Article

Realization of Paris Agreement pledges may

limit warming just below 2 °C

https://doi.org/10.1038/s41586-022-04553-z Malte Meinshausen1,2 ✉, Jared Lewis2,3, Christophe McGlade4, Johannes Gütschow2,5,

Zebedee Nicholls1,2,3, Rebecca Burdon2,6, Laura Cozzi4 & Bernd Hackmann7
Received: 26 November 2021

Accepted: 16 February 2022

Over the last five years prior to the Glasgow Climate Pact1, 154 Parties have submitted
Published online: 13 April 2022
new or updated 2030 mitigation goals in their nationally determined contributions
Check for updates
and 76 have put forward longer-term pledges. Quantifications of the pledges before
the 2021 United Nations Climate Change Conference (COP26) suggested a less than
50 per cent chance of keeping warming below 2 degrees Celsius2–5. Here we show that
warming can be kept just below 2 degrees Celsius if all conditional and unconditional
pledges are implemented in full and on time. Peak warming could be limited to
1.9–2.0 degrees Celsius (5%–95% range 1.4–2.8 °C) in the full implementation
case—building on a probabilistic characterization of Earth system uncertainties in line
with the Working Group I contribution to the Sixth Assessment Report6 of the
Intergovernmental Panel on Climate Change (IPCC). We retrospectively project
twenty-first-century warming to show how the aggregate level of ambition changed
from 2015 to 2021. Our results rely on the extrapolation of time-limited targets
beyond 2030 or 2050, characteristics of the IPCC 1.5 °C Special Report (SR1.5)
scenario database7 and the full implementation of pledges. More pessimistic
assumptions on these factors would lead to higher temperature projections.
A second, independent emissions modelling framework projected peak warming of
1.8 degrees Celsius, supporting the finding that realized pledges could limit warming
to just below 2 degrees Celsius. Limiting warming not only to ‘just below’ but to ‘well
below’ 2 degrees Celsius or 1.5 degrees Celsius urgently requires policies and actions
to bring about steep emission reductions this decade, aligned with mid-century
global net-zero CO2 emissions.

Intensifying heatwaves, coral reef bleaching, increasing flood and By mid-November 2021, 153 Parties had submitted new or updated
extreme precipitation events6,8 are climate impacts at the present-day NDCs. By that time, 75 of them had also provided long-term targets as
level of warming of 1.2 °C above 1850–1900 levels, attributable to part of their NDCs or 44 LT-LEDS (Fig. 1). In addition, India announced a
human activities9. With plummeting costs of renewable energy10 and new NDC and 2070 net-zero goal at COP2614 (we interpret India’s 2070
electrification options throughout all sectors, in addition to a series of goal conservatively as net-zero CO2, and add it to the list of Parties with
climate policies already in place, very-high-emission projections such new and updated NDCs and long-term targets, making the totals 154
as Shared Socioeconomic Pathway 5-8.5 (SSP5-8.5) are increasingly and 76, respectively).
unlikely11. Yet, for countries to achieve the long-term target of the Paris The long-term targets are mostly net-zero CO2 or net-zero green-
Agreement12 of keeping warming to ‘well below’ 2 °C and ‘resolving’ to house gas (GHG) emission targets for 2050, although a few, includ-
pursue efforts to limit warming to 1.5 °C (ref. 1), substantial mitigation ing China, aim for 2060. Recent analysis2,3 considered updated NDCs
action is required this decade—well beyond the pledges made at the received prior to August or September 2021, and indicated that current
time of the Paris Agreement5. As a major shift in emphasis to the Paris pledges with continued ambition still had a higher than 50% chance of
Agreement, the Glasgow Climate Pact1 stressed the importance of 1.5 °C exceeding 2 °C.
even more clearly.
The Paris Agreement12 contains a ratcheting-up process that requires
countries to update and progress their so-called nationally determined Bottom-up analysis in line with IPCC
contributions (NDCs), in the run-up to Glasgow1 and every five years Going beyond other analyses that focus on key countries, we revisit
thereafter13. In addition, Parties to the Paris Agreement are invited to officially submitted inventory data and targets of 196 countries, their
submit their long-term low-emission development strategies (LT-LEDS). NDCs and LT-LEDS from the time of the Paris Agreement until after
1
Climate & Energy College, School of Geography, Earth and Atmospheric Sciences, The University of Melbourne, Melbourne, Victoria, Australia. 2Climate Resource, Melbourne, Victoria,
Australia. 3International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria. 4International Energy Agency, Paris, France. 5Potsdam Institute for Climate Impact Research,
Potsdam, Germany. 6Melbourne Climate Futures, The University of Melbourne, Melbourne, Victoria, Australia. 7United Nations Climate Change, Bonn, Germany.
✉e-mail: malte.meinshausen@unimelb.edu.au

