blake, M. & Smith, P. T. International distributive justice. In The Stanford Encyclopedia of 
Nations, cities, businesses and other key actors need to set and achieve Philosophy (ed. Zalta, E. N.) (Stanford, 2022); https://plato.stanford.edu/archives/
science-based targets for reducing their environmental impacts based sum2022/entries/international-justice/.
8 | Nature | www.nature.com
20. Norlock, K. Feminist ethics. In The Stanford Encyclopedia of Philosophy (ed. Zalta, E. N.) 61. Carlisle, D. M., Wolock, D. M. & Meador, M. R. Alteration of streamflow magnitudes and 
(Stanford, 2019); https://plato.stanford.edu/archives/sum2019/entries/feminism-ethics/. potential ecological consequences: a multiregional assessment. Front. Ecol. Environ. 9, 
21. Gupta, J. et al. Reconciling safe planetary targets and planetary justice: why should social 264–270 (2010).
scientists engage with planetary targets? Earth System Governance 10, 100122 (2021). 62. Mekonnen, M. M. & Hoekstra, A. Y. Four billion people facing severe water scarcity. Sci. Adv. 
22. Gupta, J. et al. Earth system justice needed to identify and live within Earth system 2, e1500323 (2016).
boundaries. Nat. Sustain. https://doi.org/10.1038/s41893-023-01064-1 (2023). 63. Minderhoud, P. S. J., Middelkoop, H., Erkens, G. & Stouthamer, E. Groundwater extraction 
23. O’Neill, B. et al. in Climate Change 2022: Impacts, Adaptation, and Vulnerability (eds may drown mega-delta: projections of extraction-induced subsidence and elevation of 
Pörtner, H.-O. et al.) 2411–2538 (Cambridge Univ. Press, 2022). the Mekong delta for the 21st century. Environ. Res. Commun. 2, 011005 (2020).
24. Gupta, J. & Schmeier, S. Future proofing the principle of no significant harm. Int. Environ. 64. Kath, J., Boulton, A. J., Harrison, E. T. & Dyer, F. J. A conceptual framework for ecological 
Agreem. 20, 731–747 (2020). responses to groundwater regime alteration (FERGRA). Ecohydrol. 11, e2010 (2018).
25. Spijkers, O. The no significant harm principle and the human right to water. Int. Environ. 65. Döll, P., Fritsche, M., Eicker, A. & Müller Schmied, H. Seasonal water storage variations as 
Agreem. 20, 699–712 (2020). impacted by water abstractions: comparing the output of a global hydrological model 
26. Rammelt, C. et al. Impacts of meeting minimum access on critical earth systems amidst with GRACE and GPS observations. Surv. Geophys. 35, 1311–1331 (2014).
the Great Inequality. Nat. Sustain. 6, 212–221 (2022). 66. Scanlon, B. R. et al. Groundwater depletion and sustainability of irrigation in the US High 
27. Steffen, W. et al. Planetary boundaries: guiding human development on a changing Plains and Central Valley. Proc. Natl Acad. Sci. USA 109, 9320–9325 (2012).
planet. Science 347, 1259855 (2015). 67. Prüss-Ustün, A. et al. Burden of disease from inadequate water, sanitation and hygiene for 
28. Raworth, K. A doughnut for the Anthropocene: humanity’s compass in the 21st century. selected adverse health outcomes: an updated analysis with a focus on low- and 
Lancet Planet Health 1, e48–e49 (2017). middle-income countries. Int. J. Hyg. Environ. Health 222, 765–777 (2019).
29. UN GA. Transforming Our World: The 2030 Agenda for Sustainable Development General 68. UNESCO WWAP The United Nations World Water Development Report 3: Water in a 
Assembly resolution 70/1 vol. A/RES/70/1 (United Nations, 2015). Changing World (UNESCO and Earthscan, 2009); https://unesdoc.unesco.org/ark:/48223/ 
30. van Vuuren, D. P. et al. Defining a sustainable development target space for 2030 and pf0000181993.
2050. One Earth 5, 142–156 (2022). 69. WHO Guidelines for Drinking-water Quality 4th edn (World Health Organization, 2022); 
31. Hickel, J. Is it possible to achieve a good life for all within planetary boundaries? Third https://www.who.int/publications/i/item/9789240045064.
World Q. 40, 18–35 (2019). 70. Rockström, J., Lannerstad, M. & Falkenmark, M. Assessing the water challenge of a new 
32. O’Neill, D. W., Fanning, A. L., Lamb, W. F. & Steinberger, J. K. A good life for all within green revolution in developing countries. Proc. Natl Acad. Sci. USA 104, 6253–6260 
planetary boundaries. Nat. Sustain. 1, 88–95 (2018). (2007).
33. Mace, G. M. et al. Approaches to defining a planetary boundary for biodiversity. Glob. 71. Aldaya, M. M., Allan, J. A. & Hoekstra, A. Y. Strategic importance of green water in 
Environ. Change 28, 289–297 (2014). international crop trade. Ecol. Econ. 69, 887–894 (2010).
34. Gleeson, T. et al. The water planetary boundary: interrogation and revision. One Earth 2, 72. Schulte-Uebbing, L. F., Beusen, A. H. W., Bouwman, A. F. & de Vries, W. From planetary  
223–234 (2020). to regional boundaries for agricultural nitrogen pollution. Nature 610, 507–512 (2022).
35. Zipper, S. C. et al. Integrating the water planetary boundary with water management from 73. Zhang, X. et al. Quantitative assessment of agricultural sustainability reveals divergent 
local to global scales. Earth’s Future 8, e2019EF001377 (2020). priorities among nations. One Earth 4, 1262–1277 (2021).
36. Heistermann, M. HESS opinions: a planetary boundary on freshwater use is misleading. 74. Springmann, M. et al. Options for keeping the food system within environmental limits. 
