LETTER • OPEN ACCESS You may also like

- Soil microbial respiration adapts to higher

Plausible 2005–2050 emissions scenarios project and longer warming experiments at the
global scale
between 2 °C and 3 °C of warming by 2100 Lu Yang, Junxiao Pan, Jinsong Wang et
al.

- Optimizing sowing window and cultivar

To cite this article: Roger Pielke Jr et al 2022 Environ. Res. Lett. 17 024027 choice can boost China’s maize yield
under 1.5 °C and 2 °C global warming
Mingxia Huang, Jing Wang, Bin Wang et
al.

- Soil indigenous nutrients increase the

View the article online for updates and enhancements. resilience of maize yield to climatic
warming in China
Xi Deng, Yao Huang and Zhangcai Qin

This content was downloaded from IP address 137.75.80.24 on 24/08/2023 at 15:31

 Environ. Res. Lett. 17 (2022) 024027 https://doi.org/10.1088/1748-9326/ac4ebf

LETTER

Plausible 2005–2050 emissions scenarios project between 2 ◦C

and 3 ◦C of warming by 2100
OPEN ACCESS

RECEIVED
23 July 2021 Roger Pielke Jr1,∗, Matthew G Burgess1,2,3 and Justin Ritchie4
REVISED 1
21 January 2022 Department of Environmental Studies, University of Colorado Boulder, Boulder, CO, 80303, United States of America
2
Cooperative Institute for Research in Environmental Sciences, University of Colorado Boulder, Boulder, CO, 80309, United States of
ACCEPTED FOR PUBLICATION
25 January 2022
America
3
Department of Economics, University of Colorado Boulder, Boulder, CO, 80309, United States of America
PUBLISHED 4
Institute for Resources, Environment and Sustainability, University of British Columbia, 2202 Main Mall, Vancouver, BC V6T 1Z4,
11 February 2022
Canada
∗
Author to whom any correspondence should be addressed.
Original content from
this work may be used E-mail: [email protected]
under the terms of the
Creative Commons Keywords: scenarios, emissions, projections, plausible, climate
Attribution 4.0 licence.
Supplementary material for this article is available online
Any further distribution
of this work must
maintain attribution to
the author(s) and the title
of the work, journal Abstract
citation and DOI. Emissions scenarios used by the Intergovernmental Panel on Climate Change (IPCC) are central to
climate change research and policy. Here, we identify subsets of scenarios of the IPCC’s 5th (AR5)
and forthcoming 6th (AR6) Assessment Reports, including the Shared Socioeconomic Pathway
scenarios, that project 2005–2050 fossil-fuel-and-industry (FFI) CO2 emissions growth rates most
consistent with observations from 2005 to 2020 and International Energy Agency (IEA) projections
to 2050. These scenarios project between 2 ◦ C and 3 ◦ C of warming by 2100, with a median of
2.2 ◦ C. The subset of plausible IPCC scenarios does not represent all possible trajectories of future
emissions and warming. Collectively, they project continued mitigation progress and suggest the
world is presently on a lower emissions trajectory than is often assumed. However, these scenarios
also indicate that the world is still off track from limiting 21st-century warming to 1.5 ◦ C or
below 2 ◦ C.

1. Introduction The report of Working Group 1 of the IPCC

AR6 (IPCC 2021) observes that ‘No likelihood is
The Intergovernmental Panel on Climate Change attached to the scenarios assessed in this Report’
(IPCC) uses carbon dioxide (CO2 ) emissions scen- (p 1–102). Yet, at the same time the IPCC AR6 recog-
arios to imagine and evaluate ‘a range of plausible nizes that ‘the likelihood of high emissions scenarios
futures, because human development is determined such as RCP8.5 or SSP5-8.5 is considered low’ (IPCC
by a myriad of factors including human decision 2021, p 1–110) and recent ‘stated policy’ scenarios
making’ (IPCC WGIII 2014a, p 48). Scenarios used of groups such as the International Energy Agency
in the IPCC 5th Assessment Reports (AR5) (IPCC (IEA) are ‘approximately in line with the medium
2014) and 6th Assessment Reports (AR6; IPCC RCP4.5, RCP6.0 and SSP2-4.5 scenarios’ (IPCC 2021,
2021) include the ‘Representative Concentration p 1–110). The IPCC’s recent recognition of scenario
Pathways’ (RCPs) and the ‘Shared Socioeconomic likelihood is part of a broader and fast-moving dis-
Pathways’ (SSPs) (Moss et al 2010, Riahi et al 2017). cussion of scenario plausibility and implications for
The high-emissions end of the RCP/SSP range is research and policy (e.g. Ritchie and Dowlatabadi
consistent with 8.5 W m−2 radiative forcing and 2017, Pielke 2018, O’Neill et al 2020, Schwalm et al
4 ◦ C–5 ◦ C of warming relative to pre-industrial levels 2020, Hausfather and Peters 2020a, 2020b, Burgess
by 2100 (e.g. SSP5-8.5, RCP8.5) (Hausfather and et al 2021, Sognnaes et al 2021, Stammer et al 2021,
Peters 2020a). At the low end, scenarios project emis- Pielke and Ritchie 2021a, 2021b).
sions consistent with 1.9 W m−2 radiative forcing Scenario ‘plausibility’ has been discussed for
and 1.5 ◦ C of warming or slightly less by 2100 many years, but the IPCC has not evaluated the
(e.g. SSP1-1.9) (Hausfather and Peters 2020a). plausibility of scenarios central to much of climate