304 | Nature | Vol 604 | 14 April 2022

 First NDCs
200
New or updated NDCs
Long-term targets (NDCs + LT-LEDS)
Long-term strategy
150
NZE targets
in policy process
Number of countries
100
NZE targets
in law
Announced
50
NZE targets
0 0
Fig. 1 | Number of countries that have NDCs, long-term targets and net-zero
emission targets and their share of global CO2 emissions. Almost all
countries submitted at least one NDC, and the majority submitted new or
updated NDCs. Seventy-six countries (plus the 27 countries of the European
Union (EU27) as a whole) have long-term targets that are part of either NDCs or
LT-LEDs, and 44 (plus the EU27 as a whole) have LT-LEDS. In the three rightmost
columns, countries included are those with net-zero targets in the policy
process, in law or announced—including their respective share of global 2019
fossil and industrial CO2 emissions35. Status as of 11 November 2021 (Methods).
NZE, net-zero emission.
COP26 on 11 November 2021. We also consider recent pledges for inter-
national maritime transport emissions15 and international aviation16.
In our bottom-up analysis, we find that updated NDCs result in pro-
jected 2030 global GHG emissions, excluding land use, land-use change
and forestry (LULUCF), of 50.9 to 54.9 gigatonnes of CO2 equivalent
(GtCO2e; 100-yr Global Warming Potential (GWP-100) from AR6)6,
which is +7.0% to +15.4% above 2010 levels and close to 2019 levels
(−3.4% to +4.1%). The Glasgow Climate Pact refers to the IPCC finding
that limiting warming to 1.5 °C requires reducing CO2 emissions by
45% compared to 2010 levels by 2030: by contrast, on the basis of our
NDC quantifications, 2030 CO2 emissions are estimated to be 6% to
13% above 2010 levels. If, in addition to the 2030 NDC targets, the 76
long-term targets are achieved in full (including the announcement by
India), GHG emissions would be 32%–34% below 2010 levels by 2050.
One central measure of comparability across nations is their per cap-
ita emissions17. We find average estimated per capita emissions in 2030
rising for some key countries relative to 2019 levels, including China
(6% increase of average per capita emissions by 2030 over 2019 lev-
els), Turkey (33%), Brazil (16%) and Russia (21%), whereas a number
of countries with high per capita emissions levels exhibit substantial
decreases, including Canada (−46%), Germany (−40%), Japan (−34%),
UK (−43%) and Kazakhstan (range from −16% to −26%) (Fig. 2a). Total
emissions are projected to increase for India by +1.1 to +1.5 GtCO2e
(a 31%–44% increase), but its per capita emissions in 2030 are projected
to remain well below the global average (Fig. 2a). The pledged long-term
targets generally lead to a convergence of per capita emissions, but
some of the current developing countries will have higher per capita
emissions than industrialized countries by 2030. There are also several
countries where NDC targets imply higher emissions in 2030 than in a
downscaled country-level scenario with no additional climate policies;
the excess emissions above the reference scenario up to the target are
referred to as ‘hot air’ (Methods, Fig. 2b).
There are a number of challenges and uncertainties in estimating
a bottom-up aggregation of the historical and future pledge emis-
sions of countries, including: under-reported or non-reported sectors;
uncertainties due to vague NDC definitions (for example, net-zero
emissions targets that do not specify whether they refer to CO2 or all
Share of global CO2 emissions 2020, excl. LULUCF (%)
100
GHG) and conditionalities (for example, access to financial resources,
technology transfer, technical cooperation, capacity-building support,
the availability of market-based mechanisms, and absorptive capacity
75 of forests and other ecosystems); targets specified as ranges; future
pledges that are above no-climate-policy reference scenarios (hot air);
and difficulties in quantifying changes in emissions and sinks in the
LULUCF sector18 (Methods).
50
Projecting temperatures from 2030 levels
We first look at what 2030 NDCs and current energy policies would
25 imply for long-term temperature rise. Present-day energy policies, as
encapsulated by the International Energy Agency (IEA) Stated Policies
Scenario (STEPS)4, which considers only existing energy policies and
measures that are under development, implies warming of 2.6 °C (1.9 °C
to 3.7 °C) by 2100 (Extended Data Table 1). Very similar projections are
obtained by United Nations Environment Programme (UNEP)3,19 and
by the Climate Action Tracker20 for their respective ‘current policies’
estimates (Extended Data Table 1). Quantifying the temperature rise
implications on the basis of only 2030 NDC targets faces two main
challenges: first, how to translate NDCs into absolute emissions and
second, how to extend the NDC ambition beyond 203021,22, which is
necessary to estimate end-of-century warming.
To encapsulate the uncertainty within 2030 NDCs, we distinguish
eight cases for 2030 NDC emissions, namely: the permutations of
‘low-end’ and ‘high-end’ quantifications (in which NDCs are either
uncertain or provided as a range); including hot air or excluding hot
air (in which high-emission NDCs are taken at face value or emissions
are capped at reference scenarios); and considering only unconditional
NDC elements or assuming the full implementation of NDCs (Fig. 3b
and Extended Data Table 2).
To extend 2030 emissions pathways to 2100 we draw on recent litera-
ture. Studies have taken various approaches21,22. Some use integrated
assessment models to calculate implicit carbon prices based on 2030
NDC emissions levels, and extend these over the next 70 years2,3,23.
Another approach is to choose a wide variety of scenarios from different
modelling groups7 to determine the percentile of a certain emission
level within the large ensemble of scenarios and pursue that percentile
level21,24,25. One difficulty with this approach is that a number of sce-
narios in the most recent SR1.5 database7 investigate delayed-action
futures, with limited mitigation efforts prior to 2030, and then rapid
emissions reductions after 2030. For the purpose of using the database
as foundation for a percentile-based extension, these scenarios distort
the distribution over scenarios before and after 2030. In this work, to
address this discontinuity in effort, we use 2025–2030 annual growth
or reduction rates of emissions to project emissions from 2030 to 2050.
From 2050, we use an equal-quantile-walk approach24,25, providing an
approximation of potential twenty-first-century emission trajectories
in line with the 2030 NDCs only. This approach provides emission levels
and rates of change broadly similar to those of SSP2-4.5 (Fig. 3b and
Extended Data Fig. 1) and the IEA STEPS scenario4. Across the eight
scenarios that represent only 2030 NDCs, the range of conditionality,
hot air and how NDCs can be interpreted, our probabilistic method
finds a range of peak twenty-first-century warming of 2.2 °C (5%–95%
geophysical uncertainty range 1.6 °C to 3.3 °C) for the lowest scenario
to 3.0 °C (2.2 °C to 4.4 °C) for the highest (‘a’ and ‘h’, respectively, in
Extended Data Table 2 and Fig. 3c). The wide range of peak temperatures
highlights that a single projection based on 2030 NDCs is not repre-
sentative, in line with recent multi-model study findings23 (‘9’ in Fig. 3c).
Net-zero pledges get us just below 2 °C but not 1.5 °C
Compared with our estimates based on 2030 NDCs only (Fig. 3b, c),
the long-term targets, if implemented in full and on time, could bring
Nature | Vol 604 | 14 April 2022 | 305
Article
a b Hot air: unconditional 2030 NDCs
25 higher than reference
Trinidad and Tobago Trinidad and Tobago
Brunei Darussalam Brunei Darussalam
Qatar Palau Turkmenistan Qatar Income group36
United Arab Emirates Kuwait Pakistan
Oman Oman Bahrain High income
Kuwait Upper middle income
United Arab Emirates Turkey SSP5-BL
Australia Lower middle income SSP1-BL
Low income
Vietnam
USA Saudi Arabia Bunkers

20 Global average Venezuela

Bahrain
Canada Kazakhstan
Turkmenistan Total emissions (MtCO2e yr−1) Uzbekistan
Russian Federation 0
Per capita GHG emissions, excl. LULUCF, GWP-100 AR6 (tCO2e per capita)

New Zealand 2,500 Russian Federation

Luxembourg
Mongolia Saudi Arabia
7,500
Palau Bangladesh
Australia
Mongolia 12,500
Democratic People’s Republic of Korea
Kazakhstan
15 Russian Federation
Philippines

Republic of Korea
Error bars Nicaragua

USA Iran Cameroon

Iran

Luxembourg Burkina Faso

Canada
New Zealand Republic of Korea Niger
10 Germany 76 countries with long-term
Japan China targets in their NDCs
China Somalia
South Africa or LT-LEDS and India’s
1 Nov 2021 announcement
Turkey South Africa Trinidad and Tobago

Italy Cambodia
UK
France Brazil
Turkey Japan
Thailand Thailand Ethiopia
Italy Germany
Brazil Mexico Mexico
Vietnam France
5 Singapore
Vietnam
UK China Kenya
India Indonesia New Zealand
Indonesia
Canada Thailand Zimbabwe
Pakistan
India USA
Pakistan Mexico
Zambia
India
Intl aviation Intl aviation South
Indonesia Africa Brunei Darussalam
Intl maritime transport
0 Other
Intl maritime transport

2019 2030 2050 0 100 200 300 400 500 600 700 800
GHG emissions, GWP-100 AR6
All countries All countries Countries with long-term targets (MtCO2e yr−1)

Fig. 2 | Although some near-term pledges are estimated to be very weak, error bars), here capped at assumed reference level scenarios (Methods). For
implying potentially substantial overachievement (hot air), countries with comparison, global average per capita emissions in 2019 and following
long-term targets envisage substantial per capita emission reductions. scenarios A and B (Extended Data Table 2) are shown (blue horizontal bars).
a, Per capita emissions in 2019, under NDC targets in 2030 of all analysed 196 Income groups are based on ref. 36. b, Potential overachievement (‘hot air’) of
countries and under long-term targets for 2050 for 76 countries (that have current NDC in 2030 under two no-additional-climate policy reference
long-term targets), as well as international transport emissions. The range approximations (Methods). The brown box indicates hot air when assuming an
depicts conditional and unconditional NDCs in 2030 and uncertainty SSP5 baseline (SSP5-BL), and the grey bar indicates additional hot air following
regarding the quantification of countries’ emissions in 2050 (vertical grey a SSP1-baseline. Intl, international.