Hydrol. Earth Syst. Sci. 21, 3455–3461 (2017). Nature 562, 519–525 (2018).
37. Biermann, F. & Kim, R. E. The boundaries of the planetary boundary framework: a critical 75. Zhang, X. et al. Quantifying nutrient budgets for sustainable nutrient management. Glob. 
appraisal of approaches to define a ‘safe operating space’ for humanity. Annu. Rev. Biogeochem. Cycles 34, e2018GB006060 (2020).
Environ. Resour. 45, 497–521 (2020). 76. Cordell, D. & White, S. Life’s bottleneck: sustaining the world’s phosphorus for a food 
38. Wang-Erlandsson, L. et al. A planetary boundary for green water. Nat. Rev. Earth Environ. secure future. Annu. Rev. Environ. Resour. 39, 161–188 (2014).
3, 380–392 (2022). 77. Gu, B. et al. Abating ammonia is more cost-effective than nitrogen oxides for mitigating 
39. Rijsberman, F. R. & Swart, R. J. (eds) Targets and Indicators of Climate Change. Report of PM2.5 air pollution. Science 374, 758–762 (2021).
Working Group II of the Advisory Group on Greenhouse Gases (Stockholm Environmental 78. Ward, M. H. et al. Drinking water nitrate and human health: an updated review. Int. J. 
Institute, 1990). Environ. Res. Public Health 15, 1557 (2018).
40. Parmesan, C. et al. in Climate Change 2022: Impacts, Adaptation, and Vulnerability (eds 79. Tirado, R. & Allsopp, M. Phosphorus in Agriculture: Problems and Solutions. Technical 
Pörtner, H.-O. et al.) 197–377 (Cambridge Univ. Press, 2022). report (review) (Greenpeace, 2012); https://www.greenpeace.to/greenpeace/wp-content/
41. Lenton, T. M. et al. Quantifying the human cost of global warming. Nat. Sustain. https:// uploads/2012/06/tirado-and-allsopp-2012-phosphorus-in-agriculture-technical-report- 
doi.org/10.1038/s41893-023-01132-6 (2023). 02-2012.pdf.
42. Sala, E. et al. Protecting the global ocean for biodiversity, food and climate. Nature 592, 80. Haywood, J. M., Jones, A., Bellouin, N. & Stephenson, D. Asymmetric forcing from 
397–402 (2021). stratospheric aerosols impacts Sahelian rainfall. Nat. Clim. Change 3, 660–665  
43. Fedele, G., Donatti, C. I., Bornacelly, I. & Hole, D. G. Nature-dependent people: mapping (2013).
human direct use of nature for basic needs across the tropics. Glob. Environ. Change 71, 81. Krishnamohan, K. S. & Bala, G. Sensitivity of tropical monsoon precipitation to the latitude 
102368 (2021). of stratospheric aerosol injections. Clim. Dyn. 59, 151–168 (2022).
44. Vira, B. & Kontoleon, A. in Biodiversity Conservation and Poverty Alleviation: Exploring the 82. Liu, F. et al. Global monsoon precipitation responses to large volcanic eruptions. Sci. Rep. 
Evidence for a Link (eds Roe, D. et al.) 52–84 (Wiley, 2012). 6, 24331 (2016).
45. Alves, R. R. N. & Rosa, I. M. L. Biodiversity, traditional medicine and public health: where 83. Zuo, M., Zhou, T. & Man, W. Hydroclimate responses over global monsoon regions 
do they meet? J. Ethnobiol. Ethnomed. 3, 14 (2007). following volcanic eruptions at different latitudes. J. Clim. 32, 4367–4385 (2019).
46. Isbell, F. et al. Linking the influence and dependence of people on biodiversity across 84. Douville, H. et al. in Climate Change 2021: The Physical Science Basis (eds Masson- 
scales. Nature 546, 65–72 (2017). Delmotte, V. et al.) 1055–1210 (Cambridge Univ. Press, 2021).
47. Ellis, E. C. & Mehrabi, Z. Half Earth: promises, pitfalls, and prospects of dedicating half of 85. Visioni, D. et al. Seasonally modulated stratospheric aerosol geoengineering alters the 
Earth’s land to conservation. Curr. Opin. Environ. Sustain. 38, 22–30 (2019). climate outcomes. Geophys. Res. Lett. 47, e2020GL088337 (2020).
48. Garibaldi, L. A. et al. Working landscapes need at least 20% native habitat. Conserv. Lett. 86. Zhao, M., Cao, L., Bala, G. & Duan, L. Climate response to latitudinal and altitudinal 
14, e12773 (2020). distribution of stratospheric sulfate aerosols. J. Geophys. Res. 126, e2021JD035379  
49. Rocha, J. C. Ecosystems are showing symptoms of resilience loss. Environ. Res. Lett. 17, (2021).
065013 (2022). 87. Vogel, A. et al. Uncertainty in aerosol optical depth from modern aerosol‐climate models, 
50. Obura, D. O. et al. Integrate biodiversity targets from local to global levels. Science 373, reanalyses, and satellite products. J. Geophys. Res. 127, e2021JD035483 (2022).
746–748 (2021). 88. WHO Global Air Quality Guidelines: Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen 
51. Pascual, U. et al. Biodiversity and the challenge of pluralism. Nat. Sustain. 4, 567–572  Dioxide, Sulfur Dioxide and Carbon Monoxide (WHO, 2021); https://apps.who.int/iris/
(2021). handle/10665/345329.
52. Tickner, D. et al. Bending the curve of global freshwater biodiversity loss: an emergency 89. Cohen, A. J. et al. Estimates and 25-year trends of the global burden of disease 
recovery plan. Bioscience 70, 330–342 (2020). attributable to ambient air pollution: an analysis of data from the Global Burden of 
53. Reid, A. J. et al. Emerging threats and persistent conservation challenges for freshwater Diseases Study 2015. Lancet 389, 1907–1918 (2017).
biodiversity. Biol. Rev. Camb. Philos. Soc. 94, 849–873 (2019). 90. EPA. Review of the national ambient air quality standards for particulate matter. 