© 2022 The Author(s). Published by IOP Publishing Ltd

Environ. Res. Lett. 17 (2022) 024027 R Pielke Jr et al

research and assessment (see, Höök 2011, Wiek et al history and near-term IEA projections. Here, we build
2013, Walton et al 2019, Pielke and Ritchie 2021a, upon and update that analysis by comparing growth
2021b). Plausibility, as defined by Wiek et al (2013) rates in fossil fuel and industry (FFI) CO2 emissions
refers to ‘a variety of future states that are considered of IPCC and SSP baseline and policy scenarios to
“occurrable” (could happen)’ (p 137). This general identify scenarios that project growth rates in FFI CO2
definition may be operationalized in different, legit- emissions consistent with recent history (Global Car-
imate ways. Our analysis defines a ‘plausible’ scen- bon Project 2021, Ritchie and Roser 2021) and near-
ario as one in which future fossil-fuel-and-industry term IEA projections to 2050 (IEA 2021). We charac-
(FFI) CO2 emission growth rates of the scenario terize such scenarios as plausible. Our analysis does
show a consistency with observations and IEA Stated not associate individual scenarios with likelihoods or
Policies Scenario (STEPS) near-term projections. A probabilities. After identifying a subset of plausible
scenario that has already diverged from reality is, by scenarios from the larger set of available scenarios, we
definition, not ‘occurrable.’ It is theoretically pos- then assess what this subset of plausible scenarios pro-
sible for a scenario that has already diverged signific- jects to 2100 for future FFI CO2 emissions and asso-
antly from observations to later return to the same ciated global average temperature change.
projected future emissions level, though for this to
occur would imply opposite divergences in the future,
which would require a major departure from the tra- 2. Methods
jectory of the original scenario, thus making that
scenario implausible, even if it were to arrive at the We evaluate the 1184 AR5 scenarios (IPCC 2014,
same final level of emissions (see Burgess et al 2021). IPCC WGIII 2014b) and 127 SSP scenarios (Riahi
Here, we further this discussion by identifying et al 2017) with respect to projected FFI CO2 emis-
those scenarios of the AR5 and AR6 which are most sions growth rates, compared to observations (Global
in line with recent observations and near-term pro- Carbon Project 2021, Ritchie and Roser 2021) over
jections of FFI carbon dioxide emissions. The IPCC 2005–2020, and compared to 2021–2050 projections
has not conducted or assessed evaluations of scen- from the IEA’s 2021 World Energy Outlook (WEO)
ario plausibility, and until very recently little attention (IEA 2021). We focus on the AR5 and SSP databases,
has been paid to IPCC scenario plausibility in the lit- as these are central to the most recent IPCC assess-
erature (Pielke and Ritchie 2021a). Thus, our paper ment reports (IPCC 2014, 2021), and they cover the
addresses an important gap in climate research with full range of emissions pathways widely used in cli-
broad implications for research and policy. mate research. We do not include the scenario data-
Studies exploring future climate impacts typically base of the IPCC Special Report on Global Warm-
are based on a small number of scenarios as baselines ing of 1.5 ◦ C (IPCC 2018) in this analysis because
to compare with scenarios of more or less stringent it was designed based on a set of exploratory scen-
21st-century mitigation efforts (Pielke and Ritchie arios constrained to achieve the 1.5 ◦ C policy tar-
2021a). However, several recent studies have chal- get, and thus not equivalent in purpose or scope to
lenged whether commonly-used high-emission scen- the baseline and policy scenarios of the AR5 and SSP
arios (e.g. SSP5-8.5) are plausible baselines, due to databases.
their assumptions of rapid 21st-century coal expan- IEA STEPS projects that pledged climate and
sion (Ritchie and Dowlatabadi 2017, Hausfather and energy policies are enacted in varying degrees but
Peters 2020a, Burgess et al 2021, Pielke and Ritchie includes no new policies, akin to conventional ‘busi-
2021a, 2021b; see also Schwalm et al 2020, Hausfather ness as usual’ (BAU) baselines. Previous IEA WEOs
and Peters 2020b). Other studies have cast doubt on (e.g. IEA 2019) also included a second, slightly
the feasibility of rapid accelerations in decarboniza- more pessimistic, BAU-like Current Policies Scen-
tion, sufficient scaling of carbon-removal technolo- ario (CPS)—assuming no further policies would be
gies, or both, necessary to limit forcing to 1.9 W m−2 implemented beyond those already enacted. Both
(Anderson and Peters 2016, Rogelj et al 2016, Pielke CPS and STEPS CO2 emissions have been successively
2018, Liu and Raftery 2021). To ensure the rigor of cli- revised down in recent years, for reasons including
mate research designed to project plausible futures, it faster-than-projected renewable energy deployment
is thus important to continuously evaluate scenarios growth, slower coal growth, and slower economic
for plausibility. growth (Frankel 2011, de Resende 2014, Ritchie and
Burgess et al (2021) compared fundamental Dowlatabadi 2017, IEA 2019, 2020, 2021, Burgess et al
assumptions of IPCC and SSP baseline scenarios 2020, 2021). The IEA discontinued using CPS in 2020
(specifically: fossil-fuel CO2 emissions, population, due to the implausibility of a future with no new cli-
GDP per capita, energy intensity and carbon intens- mate policies (IEA 2020). IEA STEPS is commonly
ity) to data from 2005 to 2020 and to the most used as a current benchmark reference case projection
recent projections of the IEA to 2040 (IEA 2020). for plausible near-term emissions (e.g. Hausfather
That analysis identified a set of scenarios classified as 2021) and we follow this convention in this analysis
implausible because they were out of step with recent as well.