down the peak temperatures to a median of 1.9 °C or just 2.0 °C (sce- methane reductions from the Global Methane Pledge (GMP). The APS
narios A and B, Fig. 3a). This is found in two scenarios that combine low results in a temperature rise of 1.8 °C in 2100 (5%–95% ranges of 1.3 to
and high emission levels in 2030 (implementing all conditional NDC 2.6 °C). For the first time, the combination of 2030 NDCs and long-term
elements or not) with a lower or higher interpretation of long-term pledges, if they are all implemented in full (that is, including conditional
targets, here called scenarios A and B, respectively. Peak warming is elements) and on time, could bring median warming to or just below
reached around 2080–2090 (5%–95% range 2049/2059 to 2100) with the symbolic 2 °C mark.
an exceedance probability of 90%–94% for 1.5 °C, 42%–52% for 2 °C Despite this encouraging finding, our analysis also shows that
and 12%–17% for 2.5 °C (Fig. 3a, inset). Warming in 2100 is estimated broadening the scope of long-term net-zero targets is alone not
to be 1.9–2.0 °C with 5%–95% ranges of 1.3–2.8 °C and 1.4–3.0 °C for sufficient to keep warming to around 1.5 °C. That is because ambi-
scenarios A and B, respectively (Fig. 3a and Extended Data Table 2). The tion up to 2030 will be crucial to contain warming closer to 1.5 °C.
IEA has independently assessed the range of temperature outcomes of Median warming is expected to exceed 1.5 °C shortly after 2030 even
achieving all long-term energy and emission pledges in its Announced under the most ambitious emission pledge scenarios developed
Pledges Scenario (APS). In addition to all of the long-term targets in here as well as in strong decarbonization scenarios such as SSP1-1.9
scenarios A and B, the APS includes the announced net-zero pledges (Fig. 2a). The amount of further warming from today and from 2030
by Russia and Saudi Arabia, as well as an assumption about additional depends on mitigation action we take from today: strong action

306 | Nature | Vol 604 | 14 April 2022

 a b This d 700
3.5 IEA SSPs A B Percentiles study’s
70 range of
100% 5–95%

GHG emissions, GWP-100 AR6 (GtCO2e yr−1)

2030

Gobal methane emissions (MtCH4 yr−1)

B 17–83% 600
33–67% 60 extension
A
Exceedance probability

80% Median (50%) Other options

1.5 °C
comparable 50 h 500
3.0 studies
60% based on STE g
50% B 2030 NDCs 40 400
and long-
40% 2.0 °C A term targets f
30 b c
SSP2-4.5 e 300
2.5 20% d o
20 This study APS a n
2030 NDCs m A
B 200
0% 10 + long-term
2000 2100 target scenarios A
GMST rel. 1850–1900 (°C)

Year
100
B
0
2.0
A '
1b –10 0
APS 2000 2020 2040 2060 2080 2100 2000 2020 2040 2060 2080 2100
1a 2b

SSP1-2.6
2a 5 c 3.5 STE e
2.5

rel. 1850–1900 (°C)

1.5
4
3.0
2.0

GMST
SSP1-1.9 APS A

GMST rel. 1850–1900 (°C)

2.5

1.0 1.5
2.0
o n
7 9 11 1.0 m
6 8 10
1.5 0
o
f

rel. scenario A (°C)

n
0.5

Median GMST
1.0 –0.02
This study Other m
comparable
0.5 studies based
on 2030 NDCs –0.04

0 0
2000 2020 2040 2060 2080 2100 2000 2020 2040 2060 2080 2100 2000 2020 2040 2060 2080 2100
Year Year Year

Fig. 3 | Global mean temperature projections based on 2030 NDCs show a cases for 2030 NDCs and long-term targets (blue band) with individual
wide range, whereas those based also on long-term targets stay just around scenarios described in Extended Data Table 2. c, Global mean temperatures
or below 2 °C—with limited additional effects by the GMP. a, Warming due to due to the extrapolation of 2030 NDC sensitivity cases, including a comparison
NDCs and long-term targets. Global mean temperature projections based on all to COP26 analyses (labelled as per Extended Data Table 1). d, Global methane
officially submitted NDCs and long-term targets as of the end of COP26 in emissions including sensitivity cases reflecting our GMP quantification.
addition to the announcement by India (1 November 2021), with a sensitivity e, Global mean temperature projections under scenario A, as well as the GMP
case considering full implementation of NDC targets (A) and only sensitivity scenarios. f, Temperature differences of the median of the GMP.
unconditional targets (B). Inset, scenario A has a better than 50% chance and b, d also display emissions of the SR1.5 emission database (thin grey lines) and
scenario B has a roughly 50% chance, but neither scenario has a likely 67% the SSP1-1.9, SSP1-2.6 and SSP2-4.5 scenarios (grey dash-dotted lines) in the
chance to stay below 2 °C. b, Greenhouse gas emissions without LULUCF for the background. APS, IEA Announced Pledges Scenario; GMST, global mean
considered pathways extending 2030 NDCs (orange) and the two sensitivity surface temperature; rel., relative to; STE, IEA Stated Policy Scenario (STEPS).

will leave us with peak warming around 1.5 °C, whereas weak action
will see temperature continue to rise to 1.7, 1.8, 2.0 °C or higher. Uncertainties and side-shows
Remaining cumulative CO2 emissions to limit median peak warm- Our temperature projections are subject to several uncertainties, ren-
ing to 1.5 °C (estimated to be 500 GtCO2 from 20206 or approxi- dering the median warming projections as indicative. First, despite
mately 420 GtCO2 from 2022 onwards given recent emissions26) are recent advances narrowing the uncertainty range of climate sensitivity
almost entirely consumed by projected cumulative 444–457 GtCO2 down to a probable range of 2.5 °C to 4.0 °C28 compared to a previous
emissions from 2020 to 2030. By 2030, a remainder of only about broader range of 1.5 °C to 4.5 °C29, the carbon cycle, radiative forcing
1–2 years of emissions at the 2019 level (approximately 40 GtCO 2)26 and climate response uncertainties as assessed by the IPCC are still
is left in the 1.5 °C budget (Fig. 4a), consistent with the estimate substantial. Second, even though our calibrated climate emulator
by the United Nations Framework Convention on Climate Change MAGICC7 has the lowest divergence from IPCC assessed temperature
(UNFCCC)27. Between 2020 and 2050, our most optimistic imple- ranges across the four considered climate emulators (cross-chapter
mentation scenario (scenario A) has cumulative CO2 emissions of box 7.1 in ref. 28), some differences exist, including a slightly lower
around 1,000 GtCO2 (Fig. 4b). end-of-century warming in the SSP1-1.9 scenario case. Our finding of
As enshrined in Article 4.1 of the Paris Agreement, one important median warming at or just below 2 °C is, however, considered robust
near-term indicator for the energy transition is a peak in GHG global because the divergence from the SSP1-2.6 and higher scenarios (more
emissions. ‘Cost-optimal’ scenarios in the SR1.5 database that limit similar to the pathways assessed here) is within 5% (cross-chapter box 7.1
warming to around 1.5 °C or 2 °C feature a peak in GHG emissions early in ref. 28). Third, the extension of emissions beyond 2050 here uses the
in this decade, 2020–2030 (Fig. 3b). With the most recent NDC pledges multi-gas and time-evolution characteristics of the SR1.5 database of
(excluding hot air), 2030 emissions might be lower than 2025 levels scenarios. Recently, new scenarios emerged following a new scenario
(that is, emissions have peaked) by up to 1.7 GtCO2e (Fig. 5a). framework30 that is more geared towards limiting peak temperatures
To put our findings based on the most recent pledges into context, and tends to favour earlier stronger reductions31 with less reliance on
we performed hundreds of climate projections to characterize the net-negative emissions in the second half of the century. We find that
aggregate effect of countries’ ambitions each time a new NDC pledge our scenario A extension beyond 2050 is robust in the sense that it tends
was made over the last five years. They allow us to track the progres- to be conservative, that is, on the higher side of post-2050 emissions
sion of pledges in a single, consistent methodological framework to compared to this new set of scenarios with similar 2050 emission levels
estimate their probable effect on future cumulative carbon emissions (Extended Data Fig. 2).
(Fig. 4), the peaking of global emissions (Fig. 5a) and the implied future COP26 saw several announcements outside the formal negotiation
temperature outcomes (Fig. 5b)—if these targets are achieved and process, including on forestry, cars, finance and coal. In addition,
reflected in real-world emissions. the GMP was launched by the USA and the European Union, signed by an

Nature | Vol 604 | 14 April 2022 | 307

Article
a Fig. 4 | Pledges submitted at the time of the Paris Agreement (2015) and up
South Africa
Russia to after COP26 show a strong recent decrease in projected cumulative
Cumulative CO2 emissions from 2020 to 2030 (GtCO2)