54. Dodds, W. K., Perkin, J. S. & Gerken, J. E. Human impact on freshwater ecosystem services: Environmental Protection Agency. 40 CFR Part 50. Fed. Regis. Rules Regul. 85,  
a global perspective. Environ. Sci. Technol. 47, 9061–9068 (2013). 82684–82748 (2020).
55. Funge-Smith, S. & Bennett, A. A fresh look at inland fisheries and their role in food 91. European Commission. Air quality standards https://ec.europa.eu/environment/air/
security and livelihoods. Fish Fish 20, 1176–1195 (2019). quality/standards.htm (2020).
56. Poff, N. L. et al. The ecological limits of hydrologic alteration (ELOHA): a new framework 92. Shaddick, G. et al. Data integration for the assessment of population exposure to  
for developing regional environmental flow standards. Freshw. Biol. 55, 147–170 (2010). ambient air pollution for global burden of disease assessment. Environ. Sci. Technol. 52, 
57. Liu, X. et al. Environmental flow requirements largely reshape global surface water 9069–9078 (2018).
scarcity assessment. Environ. Res. Lett. 16, 104029 (2021). 93. Rao, N. D., Kiesewetter, G., Min, J., Pachauri, S. & Wagner, F. Household contributions to 
58. Hoekstra, A. Y., Mekonnen, M. M., Chapagain, A. K., Mathews, R. E. & Richter, B. D. Global and impacts from air pollution in India. Nat. Sustain. 4, 859–867 (2021).
monthly water scarcity: blue water footprints versus blue water availability. PLoS ONE 7, 94. Rao, S. et al. Future air pollution in the shared socio-economic pathways. Glob. Environ. 
e32688 (2012). Change 42, 346–358 (2017).
59. Richter, B. D., Davis, M. M., Apse, C. & Konrad, C. A presumptive standard for environmental 95. van Donkelaar, A., Martin, R. V. & Park, R. J. Estimating ground-level PM2.5 using aerosol 
flow protection. River Res. Appl. 28, 1312–1321 (2012). optical depth determined from satellite remote sensing. J. Geophys. Res. 111, D21201 
60. Rolls, R. J. & Arthington, A. H. How do low magnitudes of hydrologic alteration impact (2006).
riverine fish populations and assemblage characteristics? Ecol. Indic. 39, 179–188  96. Gupta, P. et al. Satellite remote sensing of particulate matter and air quality assessment 
(2014). over global cities. Atmos. Environ. 40, 5880–5892 (2006).
Nature | www.nature.com | 9
Article
97. Persson, L. et al. Outside the safe operating space of the planetary boundary for novel Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in 
entities. Environ. Sci. Technol. 56, 1510–1521 (2022). published maps and institutional affiliations.
98. Naidu, R. et al. Chemical pollution: a growing peril and potential catastrophic risk to 
humanity. Environ. Int. 156, 106616 (2021).
99. Bai, X. et al. How to stop cities and companies causing planetary harm. Nature 609,  Open Access This article is licensed under a Creative Commons Attribution 
463–466 (2022). 4.0 International License, which permits use, sharing, adaptation, distribution 
100. Companies taking action. Science Based Targets https://sciencebasedtargets.org/ and reproduction in any medium or format, as long as you give appropriate 
companies-taking-action (2022). credit to the original author(s) and the source, provide a link to the Creative Commons licence, 
101. Technical guidance for step 1: assess and step 2: prioritize. Draft for public comment and indicate if changes were made. The images or other third party material in this article are 
(September 2022). Science Based Targets Network https://sciencebasedtargetsnetwork. included in the article’s Creative Commons licence, unless indicated otherwise in a credit line 
org/wp-content/uploads/2022/09/Technical-Guidance-for-Step-1-Assess-and-Step-2- to the material. If material is not included in the article’s Creative Commons licence and your 
Prioritize.pdf (2022). intended use is not permitted by statutory regulation or exceeds the permitted use, you will 
102. Resources for public consultation on technical guidance for companies. Science Based need to obtain permission directly from the copyright holder. To view a copy of this licence, 
Targets Network https://sciencebasedtargetsnetwork.org/resources/public-consultation- visit http://creativecommons.org/licenses/by/4.0/.
resources/ (2022).
103. Rockström, J. et al. A safe operating space for humanity. Nature 461, 472–475 (2009). © The Author(s) 2023
10 | Nature | www.nature.com
Methods to the subglobal boundary. To derive the safe levels of monthly flow 
alteration volumes for all land area globally, we analysed water bal-
This work is an output of the Earth Commission, an independent ance model (WBM) runs coupled with the TerraClimate dataset of 
international scientific assessment initiative hosted by Future Earth monthly climate forcings (Supplementary Methods has further 
(https://earthcommission.org/). The synthesis presented here builds information).
on recent work of the Earth Commission; other scientific literature, (2) F or groundwater, our approach is based on preventing declines in 
such as the PBs; intergovernmental reports, such as those of the IPCC; local aquifer levels by setting the maximum safe average annual 
and World Health Organization guidelines. As the science component drawdown equal to the average annual recharge (Supplementary 
of the Global Commons Alliance (https://globalcommonsalliance. Methods). We estimated the annual groundwater recharge and draw-
org/), the Earth Commission’s theory of change includes providing down for all land surface areas using Gravity Recovery and Climate 
our results on ESBs to the Science-Based Targets Network, the Systems Experiment satellite data covering the period from 2003 to 2016 
Change Lab and Earth HQ. coupled with data from the Global Land Data Assimilation National 
While we acknowledge that any scientific assessment will involve Oceanic and Atmospheric Administration Land Surface Model L4 
some subjectivity, we have taken several steps to ensure the scientific v.2.1 (Supplementary Methods has more detailed information).