2
Environ. Res. Lett. 17 (2022) 024027 R Pielke Jr et al

We use IEA projections to 2050 as representing energy/GDP) and carbon intensity (CO2 emission-
current consensus expectations among energy experts s/primary energy), to explore the drivers of the FFI
for the near-term future. It is of course possible CO2 emission divergences (see SM).
that IEA projections will turn out to be inaccurate.
We use IEA projections as benchmark projections— 3. Results
consistently with previous studies (Hausfather and
Peters 2020a, Burgess et al 2021)—as the IEA’s pro- For the period 2005–2020, 64 AR5 (6%) and 7 SSP
jections are updated annually. If IEA’s projections of (6.6%) scenarios meet the ±0.1%/y error threshold,
near-term emissions continue to be revised down— and 243 AR5 (22.9%) and 23 SSP (21.9%) scen-
as they have in recent years (IEA 2019, 2020, 2021)— arios meet the ±0.3%/y threshold. For the period
then our present analysis would be biased towards 2005–2050, compared to the STEPS scenario, 26 AR5
including ‘plausible’ scenarios with higher emissions (2.4%) and 9 SSP (8.6%) scenarios meet the ±0.1%/y
and warming than was justified—meaning that scen- divergence threshold, and 85 AR5 (8%) and 19 SSP
arios currently judged as plausible may in the future (18.1%) scenarios meet the ±0.3%/y threshold. We
be judged implausible. Indeed, as the future unfolds, characterize these scenarios as ‘plausible’ because
we should fully expect that the subset of available their emissions trajectories are consistent with the
scenarios judged to be plausible shrinks in num- recent past and current reference projections of the
ber over time as reality constrains possibilities as the energy system’s expected evolution over the next three
future becomes the present. decades.
We consider both baseline (i.e. reference or no These plausible scenarios project a range of
policy) and policy scenarios of the AR5 and SSP futures to 2100, but as a group they are largely
databases (IPCC WGIII 2014b, Riahi et al 2017), consistent with continued decarbonization. Figure 1
and ask: Which scenarios have fossil-fuel CO2 emis- shows the FFI emissions in scenarios meeting the
sions growth rate errors and divergences of less than ±0.1%/y and ±0.3%/y divergence thresholds for
0.1%/y or 0.3%/y (where ‘y’ denotes year) over (a) 2005–2050. The SM contains corresponding figures,
2005–2020 (errors, compared to observations) and showing: fossil-fuel emissions only (figure S2); and
(b) 2005–2050 (divergences, compared to observa- FFI emissions of scenarios selected with: 2005–2020
tions and IEA 2021 STEPS) periods? We could not error filters (figure S3), 2005–2019 error filters (i.e.
assess divergences for 114 AR5 scenarios and 22 not including the pandemic effects; figure S4), and
SSP scenarios lacking 2005 numbers, and 11 AR5 2005–2050 divergences from 2005–2050 observations
scenarios lacking 2050 numbers, leaving a total of and IEA (2021) APS projections (figure S5). Of note,
1059 AR5 scenarios and 105 SSP scenarios in our for emissions in 2050 the median trajectory of the