India Vietnam
Brazil
480 Pakistan Intl shipping
Philippines emissions. a, Cumulative carbon emissions from 2020 to 2030 that are implicit
USA
Iran
Nigeria
in unconditional pledges (orange) and the full implementation of NDCs, that is,
475 Turkey
Iraq Iraq including conditional elements (blue). b, As in a, but for the period 2020 to
Japan
470 China 2050, including long-term targets. Step changes due to individual
country’s pledge announcements are highlighted for steps larger than 1 GtCO2
465 cumulative 2020–2030 emissions (a) or 5 GtCO2 2020–2050 emissions (b). For
countries with multiple updates of their NDCs, only the most recent NDC
460 update is displayed. The temporary withdrawal of the USA from the Paris
Kazakhstan Agreement is not shown.
455
Brazil
USA
450 Unconditional range
China
Full implementation range
USA
445
India
2044 (Fig. 3d). In the IEA APS scenario33, such a drop is achieved soon
2016 2017 2018 2019 2020 2021 2022
after 2050. As a result, the GMP could instead be seen as reconfirming
b reductions already implicit in NDCs and long-term targets, albeit bring-
Cumulative CO2 emissions from 2020 to 2050 (GtCO2)

1,400
India ing these reductions forward in time. As a consequence, the additional
UK Intl shipping
1,350 Brazil temperature reductions would be less than usually assumed.
Mexico
Philippines If the GMP is implemented as a 30% methane reduction on a
1,300 USA
Indonesia country-by-country level by signatory countries (which account for
Nigeria
1,250 Iraq approximately 54% of global methane emissions)34 we find additional
Japan Japan
Brazil methane reductions of 37–56 Mt CH4 yr−1, depending on whether full
1,200
China
USA
implementation or conditional NDCs are assumed as the appropriate
1,150 scenario to compare to (Methods). These additional methane reduc-
Cumulative emissions to keep warming below 2 °C with 67%
tions can help to reduce the peak level of warming, but the impact of
1,100 Ukraine
Saudi Arabia fully implementing the GMP is more muted than when compared to a
1,050 reference scenario that has rising methane emissions. For example, in
China
Range of results for only unconditional NDC elements
Range of results for full implementation of NDCs
our scenario A, we find that achieving the GMP would reduce median
1,000
Point in time of new NDC submission India
peak (2100) temperatures by 0.016 °C (0.009 °C) in the case of a global
2016 2017 2018 2019 2020 2021 2022 30% methane reduction, and 0.004 °C (0.003 °C) to 0.007 °C (0.005 °C)
NDC submission and evaluation date when the 30% methane reductions occur only in the 104 signatory
countries (Methods and Extended Data Table 2).
additional 103 countries (as of 2 November 2021), and was announced
as helping to shave off an additional 0.2 °C warming32. The temperature
impact of achieving the GMP depends on how its goal is interpreted, Progression and implementation challenge
the background scenario considered, and the time at which the tem- Our results provide a reason to be optimistic: warming could be limited
perature effect is evaluated. In the IEA STEPS, a 30% reduction in global to 2 °C or just below, if the pledges on the table are implemented in full
methane emissions—a drop of approximately 100 Mt CH4 yr−1 from and on time. Peaking of global GHG emissions could be achieved this
2020 levels—would reduce the temperature rise by 2050 by around decade. But our results also provide a sobering assessment of how
0.12 °C. In our scenario A, a reduction of 100 Mt CH4 yr−1 is achieved by far current pledges are from limiting warming to 1.5 °C. The IEA APS

a b
Peak median warming in 21st century rel. 1850–1900 (K)

Turkey China Russia India

Global GHG emissions in 2030 rel. 2025 (MtCO2e)

Iran Brazil UK
India 3.00 Intl shipping
–1,500 Burundi UK Brazil
Saudi Arabia Germany
Intl shipping
Egypt UAE
–1,000 Iraq Colombia 2.75 USA
Ukraine Indonesia
South Africa
–500 2.50
a Japan
bc
de Brazil Unconditional range
No peaking before 2030 f
0 g
2.25 China
Peaking before 2030 USA Full implementation range
Nigeria 2 °C
500 Kazakhstan 2.00
Indonesia
Brazil Philippines China
USA India
1,000 1.75
a. Burundi
b. Burkina Faso
c. Iraq 1.5 °C
1,500 Range of results for only unconditional NDC elements d. Japan 1.50
Range of results for full implementation of NDCs e. Saudi Arabia USA
f. China
2,000 Point in time of new NDC submission g. India 1.25
2016 2017 2018 2019 2020 2021 2022 2016 2017 2018 2019 2020 2021 2022
NDC submission and evaluation date NDC submission and evaluation date

Fig. 5 | Pledges submitted at the time of the Paris Agreement (2015) and up some lower pathways that limit median temperature to below 2 °C and feature
to after COP26 indicate the possibility of emission peaking before 2030 peak temperatures around the 2070s. Step changes due to individual pledge
and lowering median warming. a, The difference between estimated 2030 announcements are highlighted for some countries for steps larger than a
and 2025 global GHG emission emissions levels serves as an indicator for 50 MtCO2e of emission differences (a), or 0.04 °C warming (b). For countries
peaking of global emissions. b, Projected median peak warming over the with multiple updates of their NDCs, only the most recent NDC update is
twenty-first century relative to 1850–1900. Peak median warming generally displayed. The temporary withdrawal of the USA from the Paris Agreement is
occurs at the end of the century (2100) in the investigated scenarios except for not shown. UAE, United Arab Emirates.