rigour of our ESBs. (1) Our analysis is founded on a rigorous evidence (3)  For nitrogen, we used three regional environmental boundaries: 
base (Safe ESBs and Supplementary Methods). (2) Where possible, significant disruption to freshwater ecosystems (from total N run-
we determine ESBs at multiple likelihood levels (for climate change, off), groundwater potability (from nitrate leaching) and terres-
0.5 °C for low likelihood of passing climate tipping points, 1 °C for trial ecosystems (from atmospheric N deposition due to ammonia 
moderate likelihood and so on) (Table 1). (3) The nomination process and nitrogen oxide emissions) across wide areas based on critical 
for the Earth Commission and its working groups was an independent concentration limits for each. We mainly relied on a recent study72 
process managed by Future Earth (Ethics and inclusion statement). following up previous works74,104,105 that extended the approach 
(4) We report the confidence in our ESB assessments (Safe ESBs and of the original PBs27,103. This study used the Integrated Model to 
Extended Data Table 1). Assess the Global Environment (IMAGE) model to derive subglobal 
boundaries for critical nitrogen losses, surpluses and inputs based 
Safe ESBs on critical concentrations in air and water and then aggregated 
We used two main groups of approaches to setting safe ESBs: a ‘multiple these into global boundaries (Supplementary Methods has further 
elements’ approach and a ‘spatial aggregation’ approach. We describe information).
these methods here in general terms, with technical details available (4)  For phosphorus, we relied on recent work that used literature-derived 
in Supplementary Methods. These boundaries are aimed at protecting critical concentrations for avoiding eutrophication from P runoff 
Earth system stability and life-support systems for as many species as to estimate global boundaries for P mined input and surplus based 
possible, but they may not protect all species or all humans today, as on a global budget calculation, taking into account P recycling, 
further elaborated in our justice analysis. human excreta, soil and sediment retention, and global nutrient 
rebalancing74,106.
For climate and biosphere, we assessed critical thresholds for a range 
of ‘elements’ relevant to each Earth system domain through literature Our approach for the safe aerosol boundaries does not fit neatly 
review and modelling. into these two categories because we used different methods for the 
•	For climate, we based our data on those found in a recent assessment subglobal and global boundaries. Our subglobal safe boundary uses the 
of climate tipping elements16 combined with evidence on biosphere PB assessment of AODs that avoid tipping of regional monsoon systems. 
and cryosphere function and palaeoclimate variability (Supplemen- Our global assessment uses recent literature on the consequences of 
tary Methods). interhemispheric differences in aerosol concentrations on the global 
•	For functional integrity, we synthesized the literature on the area monsoon system (Quantifying ESBs and Supplementary Methods have 
needed to secure local NCP, including pollination, pest and disease further information).
control, water quality regulation, soil protection, natural hazards As a reference for a ‘safe’ Earth climate system state, we used the inter-
mitigation, and physical and psychological experiences (Supple- glacial Holocene epoch (that is, the state of the Earth system since the last 
mentary Methods). Ice Age some 11,700 years ago107,108. The Holocene’s exceptionally stable 
•	For natural ecosystem area, we examined the Earth system NCP of global climate system (oscillating <0.5–1 °C from the global pre-industrial 
carbon stocks, water flows and habitat for avoiding species extinction 14 °C mean surface temperature)107 and its configurations of global 
(Supplementary Methods). hydrology, primary production of biomass, biogeochemical cycling 
and Earth system NCP were the fundamental prerequisites for human 
From these sets of thresholds, we determined boundaries that avoid development as we know it7. We argue that only within a Holocene-like 
triggering climate tipping elements or maintain multiple local or Earth interglacial climate can Earth continue to support human well-being, 
system NCPs at different levels of likelihood. To set the climate bounda- subject to consumption behaviours and population size. There is no 
ries, we also used temperature ranges of previous Quaternary intergla- evidence that billions of humans and complex societies can thrive in 
cials and temperature ranges that maintain biosphere and cryosphere other known climates, such as a glacial ice age or ‘Hothouse Earth’7.
functioning (Supplementary Methods). We identified boundaries at multiple levels of likelihoods to reflect 
For water and nutrients, we identified subglobal boundaries relevant underlying scientific uncertainties and variabilities. These uncertain-
to these systems and then converted them into global boundaries using ties included epistemic uncertainty in the boundary value for a specific 
models or simple aggregation. Earth system process or component, such as a tipping element; variabil-
(1) F or surface water flows, we used an emerging consensus in the lit- ity in a boundary value across different places; and uncertainty when 
erature to set boundaries on the alterations (increase or decrease) aggregating multiple subglobal boundaries into a global boundary. In 
to local-scale surface water flows that protect freshwater ecosys- some cases, these levels are presented with qualitative descriptors of 
tems and fisheries (Supplementary Methods) and applied this to the each likelihood level; in other cases, they are presented as a central esti-
global land surface area. While the safe alterations can be summed mate with an uncertainty range, depending on the available evidence.
to a global alteration budget, to ensure aquatic ecosystem protec- Some of our boundary quantifications use assessments of tipping 
tion, the safe ESB is best implemented and interpreted according elements since triggering tipping can endanger Earth system stability. 