main analysis. These growth rate errors and diver- subset of plausible scenarios (figure 1) is essentially
gences result in different levels of projected accur- identical to the independent projection of the IEA
acy over time: a ±0.1% (±0.3%) annual growth WEO (2021).
rate error compounded over 100 years results in The upper limit of plausible scenario ranges
∼±10% (∼±35%) overall projection error in annual with both IEA (2021) STEPS filters is similar, at
FFI CO2 emissions. As we show in supplement- slightly less than 40 gigatons (Gt) CO2 per year
ary materials (SM) figure S1 (available online at by 2100, and FFI CO2 emissions peaking before
stacks.iop.org/ERL/17/024027/mmedia), the ranges mid-century (figure 1). Both ±0.1%/y and ±0.3%/y
of 2100 warming found in plausible scenarios are divergence subsets also include scenarios with net
qualitatively insensitive to changes in the specific val- negative emissions (i.e. the removal of carbon diox-
ues chosen for these growth rate error and divergence ide from the atmosphere) in the latter half of the
thresholds. 21st century. Scenarios selected with the 2005–2019
As a robustness check regarding the COVID-19 and 2005–2020 error filters have higher upper lim-
recession and temporary drop in CO2 emissions in its: ∼60–80 Gt CO2 /y by 2100 with the 2005–2020 fil-
2020 (Le Quéré et al 2020), in the SM we compare ters (figure S3), and ∼70–100 Gt CO2 /y by 2100 with
2005–2020 growth rates from AR5 and SSP scenarios the 2005–2019 filters (figure S4). Scenarios selected
to 2005–2019 average growth rates. In the SM, as a with the 2005–2050 APS filter have lower upper lim-
further robustness check, we also compare the AR5 its: ∼20 Gt CO2 /y by 2100 (figure S5).
and SSP scenarios to the IEA (2021) new ‘Announced Figure 2 plots the plausible scenarios (using
Pledges Scenario’ (APS), which evaluates announced the 2005–2050 STEPS filters) in terms of
ambitions and targets, even when they have not been 2000–2100 cumulative CO2 emissions and associated
anchored in specific policies. Lastly, we compare end-of-century global mean temperature increase
errors and divergences of AR5 and SSP scenarios from (relative to the pre-industrial baseline). Temperature
observations, IEA STEPS, and IEA APS in each of the increases are only available for 524 of the AR5 scen-
Kaya Identity (Kaya and Yokoburi 1997) factors: pop- arios (but all of the SSPs). Figure 2 also shows the SSP
ulation, GDP per capita, energy intensity (primary marker scenarios and scenario ranges associated with

3
Environ. Res. Lett. 17 (2022) 024027 R Pielke Jr et al

Figure 1. The trajectories of all fossil-fuel-and-industry (FFI) CO2 emissions to 2100 among 1184 AR5 and 127 SSP scenarios
(IPCC WGIII 2014b, Riahi et al 2017), along with the IEA STEPS to 2050 (IEA 2021). Shaded regions indicate envelopes of
scenarios meeting ±0.1%/y (blue) and ±0.3%/y (gray) divergence tolerances in FFI CO2 emissions (relative to observations and
IEA projections). See also figure S1.