308 | Nature | Vol 604 | 14 April 2022

scenario4 (Fig. 3b), has a strong reduction in GHG emissions to 2030,
but peak warming is still 1.8 °C; in the case of scenarios A and B, which
have a slower reduction to 2030, the peak temperature rise would be
1.9 °C or even 2.0 °C. The emission gap in 2030 between these scenarios
and scenarios in line with SSP1-1.9 or ‘below 1.5 °C’ is substantial.
Limiting peak or 2100 warming to 1.5 °C is dependent on four main
pillars. First, existing long-term pledges and targets need to be under-
pinned by policies and measures during this decade. Second, emission
reduction targets this decade need to be substantially enhanced to
reduce CO2 emissions by 45% by 2030 and to decisively reduce non-CO2
emissions. Third, countries without net-zero targets should consider
announcing and implementing such targets (with appropriate support
in the case of low-income countries). Finally, countries with net-zero
targets could bring their net-zero dates forward or aim for net-negative
targets. Commitments made so far, especially for this decade, fall far
short of what is required to limit temperature rise to 1.5 °C. Any delay in
reversing the upward trend of emissions, phasing out the unabated use
of fossil fuels and developing sustainable, additional and permanent
negative-emissions options will put this target out of reach.
Online content
Any methods, additional references, Nature Research reporting sum-
maries, source data, extended data, supplementary information,
acknowledgements, peer review information; details of author contri-
butions and competing interests; and statements of data and code avail-
ability are available at https://doi.org/10.1038/s41586-022-04553-z.
1. UNFCCC. Glasgow Climate Pact https://unfccc.int/documents/310475 (United Nations
Framework Convention on Climate Change, 2021).
2. Ou, Y. et al. Can updated climate pledges limit warming well below 2 °C? Science 374,
693–695 (2021).
3. UNEP. Emissions Gap Report 2021: The Heat is On – A World of Climate Promises Not Yet
Delivered https://www.unep.org/resources/emissions-gap-report-2021 (United Nations
Environment Programme, 2021).
4. IEA. World Energy Outlook 2021 https://www.iea.org/reports/world-energy-outlook-2021
(International Energy Agency, 2021).
5. Rogelj, J. et al. Paris Agreement climate proposals need a boost to keep warming well
below 2 °C. Nature 534, 631–639 (2016).
6. IPCC. Climate Change 2021: The Physical Science Basis. Contribution of Working Group I
to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change
(Cambridge Univ. Press, 2021).
7. Huppmann, D., Rogelj, J., Kriegler, E., Krey, V. & Riahi, K. A new scenario resource for
integrated 1.5 °C research. Nat. Clim. Change 8, 1027–1030 (2018).
8. Masson-Delmotte, V. et al. Global Warming of 1.5 °C. An IPCC Special Report on the
Impacts of Global Warming of 1.5 °C Above Pre-Industrial Levels and Related Global
Greenhouse Gas Emission Pathways, in the Context of Strengthening the Global Response
to the Threat of Climate Change, Sustainable Development, and Efforts to Eradicate
Poverty https://www.ipcc.ch/sr15 (IPCC, 2018).
9. Stott, P. A. et al. Attribution of extreme weather and climate‐related events. Wiley
Interdiscip. Rev. Clim. Change 7, 23–41 (2016).
10. IRENA. Renewable Power Generation Costs in 2020 (International Renewable Energy
Agency, 2021).
11. Hausfather, Z. & Peters, G. P. Emissions – the ‘business as usual’story is misleading. Nature
577, 618–620 (2020).
12. UNFCCC. The Paris Agreement (UNFCCC, 2015); https://unfccc.int/process-and-
meetings/the-paris-agreement/the-paris-agreement
13. UNFCCC. Common Time Frames for Nationally Determined Contributions Referred to in
Article 4, Paragraph 10, of the Paris Agreement https://unfccc.int/sites/default/files/
resource/cma3_auv_3b_CTF.pdf (UNFCCC, 2021).
14. Modi, N. National Statement by Prime Minister Shri Narendra Modi at COP26 Summit in
Glasgow https://www.mea.gov.in/Speeches-Statements.htm?dtl/34466/National+State
ment+by+Prime+Minister+Shri+Narendra+Modi+at+COP26+Summit+in+Glasgow (Ministry
of External Affairs, Government of India, 2 November 2021).
15. IMO. Adoption of the Initial IMO Strategy on Reduction of GHG Emissions from Ships and
Existing IMO Activity Related to Reducing GHG Emissions in the Shipping Sector
https://unfccc.int/sites/default/files/resource/250_IMO%20submission_Talanoa%20
Dialogue_April%202018.pdf (International Maritime Organization, 2018).
16. ICAO. Resolution A40-18: Consolidated Statement of Continuing ICAO Policies and
Practices Related to Environmental Protection – Climate Change https://www.icao.int/
environmental-protection/Documents/Assembly/Resolution_A40-18_Climate_Change.
pdf (International Civil Aviation Organization, 2019).
17. Zimm, C. & Nakicenovic, N. What are the implications of the Paris Agreement for
inequality? Clim. Policy 20, 458–467 (2020).
18. Günther, A., Gütschow, J. & Jeffery, M. L. NDCmitiQ v1.0.0: a tool to quantify and analyse
greenhouse gas mitigation targets. Geosci. Model Dev. 14, 5695–5730 (2021).
19. UNEP. Addendum to the Emissions Gap Report 2021 https://wedocs.unep.org/bitstream/
handle/20.500.11822/37350/AddEGR21.pdf (United Nations Environment Programme,
2021).
20. Climate Analytics & NewClimate. Climate Action Tracker. Warming Projections Global
Update: November 2021 https://climateactiontracker.org/documents/997/CAT_2021-11-
09_Briefing_Global-Update_Glasgow2030CredibilityGap.pdf (Climate Analytics,
NewClimate, 2021).
21. Gütschow, J., Jeffery, M. L., Schaeffer, M. & Hare, B. Extending near-term emissions
scenarios to assess warming implications of Paris Agreement NDCs. Earth’s Future 6,
1242–1259 (2018).
22. Jeffery, M. L., Gütschow, J., Rocha, M. R. & Gieseke, R. Measuring success: improving
assessments of aggregate greenhouse gas emissions reduction goals. Earth’s Future 6,
1260–1274 (2018).
23. Sognnaes, I. et al. A multi-model analysis of long-term emissions and warming
implications of current mitigation efforts. Nat. Clim. Change 11, 1055–1062 (2021).
24. Meinshausen, M. et al. Multi-gas emission pathways to meet climate targets. Clim.
Change 75, 151–194 (2006).
25. Lamboll, R. D., Nicholls, Z. R. J., Kikstra, J. S., Meinshausen, M. & Rogelj, J. Silicone v1.0.0:
an open-source Python package for inferring missing emissions data for climate change
research. Geosci. Model Dev. 13, 5259–5275 (2020).
26. Friedlingstein, P. et al. Global carbon budget 2021. Preprint at Earth System Science Data
https://doi.org/10.5194/essd-2021-386 (2021).
27. UNFCCC. Nationally Determined Contributions Under the Paris Agreement – Revised
Synthesis Report by the Secretariat. Report no. FCCC/PA/CMA/2021/8/Rev.1, 42
https://unfccc.int/process-and-meetings/the-paris-agreement/nationally-determined-
contributions-ndcs/nationally-determined-contributions-ndcs/ndc-synthesis-report
(UNFCCC, 2021).
28. Forster, P. et al. in Climate Change 2021: The Physical Science Basis. Contribution of
Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on
Climate Change (eds Masson-Delmotte, V. et al.) Ch. 7 (Cambridge Univ. Press, 2021).
29. Collins, M. et al. in Climate Change 2013: The Physical Science Basis. Contribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change (eds Stocker, T. F. et al.) 1029–1136 (Cambridge Univ. Press, 2013).
30. Rogelj, J. et al. A new scenario logic for the Paris Agreement long-term temperature goal.
Nature 573, 357–363 (2019).
31. Riahi, K. et al. Cost and attainability of meeting stringent climate targets without
overshoot. Nat. Clim. Change 11, 1063–1069 (2021).
32. European Commission. Launch by United States, the European Union, and partners of the
Global Methane Pledge to keep 1.5C within reach. European Commission Press Corner
https://ec.europa.eu/commission/presscorner/detail/en/statement_21_5766 (2 November
2021).
33. IEA. World Energy Outlook 2021: Technical Note on the Emissions and Temperature
Implications of COP26 Pledges https://iea.blob.core.windows.net/assets/aa17bd09-2ad0-
4d0a-b5aa-ee418900c4af/Theimpactsofnewemissionspledgesonlongtermtemperatures.
pdf (International Energy Agency, 2021).
34. Gütschow, J., Günther, A. & Pflüger, M. The PRIMAP-hist national historical emissions time
series (1750–2019) v2.3.1. Zenodo https://doi.org/10.5281/zenodo.5494497 (2021).
35. IEA. World Energy Outlook 2020 https://www.iea.org/reports/world-energy-outlook-2020
(International Energy Agency, 2020).