Article
Tipping elements commonly undergo changes that are abrupt (that Our concept of harm derives from the justice literature and connects 
is, faster than the forcing), large and difficult to reverse109, although to the terms impact and risk used in the assessment literature. For 
a particular tipping element may not display all three characteristics example, IPCC defines116 risk as the potential for adverse consequences 
simultaneously (for example, table 4.10 in ref. 9). We identified bounda- for human or ecological systems, including to lives; livelihoods; health 
ries based on tipping elements that accelerate or lock in change in the and well-being; economic, social and cultural assets; infrastructure; 
same Earth system component or process, such as climate tipping services; and ecosystems. These risks are a result of exposure (the pres-
accelerating further climate change or triggering the inevitable loss of ence of people or other assets in regions of Earth system change or 
an ice sheet, or that trigger a tipping element in another Earth system hazards, such as populations living near sea level) and of vulnerability 
domain, such as phosphorus concentration reaching a level that trig- (the propensity or disposition to be adversely affected, such as the poor 
gers eutrophication and disruption of freshwater ecosystems (Table 1). who live in precarious homes or health status). Impact is defined by 
IPCC as realized risk or consequences. Our harm estimates are mostly 
Safe ESBs: confidence levels based on exposure at different levels of Earth system change.
We also assessed the levels of confidence in our safe boundaries We recognize four caveats in the justice approach applied in this 
(Extended Data Table 1). ‘Confidence’ in this context can be read as paper. (1) While staying within the just boundaries as set in this paper 
‘degree of certainty in’ or ‘confidence in the validity of’ a specific ESB is crucial to avoid harm to significant sections of the human popu-
quantification. We use the same scheme for assessing and communi- lation, they are by no means guaranteeing just outcomes. Since just 
cating confidence as the IPCC110,111, which sets out two components: ends can be achieved with unjust means, meeting these boundaries 
(1) robustness of the evidence base, judged as limited, medium or robust, without transformation could significantly harm current generations. 
considering its type, amount, quality and consistency and (2) degree (2) While harm to humans is caused in part by increased exposure to bio-
of scientific agreement across the peer-reviewed literature and among physical changes, we recognize that harm is also a function of people’s 
the members of each Earth Commission Working Group, judged as low, social–economic vulnerability and lack of adaptive capacities. This is 
medium or high. Based on these two dimensions, five qualifiers can be beyond the scope of the present paper. (3) Our high levels of aggrega-
used to express the level of confidence in a particular ESB quantification: tion preclude systematic analysis of distributional justice issues in 
very low, low, medium, high and very high. This self-assessment is an terms of which social subgroups are most harmed under what scenarios. 
expert judgement based on our understanding of the available literature. (4) We do not explicitly address possible trade-offs between the three 
justice criteria. For example, policy instruments for achieving ‘I1’ may 
Just (NSH) ESBs well undermine ‘I3’ (for example, limit access to resources for marginal 
We adopt an Earth system justice lens22 for both intrinsic and instru- people). Hence, we call for redistribution, liability and compensation.
mental reasons. We show that some safe ESBs are not strong enough to Each safe ESB has been dealt with slightly differently, with some 
protect humans and other species today and that we cannot achieve and domains looking at when the system crosses tipping points (for exam-
live within the safe ESBs if inequality is high and resources are unjustly ple, climate change), others arguing that tipping points were crossed 
distributed. The evidence from behavioural experiments in public in the past and trying to recreate boundaries that allow species and 
goods provision shows that perceptions of fairness significantly alter systems to function (for example, surface water) and still others taking 
the outcomes of such experiments. In particular, individuals in disad- existing constraints into account in doing so (for example, groundwa-
vantageous positions insist on fairness even at the risk of large losses ter). Although the proposals from a safe (and I1) approach fulfil I2b in 
by doing so; such experiments suggest that climate change mitigation that they makes space for future generations of humans, they may not 
may not be achieved if rich countries are not perceived as pulling their guarantee safety for humans today (I2a; for example, climate change; 
weight112,113. In common pool resource experiments, rising income hence, we call for more stringent targets), do not address local human 
inequality leads to a downward spiral of resource overexploitation exposure to pollutants (for example, air pollution; hence, we comple-
and scarcity114. In such experiments, viewing the problem in terms of ment with local standards) or may limit access to resources (hence, 
fairness can lead to norms that motivate restraining from harvesting115. calling for redistribution26, liability, compensation and so on). Finally, 
A justice analysis is all the more needed as all science emerges from while I2a has an explicit temporal dimension, intragenerational justice 
the value systems that apply in that domain, although these are often has an explicit spatial dimension and focuses on whether all people have 
not made transparent. access to minimum resources and services26; how scarce resources are 
Within the context of our Earth system justice approach22, we use divided or shared between countries, communities and people and the 
three justice criteria or the ‘3Is’: interspecies justice and Earth system varied justice issues that arise per domain; how environmental risks 
stability (I1)17, intergenerational justice18 (I2) and intragenerational are spread worldwide and who is most exposed (through, for example, 
justice (I3). Our research into interspecies and multispecies justice mapping exposure and vulnerability) and how responsibilities are 
reveals details regarding the scholarly approaches to these concepts, shared between different actors.
but there have been no attempts to operationalize these concepts To calculate the population exposed to different levels of climate 
deductively. In our research, we have combined interspecies justice change (Fig. 2), we draw on literature for exposure to sea-level rise at 
with Earth system stability because Earth system instability undermines different levels of warming, as well as our own calculations of extreme 
non-human species and inductively identified, through domain-specific heat based on output of global models. We acknowledge that these 
(for example, climate, biosphere and aerosol loading) approaches, include a limited number of the possible impacts of climate change.