Figure 2. AR5 (IPCC WGIII 2014b) and SSP scenarios (Riahi et al 2017, IIASA 2018) plotted in terms of cumulative FFI CO2
emissions in the 21st century, and warming by 2100, relative to the pre-Industrial baseline. Major SSP scenario ranges (dashed
boxes) and marker scenarios are highlighted. Blue circles and light-blue triangles represent scenarios meeting the ±0.1%/y and
±0.3%/y divergence tolerances, respectively, in FFI CO2 emissions (relative to observations and IEA STEPS projections from 2005
to 2050). Figure S6 zooms in on the plausible scenarios.

different levels of radiative forcing in 2100 (1.9, 2.6, by 2100 (figures 2 and 3). High-emission baseline
3.4, 4.5, 6.0, 7.0, and 8.5 W m−2 ). scenarios of the SSPs—SSP3-7.0 (7.0 W m−2 , ∼4 ◦ C
Most of the plausible scenarios (62% with the of warming) and SSP5-8.5 (8.5 W m−2 , ∼5 ◦ C of
0.1%/y filter; 56.3% with the 0.3%/y filter) project warming)—lie far outside the envelope of plausible
emissions trajectories to 2100 that are consistent with scenarios (figure 2). None of the plausible scenarios
SSP 3.4 (3.4 W m−2 , 2.0 ◦ C–2.4 ◦ C of warming). All fall within the range of SSP 6.0 scenarios (figure 2; see
of the plausible scenarios fall between 2 ◦ C and 3 ◦ C also figure S6).

4
Environ. Res. Lett. 17 (2022) 024027 R Pielke Jr et al

Figure 3. Projected temperature increases (2100, compared to the pre-Industrial baseline) in sets of AR5 (IPCC WGIII 2014b)
and SSP scenarios (Riahi et al 2017, IIASA 2018) selected by the various filters used in our analysis, in terms of divergences in FFI
CO2 emission growth rates observed and projected by IEA (2021) STEPS or APS scenarios. Boxes denote the 25th to 75th
percentile ranges; white line denotes the medians; and whiskers denote the minima and maxima.

In contrast, the 2005–2019 and 2005–2020 fil- (see SM): population, GDP per capita, energy intens-
ters admit a small number of scenarios with greater ity (primary energy/GDP), and carbon intensity
than 3 ◦ C of warming, and the 2005–2050 APS fil- (CO2 emissions/primary energy). The AR5 and SSP
ter admits scenarios between 1.5 ◦ C and 2 ◦ C of scenarios our analysis identifies as plausible are more
warming (figure 3). Thus, the implausibility of >3 ◦ C consistent with observations and IEA (2021) STEPS
warming scenarios depends on the accuracy of IEA projections than other scenarios in carbon intens-
(2021) STEPS projections, and specifically its pro- ity (figure S8). Divergences in population and GDP
jection that global FFI carbon dioxide emissions will per capita were similar across scenarios satisfying and
at most plateau over the coming decades (figure 1). not satisfying the filters, and plausible scenarios often
Similarly, further progress in decarbonization bey- had negative energy intensity divergences compens-
ond what IEA (2021) STEPS scenario envisions could ating for the positive divergences in GDP per cap-
make sub-2-degree targets feasible. ita growth that exist across most scenarios (figure
These results are qualitatively insensitive to our S8) (see Burgess et al 2021). Similar patterns exist
choice of divergence threshold, as most scenarios in Kaya comparisons using other filters (2005–2019,
having 2100 warming outside the 2 ◦ C–3 ◦ C range 2005–2020, 2005–2050 APS; figures S9–S11).
have 2005–2050 divergences greater than 0.5%/y
(figure S1). The effect of the COVID-19 pandemic
for projections of near-term emissions trajectories is 4. Discussion
evident in the difference in warming ranges given by
the 2005–2019 and 2005–2020 filters (figures 3, S3 The subset of IPCC scenarios we identify as
and S4). ‘plausible’—most consistent with observations to
Warming projections associated with emissions 2020 and IEA (2021) projections to 2050 (figure 1)—
scenarios are also uncertain due to various climate are also consistent with a world currently well-
factors. To illustrate this, figure S7 shows ranges positioned for future climate policy success via con-
of temperature (Tebaldi et al 2021) and cumulative tinued implementation of policies that move the
emissions (Liddicoat et al 2021) consistent with SSP world closer to achieving or nearly achieving a 2 ◦ C
marker scenarios, based on the Coupled Model Inter- target by 2100 (figure 2; see also figures S1 and S6).
comparison Project (CMIP6). At the high end, these This finding is cautiously encouraging, but it is also
ranges allow for warming of >3 ◦ C by 2100 even if limited to the scope of previously published scen-
emissions stay within ranges of the plausible scen- arios, which will not capture full uncertainty about
arios. At the low end, plausible scenarios could allow future CO2 emissions or climate response. At the
warming lower than 2 ◦ C by 2100. same time our results are consistent with projec-
To assess drivers of divergences between AR5 tions of the implications of current policies for future
and SSP scenarios and observations and IEA (2021) emissions and global temperature changes projec-
scenarios, we assess growth rate divergences across ted by the Climate Action Tracker (2021) and the
the Kaya Identity (Kaya and Yokoburi 1997) factors United Nations Environment Programme’s (UNEP)