36. World Bank. World Bank Country and Lending Groups - Country Classification by Income
(World Bank, 2021); https://datahelpdesk.worldbank.org/knowledgebase/articles/9065
19-world-bank-country-and-lending-groups (11 November 2021)
37. Gütschow, J. et al. The PRIMAP-hist national historical emissions time series. Earth Syst.
Sci. Data 8, 571–603 (2016).
38. Houghton, J. T. et al. (eds) Revised 1996 Guidelines for National Greenhouse Gas
Inventories https://www.ipcc-nggip.iges.or.jp/public/gl/invs1.html (IPCC, 1996).
39. Gütschow, J., Jeffery, M. L., Günther, A. & Meinshausen, M. Country-resolved combined
emission and socio-economic pathways based on the Representative Concentration
Pathway (RCP) and Shared Socio-Economic Pathway (SSP) scenarios. Earth Syst. Sci. Data
13, 1005–1040 (2021).
40. King, L. C. & van den Bergh, J. C. J. M. Potential carbon leakage under the Paris
Agreement. Clim. Change 165, 52 (2021).
41. Gidden, M. J. et al. A methodology and implementation of automated emissions
harmonization for use in Integrated Assessment Models. Environ. Modelling Softw. 105,
187–200 (2018).
42. Grassi, G. et al. Critical adjustment of land mitigation pathways for assessing countries’
climate progress. Nat. Clim. Change 11, 425–434 (2021).
43. Meinshausen, M., Raper, S. C. B. & Wigley, T. M. L. Emulating coupled atmosphere–ocean
and carbon cycle models with a simpler model, MAGICC6 – Part I – model description
and calibration. Atmos. Chem. Phys. 11, 1417–1456 (2011).
44. Nicholls, Z. et al. Reduced Complexity Model Intercomparison Project Phase 2:
synthesizing Earth system knowledge for probabilistic climate projections. Earth’s Future
9, e2020EF001900 (2021).
45. UN. United Nations World Population Prospects 2019, Online Edition, Rev. 1 https://
population.un.org/wpp/Download/Standard/Population (United Nations, 2019).
46. Gössling, S. & Humpe, A. The global scale, distribution and growth of aviation:
implications for climate change. Global Environ. Change 65, 102194 (2020).
47. Gulev, S. K. et al. in Climate Change 2021: The Physical Science Basis. Contribution of
Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on
Climate Change (eds V. Masson-Delmotte et al.) Ch. 2 (Cambridge Univ. Press,
2021).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2022
Nature | Vol 604 | 14 April 2022 | 309
Article
Methods
Emission data aggregation
We use the updated compilation data product PRIMAP-hist37, version
2.3.1 (ref. 34) to infill UNFCCC-submitted inventory data. When quan-
tifying the NDCs, we take into account which gases (CO2, CH4, N2O,
hydrofluorocarbons, perfluorocarbons, SF6) and IPCC 1996 guideline38
sectors (energy; industry, waste and other; agricultural; LULUCF) are
covered by the NDCs, finding that 24 countries increased the coverage
of their total GHG emissions compared to the initial NDCs under the
Paris Agreement.
Countries state their mitigation targets in NDCs in different ways.
Some are, for example, formulated as reductions below a stated or
assumed business-as-usual scenario, others are intensity targets.
Increasingly, countries state their NDCs as absolute emission levels or
reductions in relation to a fixed historical emission level. We implement
those targets on the covered sectors and gases. Emissions not covered
by an NDC target are assumed to follow a country-downscaled SSP5
reference scenario39. Even if considered less and less likely at the global
level11, this downscaled higher-emission SSP5 reference scenario was
chosen because of considerable uncertainty in downscaled scenarios
(we choose to err on the higher-emission side given the focus of this
study) and because carbon leakage could put upward pressure on emis-
sions by moving emission-intensive activities to countries and sectors
not covered by NDCs40. These country-level whole-economy emission
quantifications in line with the submitted NDC are then aggregated
to the global level, adding emissions from international aviation and
shipping. These aggregations were performed each time a new NDC
was submitted, aggregating all NDCs submitted up to that point in
time, resulting in 354 sets of scenarios. As a default assumption, the
emission scenarios shown in Fig. 4 exclude hot air but represent the
four options of full implementation and conditional implementation
with their respective low and high emission ranges (where NDCs are
provided as ranges or their quantification is uncertain), times the 354
individual emission aggregations for each time that an initial or new
or updated NDC was submitted, resulting in a total of 1,416 scenarios.
To construct global gas-by-gas emission scenarios for greenhouse
gases and aerosol compounds reflective of the global pledges at each
point in time when an NDC was submitted or updated, we use infill-
ing, harmonization (Aneris, version 0.2.0+41.g6ddf6e5)41 and exten-
sion algorithms (Silicone, version 1.2.1)25. An adapted version of the
equal-quantile-walk approach24,25 was used here, which separates GHG
emission trajectories into corresponding gases in line with the SR1.5
database. Individual GHG gas distributions (CH4, N2O, and so on) from
the SR1.5 database were aggregated along the same percentiles for each
year using 100-year time horizon Global Warming Potential (GWP-100)
metric values from IPCC WG1 AR66, thereby creating a GHG distribu-
tion that permits a look-up operation in reverse, that is, looking up
individual gas emission levels in each year for a particular input GHG
emission level.
Extrapolation choices
Including long-term targets in the analysis limits the effect of subjec-
tive extrapolation choices, as 75% of 2020 emissions are covered by
long-term targets mentioned in NDCs, LT-LEDS and India’s net-zero
announcement (interpreted here conservatively as net-zero CO2)
(Fig. 1). A range of country-level extrapolation choices from 2030 to
2050 for the countries without long-term targets yield a range of ±0.1 °C
in the median peak warming projection (scenarios ‘A-constant’, ‘A-rate’
and ‘A’ in Extended Data Table 2). As sensitivity cases, we considered
continuous reduction rates from 2020 to 2030 (A-rate), constant 2030
emission levels (A-constant), and country-downscaled SSP1-reference
projections growth rates (A)39.
The scenarios that investigate temperatures based on various NDC
quantifications and their implied 2030 emission levels and that do not
consider the long-term targets (scenarios ‘a’ to ‘h’ in Extended Data
Table 2) are extended by continuing global 2025–2030 growth rates
up to 2050 (Fig. 3b).
In all our scenarios (Extended Data Table 2), the extrapolation of
emissions beyond 2050 is based on the SR1.5 scenario database7 and the
equal-quantile-walk approach and hinges upon assumptions underly-
ing the used 411 scenarios in that SR1.5 database. A recent change in the
scenario design paradigm30, which tends to lead to earlier reductions
and less-negative emissions in the second half of the century, could
affect the temperature evolution of the extrapolated scenarios. We
compare our scenario A post-2050 extension to selected scenarios
under this new framework31, which share similar 2050 GHG emission
levels (we choose the 25 closest scenarios to the 2050 Kyoto-GHG GWP-
100 AR6 emission levels of scenario A). We find that our post-2050
extensions of GHG emissions turn out to be on the higher side (hence
again, erring on the higher temperature side given the focus of this
study) (Extended Data Fig. 2).
Sensitivity cases
We distinguish eight cases for 2030 NDCs, namely the permutations of
low-end and high-end quantifications (where NDCs are either uncer-
tain or provided as a range), including hot air or excluding hot air, and
considering only unconditional NDC elements or assuming the full
implementation of NDCs including all conditional elements (Fig. 3b and
Extended Data Table 2). The wide range of peak (in this case equivalent
to 2100) temperatures resulting from these eight cases indicates that a
single projection based on 2030 NDCs is not representative.
More specifically, the low-end and high-end quantifications mark
the upper and lower emission levels of the estimated emission range,
when quantifying the NDC of a country that is subject to unclear or
uncertain assumptions (for example, a lack of clarity about whether
a net-zero target refers to CO2 or GHGs), or for those countries that
provide a target range instead of a single target. Hot air is here defined