boundaries based on existing scholarship and the logic of that domain. (1)  Projections of sea-level rise need to account for dynamic processes 
Intergenerational justice refers to the justice between past and present of different complexity and for various spatiotemporal scales. In 
generations (I2a) and between present and future generations (I2b). In particular, the immediate response of several sea-level rise con-
general, although not always, our ESBs meet the I2b criteria because they tributors (such as ice sheets and inland glaciers) to global warm-
protect future generations but not the present (I2a). Intragenerational ing is only marginal due to their high inertia but can be orders of 
justice (I3) combines justice between countries19, communities and magnitude higher on centennial timescales. Therefore, to draw a 
individuals through an intersectional lens20. In balancing between the meaningful connection between selected temperature levels and 
different justice criteria, we recognize that protecting future genera- triggered sea-level rise, recent literature117,118 has resorted to a two-
tions may impose many trade-offs with the use of resources today and fold approach. The transiently realized sea-level rise throughout the 
that promoting intragenerational justice will also raise difficult issues twenty-first century is assessed by pooling Shared Socioeconomic 
regarding how to share resources, risks and responsibilities. Pathway and Representative Concentration Pathway scenarios by 
their end-of-century stabilization temperature. Those pools (for harmed by biophysical degradation is a key constraint. There is also con-
example, all scenarios that end up at 2 ± 0.25 °C) are used to drive siderable uncertainty regarding impacts on current generations, future 
localized models of sea-level rise, resulting in estimates for sea-level generations, and specific countries and communities. In this paper, we 
rise at 2100 for different end-of-century warming stabilization lev- also do not quantify issues of access26, explore the implications of access 
els117,119. Additionally, these twenty-first century projections can be for the safe and just corridor or discuss why it is difficult to meet issues 
complemented with multi-centennial estimates since long-term of access without transforming our governance systems.
sea-level rise is governed by the equilibria of the cryosphere ele-
ments and ocean thermal expansion120. In the next step, assessing Ethics and inclusion statement
exposure on these different timescales would require population Earth Commissioners were selected by the Future Earth Advisory Com-
projections, which are available for the twenty-first century but mittee following an open call for nominations with consideration for 
futile for longer timescales. For consistency, we therefore refer balancing gender, geographical region and expertise to the extent 
to a recent study that quantifies the number of people currently possible. Members of working groups were selected by the working 
(baseline from that paper: 2010 population of 6.8 billion people) group co-leads following an open call and approved by the Earth Com-
inhabiting land that is subject to inundation by end of this century mission, with attention paid to balancing gender, geographical region 
or on a multi-centennial timescale, without accounting for potential and expertise to the extent possible.
adaptation through migration, coastal defences and so on117.
(2) W et bulb temperature (TW) exposure was calculated for the his- Reporting summary
torical time period of 1979–2014 and the Shared Socio-Economic Further information on research design is available in the Nature Port-
Pathway 2-4.5 future scenario for 2015–2100. Wet bulb tempera- folio Reporting Summary linked to this article.
ture was calculated following the Davies-Jones121 method. Glob-
al gridded temperature and relative humidity data with a grid 
spacing of 1.25° × 1.25° at 6-h intervals were downloaded from Data availability
a bias-corrected global dataset122 based on 18 models from the The data supporting Figs.  2 and 3 are available at https://doi.
Coupled Model Intercomparison Project Phase 6 and the Euro- org/10.6084/m9.figshare.22047263.v2 and https://doi.org/10.6084/
pean Centre for Medium-Range Weather Forecasts Reanalysis m9.figshare.20079200.v2, respectively. We rely on other published 
5 dataset. We aggregated the data to create a maximum daily TW datasets for the climate boundary
16, N boundary72 (model files are at 
dataset and then interpolated this to match the 1° × 1° grid spacing https://doi.org/10.5281/zenodo.6395016), phosphorus73,74 (scenario 
of the spatially explicit data for the 2020 population distribu- breakdowns are at https://ora.ox.ac.uk/objects/uuid:d9676f6b-abba-
tion (most recent available, global total 7.7 billion people) from 48fd-8d94-cc8c0dc546a2, and a summary of agricultural sustainability 
the UN WPP-Adjusted Population Count, v.4.11 (ref. 123). We then indicators is at https://doi.org/10.5281/zenodo.5234594), current N 
calculated the wet bulb exposure by summing up the population surpluses129,130 (the repository at https://dataportaal.pbl.nl/down-
count for all cells with at least 1 day with a maximum TW > 35 °C. The loads/IMAGE/GNM) with the critical N surplus limit
72 subtracted, and 
TW threshold of 35 °C was chosen as it is often considered to be a estimated subglobal P concentration in runoff based on estimated 
human physiological limit of tolerance to heat stress. The human P load to freshwater131 and local runoff data132,133. Current functional 
body is unable to cool itself beyond TW = 35 °C (ref. 124,125). An integrity is calculated from the European Space Agency WorldCover 
average 1 day per year over this temperature per year is therefore a 10-metre-resolution land cover map (https://esa-worldcover.org/en). 
conservative indicator in assessing human exposure to heat stress, The safe boundary and current state for groundwater are derived 
which does not account for annual variability. We then plotted the from the Gravity Recovery And Climate Experiment (http://www2.
total number of people exposed to 1 day with a maximum TW > 35 °C csr.utexas.edu/grace/RL06_mascons.html) and the Global Land Data 
in a year against the mean annual global warming associated with Assimilation System (https://disc.gsfc.nasa.gov/datacollection/
that year to construct an exposure–temperature response curve. GLDAS_NOAH025_3H_2.1.html). More information is available in ‘Code 
(3) W e calculate the number of people displaced from the human cli- availability’ and Supplementary Methods. Source data for Fig. 2 are 
mate niche8 at different levels of warming, following the method provided with this paper.
of Lenton et al.41. The number of people exposed to mean annual 
temperatures greater than 29 °C was calculated for different 
global mean temperature increases under four different Shared Code availability
Socio-Economic Pathways. We used the downscaled spatially The code used to produce Figs. 2 and 3 are available at https://doi.
explicit output from the Coupled Model Intercomparison Project org/10.6084/m9.figshare.22047263.v2 and https://doi.org/10.6084/
phase 6 available from the WorldClim v.2.0 database at 0.0833° m9.figshare.20079200.v2, respectively. The code used to make the 
(approximately 10-km) resolution (available at https://worldclim. nutrient Earth system boundary layers in Fig. 3 is available at https://doi.