5
Environ. Res. Lett. 17 (2022) 024027 R Pielke Jr et al

(2021) Emissions Gap analyses. Our finding that the decarbonization of the global economy if the traject-
plausible scenarios fall within the 2 ◦ C–3 ◦ C warm- ory of currently plausible scenarios is to be achieved
ing range is also consistent with a recent analysis by or exceeded in the future.
Sognnaes et al (2021) that explores scenario uncer- It is notable that all of the baseline scenarios (often
tainties beyond the scenarios in the AR5 and SSP used as ‘business as usual’ or ‘reference’ scenarios
databases. in climate research; Pielke and Ritchie 2021a) over-
It is clear from this analysis that from the per- projected 2005–2050 emissions growth, relative to
spective of the scenarios of the IPCC, the world observations and IEA (2021) STEPS, by more than
is not presently far off-track from a trajectory of 0.1%/y (all but one, AIM/CGE’s SSP1 baseline, over-
FFI emissions envisioned in IPCC scenarios to be projected by 0.3%/y or more) (figure S1) (see also
consistent with achieving 2 ◦ C policy goals. This Burgess et al 2021). This suggests that climate research
conclusion is also consistent with the IPCC AR6 and policy are currently overly focused on implaus-
(2021) recently identifying high-emission scenarios ibly pessimistic scenarios of the future (Hausfather
(e.g. SSP5-8.5) to be low likelihood and mid-range and Peters 2020a, Pielke and Ritchie 2021a, 2021b).
scenarios (e.g. SSP2-4.5) as more likely. If scenario Relying on implausible scenarios can mislead policy
developers of several decades ago were to have known analyses. For instance, using baseline scenarios that
actual emissions trajectories to 2020 and projected in over-estimate near-term emissions requires assuming
2021 for 2050, they likely would have been encour- a need for unnecessary amounts of late-21st century
aged by the path the world was on, as compared to the deployment of carbon removal technologies in policy
full set of possibilities that they had then envisioned scenarios.
across the full set of scenarios. It is also notable that the vast majority of scen-
Importantly, in the scenarios our analysis identi- arios that project futures to 2100 failed our simple cri-
fies as plausible, future decarbonization rates accel- teria of plausibility by 2020, even though they were
erate relative to the present, and many include sub- developed in recent years and decades. This raises
stantial deployment of carbon removal technologies questions about the appropriate use of long-term
in the latter half of the century, the feasibility of which scenarios as projections of plausible futures, rather
our analysis does not assess. Comparing figures 1 than as exploratory tools (Bankes 1993), and sug-
and S2, we see that carbon removal has little effect gests a need for policy-relevant scenarios that are
on the high end of the range of emissions in scen- updated much more regularly with new observational
arios we identify as plausible, but—unsurprisingly— information, similarly to the IEA’s near-term scen-
reduces the lower range of the scenario envelope. arios (Burgess et al 2021, O’Neill et al 2020, Pielke and
In 2100, the median carbon emissions created from Ritchie 2021a).
fossil-fuel combustion (i.e. which does not account
for carbon capture or removal) of the plausible scen- 5. Conclusion
arios are ∼10 GtCO2 /y (figure S2), compared to zero
when the effects of carbon removal technologies are The IEA (2021) STEPS near-term projections of CO2
applied (figure 1). A median of ∼10 GtCO2 /y from emissions support expectations for the next several
fossil combustion by 2100 would reflect greater than decades of a long-plateau in CO2 emissions and
two-thirds reduction from today’s levels. Thus, in projections of less than 3 ◦ C of warming by 2100
the scenarios our analysis identifies as most consist- (see, Hausfather and Peters 2020a). Such expecta-
ent with global energy system developments toward tions are independently supported by the envelope
mid-century, large-scale carbon removal is necessary of scenarios identified as plausible in our analysis
for achieving net-zero emissions this century as fossil (figures 1–3). Of course, the future is uncertain, con-
technologies are projected to continue to play such a tingent on policy choices and expectations that will
substantial role in the energy system. evolve as we move into that future.
Carbon removal technologies presently do not Deep decarbonization remains an enormous
exist at scale (IEA 2020), and their future technical challenge, and net-zero CO2 emissions by 2050—a
and political plausibility has been questioned (e.g. common policy goal—remains outside the envelope
Anderson and Peters 2016). Liu and Raftery (2021) of even the plausible scenario trajectories (figure 1).
show that countries must increase their decarboniz- Of course, decision makers around the world could
ation rates by 80% relative to Paris commitments to choose to intentionally grow CO2 emissions, but
limit warming to 2 ◦ C by 2100. Similarly, if the pace that currently seems highly unlikely based on ana-
of global decarbonization fails to keep up with IEA lyses such as the UNEP (2021) Emission Gap report.
(2021) STEPS projections, scenarios having greater Our analysis suggests that the world thus sits in an
than 3 ◦ C warming by 2100 would again become enviable position to take on the challenge of deep
plausible (figure 3). However, observed decarboniz- decarbonization, at least as compared to where IPCC
ation has exceeded IEA projections in recent years; baseline scenarios and some of the public discourse
see IEA (2019, 2020, 2021). Thus, our findings under- projected the world to be in 2021 (Pielke and Ritchie
score the urgency of continued efforts to accelerate 2021a). To support continued efforts to achieve deep