as the extent to which quantified NDC emission levels exceed those
implied by national emissions, when sectoral and gas specific SSP5
baseline (BL) growth rates39 are assumed. The target of any country
is considered as ‘conditional’ if it is stated as conditional on financial
or other international support. In this analysis, the long-term targets
beyond 2030 are not distinguished by their conditionality. In Fig. 2,
hot air is indicated for both the lower SSP1-BL and the higher SSP5-BL
reference scenarios39.
LULUCF sector accounting
In the LULUCF sector (here considered equivalent to CO2 ‘forestry and
other land use’emissions), countries tend to report all carbon fluxes on
managed land42, even if a large part of the carbon flux can be attributed
to CO2 fertilization and hence constitutes a natural sink in response to
anthropogenically elevated CO2 concentrations, rather than a directly
human-induced CO2 uptake. We account for this mismatch between
reported and directly anthropogenically induced sinks so that our
emission pathways represent directly anthropogenic emissions only.
Specifically, we add the SSP1-1.9 aligned estimates from supplementary
table 8 in ref. 42 to our bottom-up country-data aggregation to derive
global LULUCF emissions. Although those authors42 suggest adjusting
the integrated assessment model scenarios to match the officially
reported LULUCF data, we here pursue the opposite adjustments to
ensure comparability with the SR1.5 database of scenarios, similar to
the approach taken by the UNFCCC NDC Synthesis Report27.
Climate system uncertainties
To represent climate system and carbon cycle uncertainties in the trans-
lation from emissions to warming in line with the latest IPCC AR6 WG1
assessment, we use the probabilistic reduced complexity model MAG-
ICC743. Our climate emulator is calibrated to the assessed SSP warm-
ing ranges, ocean heat uptake and other key climate characteristics
(cross-chapter box 7.1 in IPCC WG1 AR6)28 with 600 ensemble members
that result in climate projections that represent the IPCC AR6 WG1
assessment most closely in terms of temperature projections from four
reduced-complexity models assessed in IPCC AR6 WG1. We previously
also used this IPCC AR6-consistent probabilistic set-up to compute
climate projections for the UNEP gap report3 and IEA World Energy
Outlook 2021 (WEO2021) projections4,33. The general techniques are in
close alignment with the community intercomparison project RCMIP44.
The computationally expensive part of our study was calculating proba-
bilistic climate projections for a total of 1,416 global aggregate emission
scenarios (resulting from 354 submissions of first or new or updated
NDCs, in four variations, namely conditional and full implementation
with their respective minimum–maximum ranges), resulting in a set
of 849,600 emulator runs.
Per capita emissions
World population data are taken from the UN Department of Economic
and Social Affairs45 and its medium variant projection. Per capita emis-
sions of international aviation are divided by 5% as a conservative high
estimate of the 2%–4% of the world population that is estimated to use
international aviation in a given (non-COVID) year46; per capita emis-
sions of international maritime bunkers are indicated as average for
the world population.
Present-day global mean temperatures
We refer to 1.2 °C as the ‘current’ global mean temperatures relative to
1850–1900 based on the IPCC WG1 AR647 assessment of 1.09 °C assessed
warming between 1850–1900 and the period 2011–2020, taking into
account a warming trend of approximately 0.2 °C per decade. Adjust-
ing for that decadal trend yields an approximate warming assessment
for around 2020 of 1.2 °C.
Global Methane Pledge quantification
In our bottom-up analysis, we consider the extent to which methane
emissions might be reduced beyond methane reductions implicit in
NDCs and long-term targets. Specifically, for the 104 countries (the
USA and the 103 signatory countries of the GMP as of mid-November
2021), we consider how 30% domestic methane reductions compare
to conditional and full implementation levels of NDCs, and how those
reductions—aggregated across the 104 countries—compare to a 30%
global reduction. The GMP does not specify a 30% reduction of meth-
ane in each signatory country, but in the absence of country-level GMP
commitments, our country-level bottom-up assessment is based on
that assumption. Methane emissions in each country in 2019 are taken
from officially reported and PRIMAP-hist country-reported (CR) data34.
Any sectoral methane emissions not covered by an NDC are assumed to
evolve according to the SSP5-BL reference scenario39. Methane emis-
sions covered by an NDC are subject to the GHG targets, taking into
account the fractional shift between CO2, CH4, N2O and hydrofluoro-
carbons, perfluorocarbons and SF6 emissions up to the target year39.
The sum of the NDC-covered and non-covered methane emissions are
then compared to 2019 emission levels (which is used as a proxy for
2020) for both the unconditional and full-implementation NDC cases.
The extent to which a 30% methane reduction yields lower methane
emissions in 2030 than in the NDC scenarios is then considered to be
the additional country-level methane reduction due to the GMP.
Data availability
The data archive for this study is available at https://doi.org/10.5281/
zenodo.5886866, which includes the scenario data, temperature
outcomes and supplementary tables48. Country-level emission
quantifications of NDCs and LT-LEDs (the ‘NDC Factsheets’)49 are
available at https://doi.org/10.5281/zenodo.5768669 and are also
interactively available at https://climate-resource.com/tools/ndcs.
The source data PRIMAP-hist CR is provided at https://doi.org/10.5281/
zenodo.5494497. Source data are provided with this paper.
Code availability
The analysis code used in this study is provided by Silicone (version
1.2.1) for the infilling, available at https://github.com/GranthamImpe-
rial/silicone, and by the Anaeris package (version 0.2.0+41.g6ddf6e5)
for the harmonization, available at https://github.com/iiasa/aneris.
The climate model MAGICC7.5 is in the process of becoming open
source (re-factoring ongoing), the model is available as an interactive
mode at http://live.magicc.org. The source code for extrapolating
and infilling of the NDC scenarios are available at https://gitlab.com/
climate-resource/ndc-realisations-2021.
48. Meinshausen, M. et al. Data for study “Realisation of Paris Agreement climate pledges
may limit warming just below 2 °C”. Zenodo https://doi.org/10.5281/zenodo.5886866
(2022).
49. Meinshausen, M., Lewis, J., Nicholls, Z. & Guetschow, J. NDC factsheets. Zenodo https://
doi.org/10.5281/zenodo.5768669 (2021).
50. Climate Action Tracker. Temperatures https://climateactiontracker.org/global/
temperatures/ (2021).
Acknowledgements We acknowledge discussions and collaborations with several colleagues
of the UNEP Gap report team and the IAM modellers contributing to the SR1.5 and ENGAGE
scenario databases. M.M. is the recipient of an Australian Research Council (ARC) Future
Fellowship (grant number FT130100809). C.M., L.C. and B.H. state that the views expressed in
this article are in the author’s personal capacity and do not necessarily reflect the views of the
United Nations, United Nations Climate Change or the International Energy Agency.
Author contributions All authors contributed to writing the study. M.M. designed the study
and quantified the NDCs for the bottom-up NDC analysis; J.L. performed the global
harmonization, climate model runs and analysis; C.M. and L.C. provided the analysis of
net-zero emission targets and the IEA scenarios; Z.N. provided the code base for the
probabilistic climate model analysis and analysis of the ENGAGE database; and J.G. provided
the PRIMAP-hist data.
Competing interests The authors declare no competing interests.
Additional information
Supplementary information The online version contains supplementary material available at
https://doi.org/10.1038/s41586-022-04553-z.
Correspondence and requests for materials should be addressed to Malte Meinshausen.
Peer review information Nature thanks Kalyn Dorheim, Zeke Hausfather and the other,
anonymous, reviewer(s) for their contribution to the peer review of this work.
Reprints and permissions information is available at http://www.nature.com/reprints.
Article
Total CO2 emissions
70