org). The exposed population is based on a 2010 population of org/10.5281/zenodo.7636716. The code used to make the surface water 
6.9 billion with spatial distribution as given by the History Data- layer in Fig. 3 and derive the subglobal Earth system boundaries for 
base of the Global Environment 3.2 database126. The mean annual surface water is available at https://doi.org/10.5281/zenodo.7674802. 
temperature threshold of 29 °C was chosen as it is beyond what The code to estimate current functional integrity is available at https://
humans have historically been exposed to8. figshare.com/articles/software/integrity_analysis/22232749/2. The 
code to derive the groundwater layer in Fig. 3 and derive the total 
To calculate current subglobal ESB transgressions (Fig. 3), we use data annual groundwater recharge is available at https://doi.org/10.5281/
for the above wet bulb and low-elevation coastal zones127 as proxies for cli- zenodo.7710540.
mate impacts, biosphere functional integrity (Supplementary Methods),  
surface water and groundwater (Supplementary Methods), exceedance 
of local safe and just nitrogen surplus and phosphorus concentration 104. de Vries, W., Schulte-Uebbing, L., Kros, H., Voogd, J. C. & Louwagie, G. Spatially explicit boundaries for agricultural nitrogen inputs in the European Union to meet air and water 
(Supplementary Methods) and PM  concentrations1282.5 . For population, quality targets. Sci. Total Environ. 786, 147283 (2021).
we used the UN WPP-Adjusted Population Count v.4.11 (ref. 123). 105. Schulte-Uebbing, L. & de Vries, W. Reconciling food production and environmental 
There are many uncertainties and limitations in this justice analysis. boundaries for nitrogen in the European Union. Sci. Total Environ. 786, 147427 (2021).106. Zhang, X. et al. Quantification of global and national nitrogen budgets for crop production. 
Lack of sufficient data on humans, communities and countries worldwide Nat. Food 2, 529–540 (2021).
Article
107. Osman, M. B. et al. Globally resolved surface temperatures since the Last Glacial 129. Beusen, A. H. W., Van Beek, L. P. H., Bouwman, A. F., Mogollón, J. M. & Middelburg, J. J. 
Maximum. Nature 599, 239–244 (2021). Coupling global models for hydrology and nutrient loading to simulate nitrogen and 
108. Kaufman, D. et al. Holocene global mean surface temperature, a multi-method phosphorus retention in surface water—description of IMAGE–GNM and analysis of 
reconstruction approach. Sci. Data 7, 201 (2020). performance. Geosci. Model Dev. 8, 4045–4067 (2015).
109. Biggs, R. et al. in Encyclopedia of Theoretical Ecology (eds Hastings, A. & Gross, L.)  130. Beusen, A. H. W. et al. Exploring river nitrogen and phosphorus loading and export to 
609–617 (Univ. of California Press, 2012). global coastal waters in the shared socio-economic pathways. Glob. Environ. Change 72, 
110. Reisinger, A. et al. The Concept of Risk in the IPCC Sixth Assessment Report: a Summary of 102426 (2022).
Cross-working Group Discussions (IPCC, 2020); https://www.ipcc.ch/site/assets/uploads/ 131. Mekonnen, M. M. & Hoekstra, A. Y. Global anthropogenic phosphorus loads to freshwater 
2021/02/Risk-guidance-FINAL_15Feb2021.pdf. and associated grey water footprints and water pollution levels: a high‐resolution global 
111. Mastrandrea, M. D. et al. Guidance Note for Lead Authors of the IPCC Fifth Assessment study. Water Resour. Res. 54, 345–358 (2018).
Report on Consistent Treatment of Uncertainties (IPCC, 2010); https://www.ipcc.ch/site/ 132. Fekete, B. M., Vörösmarty, C. J. & Lammers, R. B. Scaling gridded river networks for 
assets/uploads/2017/08/AR5_Uncertainty_Guidance_Note.pdf. macroscale hydrology: development, analysis, and control of error. Water Resour. Res. 
112. Gampfer, R. Do individuals care about fairness in burden sharing for climate change 37, 1955–1967 (2001).
mitigation? Evidence from a lab experiment. Clim. Change 124, 65–77 (2014). 133. Wisser, D., Fekete, B. M., Vörösmarty, C. J. & Schumann, A. H. Reconstructing 20th century 
113. Marotzke, J., Semmann, D. & Milinski, M. The economic interaction between climate global hydrography: a contribution to the Global Terrestrial Network Hydrology (GTN-H). 
change mitigation, climate migration and poverty. Nat. Clim. Change 10, 518–525 (2020). Hydrol. Earth Syst. Sci. 14, 1–24 (2010).