6
Environ. Res. Lett. 17 (2022) 024027 R Pielke Jr et al

decarbonization, climate research and policy depend Hausfather Z and Peters G 2020a Emissions—the ‘business as
on the development and regular update of plausible usual’ story is misleading Nature 577 618–20
Höök M 2011 Future coal production outlooks in the IPCC
scenarios.
Emission Scenarios: are they plausible? Energy Environ.
22 837–57
Data availability statement IEA 2020 World Energy Outlook 2020 (IEA)
IEA 2021 World Energy Outlook 2021 (IEA)
Data and code used in our analysis are available Intergovernmental Panel on Climate Change (IPCC) 2014
from https://github.com/mattgburgess/climatekaya. Climate Change 2014: Synthesis Report. Contribution of
Working Groups I, II and III to the Fifth Assessment Report of
All other data needed to evaluate the conclusions in
the Intergovernmental Panel on Climate Change (IPCC)
the paper are present in the paper or the supplement- International Energy Agency (IEA) 2019 World energy outlook
ary materials. 2019 (IEA)
The data that support the findings of this study International Institute for Applied Systems Analysis (IIASA) 2018
are openly available at the following URL/DOI: SSP database (shared socioeconomic pathways)—version
2.0 (available at: https://tntcat.iiasa.ac.at/SspDb/dsd)
https://github.com/mattgburgess/climatekaya. IPCC WGIII 2014b AR5 scenario database version 1.0.2, (IIASA)
(available at: https://tntcat.iiasa.ac.at/AR5DB/dsd)
Acknowledgments IPCC 2018 Special report on global warming of 1.5 ◦ C(IPCC)
IPCC 2021 Climate change 2021: the physical science basis.
Contribution of Working Group I to the Sixth Assessment
We thank Zeke Hausfather, Oliver Geden, Glen Report of the Intergovernmental Panel on Climate Change
Peters, Ken Caldeira, Richard Betts, Chris Green and (IPCC)
Ken Rice for comments on an earlier version of the Kaya Y and Yokoburi K 1997 Environment, Energy, and Economy:
Strategies for Sustainability (Tokyo: United Nations
manuscript. We acknowledge funding from the Uni-
University Press)
versity of Colorado Boulder (start-up grant to M G B) Le Quéré C et al 2020 Temporary reduction in daily global CO2
and The University of British Columbia and the emissions during the COVID-19 forced confinement Nat.
Canada Research Chairs Program (J R). Clim. Change 10 647–53
Liddicoat S K et al 2021 Compatible fossil fuel CO2 emissions in
the CMIP6 earth system models’ historical and shared
ORCID iDs socioeconomic pathway experiments of the twenty-first
century J. Clim. 34 2853–75
Roger Pielke Jr  https://orcid.org/0000-0002-9617- Liu P R and Raftery A E 2021 Country-based rate of emissions
8764 reductions should increase by 80% beyond nationally
determined contributions to meet the 2 ◦ C target Commun.
Matthew G Burgess  https://orcid.org/0000-0002- Earth Environ. 2 1–10
3750-4347 Moss R H et al 2010 The next generation of scenarios for climate
Justin Ritchie  https://orcid.org/0000-0001-6679- change research and assessment Nature 463 747–56
3953 O’Neill B et al 2020 Achievements and needs for the climate
change scenario framework Nat. Clim. Change 10 1074–84
Pielke R Jr 2018 Opening up the climate policy envelope Issues Sci.
References Technol. 34 30–36
Pielke R Jr and Ritchie J 2021a Distorting the view of our climate
Anderson K and Peters G 2016 The trouble with negative future: the misuse and abuse of climate pathways and
emissions Science 354 182–3 scenarios Energy Res. Soc. Sci. 72 101890