60

50
CO2 emissions (Gt CO2 / yr)


40

30


20 
  
10  


0

-10
2000 2020 2040 2060 2080 2100

Extended Data Fig. 1 | CO2 emissions for various scenarios. Total CO2
emissions of the considered scenarios A and B, the 2030 extension scenarios ‘a’
to ‘h’ (Extended Data Table 2), the SSP scenarios (dashed grey lines) and the
SR1.5 scenario database. As in Fig. 3b, but for total CO2 emissions instead of
GHG emissions.
 50 The ENGAGE
This study’s
GHG emissions excl. LULUCF (GtCO2eq / yr, GWP-100 AR6)
database does
Scenario ‘A’ not suggest
a systematic low
bias of extrapolated
40 scenario ‘A’ post-2050
emission levels
given it’s 2050
level of emissions.
For a similar level
30 of 2050 emissions,
ENGAGE scenarios
tend in fact to suggest
lower post-2050 emissions.
20 25 ENGAGE database
scenarios closest
to scenario A’s
2050 GHG emission
10 level (Riahi et al. 2021)

-10

2010 2020 2030 2040 2050 2060 2070 2080 2090 2100

Extended Data Fig. 2 | Sensitivity test of database temperature projections respective emission level around 30 GtCO2e by 2050, more of the selected
in the ENGAGE database31. The 25 closest scenarios in terms of their 2050 GHG ENGAGE scenarios suggest lower post-2050 GHG emission levels than our
emissions (GWP-100 AR6, excluding LULUCF) to this study’s scenario A (thick scenario A, which had been extrapolated using an equal-quantile-walk
blue line, Extended Data Table 2) are selected (thin coloured lines). For the approach on the basis of the SR1.5 scenario database7.
Article
Extended Data Table 1 | 2100 warming relative to 1850–1900 of comparison studies shown in sidebars of Fig. 3a, c

Projected 2100 Warming (°C

rel. 1850-1900)
Label Source of Paris Agreement pledge analysis
Median Lower Upper

Quantification of 2030 NDC targets as well as long-term targets

1a UNEP Gap Report, November update 2021 Unconditional NDCs + Net-zero (UNEP, 2021, ref. 1.9 1.8 2.1
19), median climate projection, range reflecting extension uncertainty
1b UNEP Gap Report, November update 2021 Unconditional NDCs + Net-zero (UNEP, 2021, ref. 1.9 2.5
19), range reflecting geophysical median and upper 90% uncertainty
2a UNEP Gap Report, November update 2021 Conditional NDCs + Net-zero (UNEP, 2021, ref. 19), 1.8 1.7 2.0
median climate projection, range reflecting extension uncertainty
2b UNEP Gap Report, November update 2021 Conditional NDCs + Net-zero (UNEP, 2021, ref. 19), 1.8 2.4
range reflecting geophysical median and upper 90% uncertainty
APS IEA Announced Pledges Scenario, NDCs + all long-term net-zero announcements33, range 1.8 1.3 2.6
reflecting 5%-95% geophysical uncertainty
4 CAT Optimistic scenario: Best case scenario assuming full implementation of all announced 1.8 1.5 2.4
targets including net zero targets, LT-LEDs and NDCs without Hot Air (ref. 20)
5 CAT Pledges and targets scenario: Full implementation of unconditional submitted and binding 2.1 1.7 2.6
long-term targets and 2030 NDC targets, without Hot Air (ref. 20), with range indicating 68%
geophysical uncertainty range, i.e., 16th to 84th percentile. Assuming implementation of the more
ambitious end of the pledges, including those that are conditional on finance and where a range
is provided, would limit median warming to 1.9°C (footnote 3 in ref. 50).

Quantification of 2030 targets only with extensions beyond 2030.

6 UNEP Gap Report, October 2021 Unconditional NDCs (UNEP, 2021, ref. 3), median climate 2.5 2.0 2.9
projection, range reflecting extension uncertainty
7 UNEP Gap Report, October 2021 Conditional NDCs (UNEP, 2021, ref. 3), median climate 2.4 1.9 2.8
projection, range reflecting extension uncertainty

8 CAT “2030 targets only” NDCs (CAT, 2021, ref. 20), with range indicating 68% geophysical 2.4 1.9 3.0
uncertainty range, i.e., 16th to 84th percentile.
9 Sognnaes et al. (2021), 2100 temperature range under NDC and CP scenarios (and individual 2.2 to 2.9
studies as provided in data for Figure 1c in ref. 23.
STE IEA Stated Policies Scenario, sectoral analysis of existing and under development energy 2.6 1.9 3.7
policies and measures4 – with uncertainties reflecting 5-95% geophysical uncertainties.
10 UNEP “Current policies” scenario and best-estimate (50%) warming projection (with the 2.6 2.1 3.0
uncertainty reflecting the methodological choices for emission extensions beyond 2030 (ref. 3)).
11 CAT “Current Policies” scenario as of November 2021; with range indicating 68% geophysical 2.7 2.0 3.6
uncertainty range, i.e., 16th to 84th percentile (ref. 20).

Refs. 3,4,19,20,23,33,50.
Extended Data Table 2 | Overview of shown pathways and their characteristics
Label Ambition Country- Hot Air Global Exceedance Probability Peak Warming (°C) Year of Peak warming 2100 Warming (°C)
Level – level (E = Methane for temperature level (%) (Year)
Condition extension excl.; I. Pledge 1.5C 2.0C 2.5C 3.0C 5% Median 95% 5% Median 95% 5% Median 95%
ality 2030-2050 = incl.)
(C=full
implement
ation; U =
unconditio
nal)
A High-C SSP1BL E. n/a 90% 42% 12% 2% 1.41 1.92 2.80 2049 2082 2100 1.33 1.88 2.80

B Low-U SSP1BL E n/a 94% 52% 17% 5% 1.47 2.03 3.00 2059 2094 2100 1.43 2.02 3.00

a High-C n/a E n/a 98% 69% 28% 11% 1.61 2.24 3.31 2093 2100 2100 1.59 2.24 3.31

b High-C n/a I n/a 99% 79% 39% 14% 1.70 2.36 3.50 2093 2100 2100 1.68 2.36 3.50

c Low-C n/a E n/a 99% 80% 41% 15% 1.72 2.39 3.54 2093 2100 2100 1.70 2.39 3.54

d Low-C n/a I n/a 99% 82% 47% 17% 1.76 2.46 3.64 2094 2100 2100 1.76 2.46 3.64

e High-U n/a E n/a 100% 87% 52% 20% 1.81 2.52 3.72 2100 2100 2100 1.81 2.52 3.72

f Low-U n/a E n/a 100% 94% 67% 31% 1.98 2.74 4.03 2100 2100 2100 1.98 2.74 4.03

g High-U n/a I n/a 100% 98% 78% 43% 2.12 2.90 4.26 2100 2100 2100 2.12 2.90 4.26

h Low-U n/a I n/a 100% 98% 82% 51% 2.20 3.01 4.39 2100 2100 2100 2.20 3.01 4.39

m High-C SSP1BL E full CH4 89% 40% 12% 2% 1.39 1.90 2.78 2049 2082 2100 1.32 1.87 2.78

n High-C SSP1BL E cond. 90% 42% 12% 2% 1.40 1.92 2.79 2049 2082 2100 1.33 1.88 2.79
CH4
o High-C SSP1BL E uncond. 89% 41% 12% 2% 1.40 1.91 2.79 2049 2082 2100 1.33 1.87 2.79
CH4
A-constant High-C constant E n/a 89% 36% 10% 1% 1.39 1.87 2.72 2048 2069 2100 1.26 1.79 2.71

A-hot air High-C SSP1BL I n/a 91% 45% 14% 3% 1.42 1.95 2.85 2049 2082 2100 1.36 1.92 2.85

A-rate High-C Constant E n/a 91% 45% 14% 3% 1.42 1.95 2.85 2049 2093 2100 1.36 1.92 2.85
2020-2030
rate
Pathways are characterized by (1) their ‘low’ or ‘high’ ambition (in the case of uncertain NDCs or provided target ranges, with ‘high ambition’ indicating the lower emission variant, and ‘low ambi-
tion’ indicating the higher emission variant), conditionality (C, full implementation, including conditional elements; U, unconditional); (2) the extension method from 2030 to 2050 (‘SSP1BL’
indicates using country-specific and gas-specific growth rates from ref. 38; ‘n/a’ indicates not applicable as country data only used up to 2030 and global extensions use 2025–2030 global
growth rates; ‘Constant 2020-2030 rate’ indicates assuming the constant 2020–2030 country-level emission rate; and ‘constant’ indicates constant 2030 levels of emissions up to 2050); (3)
the inclusion of ‘hot air’ (E, excluded; I, included); (4) alterations of the global methane emissions (‘full CH4’ indicates global CH4 reductions linearly ramped from 2020 levels to −30% by 2030
and constant thereafter or NDC scenario, whichever lower; ‘cond. CH4’ indicates scenario adjusted with sum of 30% country-level methane reductions that go beyond country-level conditional
NDC scenario; ‘uncond. CH4’ indicates same as ‘cond. CH4’ but using sum of country-level methane reductions below country-level unconditional NDC scenario). Also shown are peak exceed-
ance probabilities, peak warming, time of peak warming, and 2100 warming for the 5th and 95th percentiles as well as the median.