114. Owusu, K. A., Kulesz, M. M. & Merico, A. Extraction behaviour and income inequalities 
resulting from a common pool resource exploitation. Sustain. Sci. Pract. Policy 11, 536 
(2019). Acknowledgements This work is part of the Earth Commission, which is hosted by Future 
115. Liebrand, W. B. G., Jansen, R. W. T. L., Rijken, V. M. & Suhre, C. J. M. Might over morality: Earth and is the science component of the Global Commons Alliance. The Global Commons 
social values and the perception of other players in experimental games. J. Exp. Soc. Alliance is a sponsored project of Rockefeller Philanthropy Advisors, with support from the 
Psychol. 22, 203–215 (1986). Oak Foundation, MAVA, Porticus, the Gordon and Betty Moore Foundation, the Tiina and 
116. IPCC Climate Change 2022: Impacts, Adaptation, and Vulnerability (eds Pörtner, H.-O. Antti Herlin Foundation, William and Flora Hewlett Foundation and the Global Environment 
et al.) (Cambridge Univ. Press, 2022). Facility. The Earth Commission is also supported by the Global Challenges Foundation and  
117. Strauss, B. H., Kulp, S. A., Rasmussen, D. J. & Levermann, A. Unprecedented threats to the Frontiers Research Foundation. Individual researchers were supported by the European 
cities from multi-century sea level rise. Environ. Res. Lett. 16, 114015 (2021). Research Council (Grant on Climate Change and Fossil Fuel 101020082 to J.G. and Advanced 
118. Fox-Kemper, B. et al. in Climate Change 2021: The Physical Science Basis (eds Grant grant ERC-2016-ADG 743080 to J. Rockström), the Open Society Foundations (J.F.A. and 
Masson-Delmotte, V. et al.) 1211–1362 (Cambridge Univ. Press, 2021). T.M.L.), the Australian Government (Australian Research Council Future Fellowship FT200100381 
119. Rasmussen, D. J. et al. Extreme sea level implications of 1.5 °C, 2.0 °C, and 2.5 °C to S.J.L. and Australian Research Council Discovery Early Career Researcher Award DE230101327 
temperature stabilization targets in the 21st and 22nd centuries. Environ. Res. Lett. 13, to C.N.) and the Swedish Research Council Formas (Grant 2020-00371 to S.J.L.).
034040 (2018).
120. Levermann, A. et al. The multimillennial sea-level commitment of global warming. Proc. Author contributions J. Rockström, J.G., D.Q., X.B., G.B., S.E.B., F.D., K.E., C.G., N.K., T.M.L., 
Natl Acad. Sci. USA 110, 13745–13750 (2013). D.M.L., N.N., D. Obura, D.v.V., P.H.V. and R.W. conceptualized the work. J.F.A., L.S.A., D.I.A.M., 
121. Davies-Jones, R. An efficient and accurate method for computing the wet-bulb D.C., L.G., S.H., T.M.L., S.L., A.M., D. Ospina, K.P., C.R., B.S., J.S., B.S.-K., T.T., C.Z., E.M.B., S.B., 
temperature along pseudoadiabats. Mon. Weather Rev. 136, 2764–2785 (2008). W.B., P.G., L.H., L.J., C.N., S.P., J. Rocha, M.S., L.S.-U., W.d.V., C. Xiao, C. Xu, X.X., N.Z.-C. and X.Z. 
122. Xu, Z., Han, Y., Tam, C.-Y., Yang, Z.-L. & Fu, C. Bias-corrected CMIP6 global dataset for gathered and analysed data. J. Rockström, J.G., D.Q., S.J.L., X.B., G.B., S.E.B., F.D., K.E., C.G., 
dynamical downscaling of the historical and future climate (1979–2100). Sci. Data 8, 293 N.K., T.M.L., D.M.L., N.N., D. Obura, D.v.V., P.H.V., R.W., J.F.A., L.S.A., D.I.A.M., D.C., L.G., S.H., S.L., 
(2021). A.M., D. Ospina, K.P., C.R., B.S., J.S., B.S.-K., T.T., C.Z., E.M.B., P.G., C.N., L.S.-U., W.d.V. and X.Z. 
123. CIESIN. Gridded population of the world, version 4 (GPWv4): population count adjusted wrote the paper. S.J.L. coordinated writing.
to match 2015 revision of UN WPP country totals, revision 11. Center for International  
Earth Science Information Network, Columbia Univ. https://doi.org/10.7927/H4PN93PB Funding Open access funding provided by Stockholm University.
(2018).
124. Im, E.-S., Pal, J. S. & Eltahir, E. A. B. Deadly heat waves projected in the densely populated Competing interests The authors declare no competing interests.
agricultural regions of South Asia. Sci. Adv. 3, e1603322 (2017).
125. Shaw, R. et al. in Climate Change 2022: Impacts, Adaptation, and Vulnerability (eds Additional information
Pörtner, H.-O. et al.) 1457–1579 (Cambridge Univ. Press, 2022). Supplementary information The online version contains supplementary material available at 
126. Klein Goldewijk, K., Beusen, A., Doelman, J. & Stehfest, E. Anthropogenic land use https://doi.org/10.1038/s41586-023-06083-8.
estimates for the Holocene–HYDE 3.2. Earth Syst. Monit. 9, 927–953 (2017). Correspondence and requests for materials should be addressed to Johan Rockström or 
127. CIESIN-CIDR. Low elevation coastal zone (LECZ) urban-rural population and land area Steven J. Lade.
estimates, version 3. Columbia Univ. and CUNY Institute for Demographic Research, City Peer review information Nature thanks Stephen Humphreys, Thomas Nesme, Henrique Pereira 
Univ. of New York https://doi.org/10.7927/d1x1-d702 (2021). and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. 
128. van Donkelaar, A. et al. Monthly global estimates of fine particulate matter and their Peer reviewer reports are available.
uncertainty. Environ. Sci. Technol. 55, 15287–15300 (2021). Reprints and permissions information is available at http://www.nature.com/reprints.
Extended Data Fig. 1 | Alternative visualizations of safe and just Earth b, left circle) show current global states; a single current state cannot be 
system boundaries (Fig. 1). Concentric (a) and parallel (b) visualizations of defined sub-globally. Short concentric lines (that extend across less than the 
global (a, inner circle; b, left circle) and sub-global (a, outer circle; b, right full width of a wedge) represent alternative likelihood levels (safe) or levels of 
circle) safe and just ESBs. Colours are as in Fig. 1. Global rings (a, inner circle;  exposure ( just NSH) (Table 1).
Article
Extended Data Table 1 | Assessment of levels of confidence in each domain’s safe Earth system boundaries
For more information see Methods. The robustness of evidence and degree of agreement of all ESB quantifications are based on the assessment of available literature and working group 
experts’ views.