Bankes S 1993 Exploratory modeling for policy analysis Oper. Res. Pielke R and Ritchie J 2021b How climate scenarios lost touch
41 435–49 with reality: a failure of self-correction in science has
Burgess M G, Langendorf R E, Ippolito T and Pielke R Jr 2020 compromised climate science’s ability to provide plausible
Optimistically biased economic growth forecasts and views of our collective future Issues Sci. Technol.
negatively skewed annual variation SocArXiv vndqr 37 74–84
Burgess M G, Ritchie J, Shapland J and Pielke R Jr 2021 IPCC Riahi K et al 2017 The shared socioeconomic pathways and their
baseline scenarios have over-projected CO2 emissions and energy, land use, and greenhouse gas emissions
economic growth Environ. Res. Lett. 16 014016 implications: an overview Glob. Environ. Change
Climate Action Tracker 2021 2100 warming projections 42 153–68
(available at: https://climateactiontracker.org/global/ Ritchie H and Roser M 2021 CO2 and greenhouse gas emissions
temperatures/) OurWorldInData.org (available at: https://ourworldindata.
de Resende C 2014 An assessment of IMF medium-term forecasts org/co2-and-other-greenhouse-gas-emissions)
of GDP growth IEO Background Paper No. BP/14/01 Ritchie J and Dowlatabadi H 2017 Why do climate change
(Washington: Independent Evaluation Office of the IMF) scenarios return to coal? Energy 140 1276–91
Frankel J 2011 Over-optimism in forecasts by official budget Rogelj J, den Elzen M, Höhne N, Fransen T, Fekete H, Winkler H,
agencies and its implications Oxford Rev. Econ. Policy Schaeffer R, Sha F, Riahi K and Meinshausen M 2016 Paris
27 536–62 agreement climate proposals need a boost to keep warming
Global Carbon Project 2021 Supplemental data of global carbon well below 2 ◦ C Nature 534 631–9
budget 2021 (version 1.0) (Global Carbon Project) (https:// Schwalm C R, Glendon S and Duffy P B 2020 RCP8.5 tracks
doi.org/10.18160/gcp-2021) cumulative CO2 emissions Proc. Natl Acad. Sci.
Hausfather Z 2021 Flattening the curve of future emissions (The 117 19656–7
Breakthrough Institute) (available at: https:// Sognnaes I et al 2021 A multi-model analysis of long-term
thebreakthrough.org/issues/energy/flattening-the-curve-of- emissions and warming implications of current mitigation
future-emissions) efforts Nat. Clim. Change 11 1055–62
Hausfather Z and Peters G P 2020b RCP8.5 is a problematic Stammer D et al 2021 Hamburg climate futures outlook 2021:
scenario for near-term emissions Proc. Natl Acad. Sci. assessing the plausibility of deep decarbonization by 2050
117 27791–2 (University of Hamburg)

7
Environ. Res. Lett. 17 (2022) 024027 R Pielke Jr et al

Tebaldi C et al 2021 Climate model projections from the scenario Wiek A, Withycombe Keeler L, Schweizer V and Lang D J 2013
model intercomparison project (ScenarioMIP) of CMIP6 Plausibility indications in future scenarios Int. J. Foresight
Earth Syst. Dyn. 12 253–93 Innov. Policy 9 133–47
United Nations Environment Programme (UNEP) 2021 Working Group III of the Intergovernmental Panel on Climate
Emissions gap report 2021 (UNEP) Change (IPCC WGIII) 2014a Climate change 2014:
Walton S, O’Kane P and Ruwhiu D 2019 Developing a theory of mitigation of climate change Working Group III contribution
plausibility in scenario building: Designing plausible to the fifth assessment report of the Intergovernmental Panel
scenarios Futures 111 42–56 on Climate Change (IPCC)