Download Review of existing emissions pathways and evaluation of

Review of existing emissions
pathways and evaluation of
decarbonisation rates
June 2014
Authors: Olivier Dessens1, Gabrial Anandarajah1, and Ajay
Gambhir2.
1
UCL Energy Institute
2
Grantham Institute, Imperial College London
Version: 1.0
Reference: AVOID 2 WPC1
Funded by
This work was supported by AVOID 2 programme (DECC) under contract reference no. 1104872
For more information about the AVOID 2 programme visit www.avoid.uk.net
2
Executive Summary
This report presents an assessment of recent studies examining greenhouse gas (GHG)
emission pathways that are compatible with limiting average global temperature rise to levels
close to 2°C or at least below 2.5°C by 2100.
Whilst the studies examined show some variation due to differing underlying assumptions in
areas such as technology constraints and socioeconomic factors, strong commonalities
could be found between the studies and the following key conclusions drawn.
1. Mitigation to 2°C or below is still possible.
In order to meet stringent climate targets compatible with the 2°C goal it is necessary
to limit atmospheric CO2-eq concentrations to between 450-500 parts per million by
the end of the 21st Century. However, this level of climate mitigation will require a
significant transformation in the way we produce and consume energy and the way
we use the natural environment.
In most studies examined, remaining below the 2°C target required net global
emissions in 2100 to be zero or negative. This therefore required relying on
technologies such as biomass with carbon capture and sequestration in order to
counterbalance the unavoidable GHG emissions resulting from continued economic
activity.
2. Mitigation to 2.5°C is easier to achieve.
Achieving the less ambitious goal of 2.5°C typically requires limiting CO2-eq
concentrations to under 550 ppm by the end of the century. However, whilst this
level of mitigation may be easier to achieve, it will still require strong GHG emissions
reductions typically to less than half the present day.
3. Removing key technologies will make mitigation more difficult and/or more costly.
It was found that in order to remain under the 2°C target early action and a full
portfolio of low-carbon technologies is needed in order to keep global mitigation costs
down. Therefore the availability, cost and future performance of key technologies
has an important role in achieving climate targets.
Widespread electrification of the energy end use sectors combined with strong
decarbonisation of electricity production occur in most mitigation scenarios. Nonfossil energy sources replace coal in the near term and gas in the medium and longterm in the electricity sector. Renewable electricity generation deployment is not by
itself sufficient to achieve the required levels of electricity decarbonisation, and most
of the 2°C scenarios also depend, during the second-half of the century, on the largescale deployment of CO2 capture technologies.
Finally energy demand reduction can significantly reduce mitigation costs as well as
the reliance on carbon capture and storage (CCS); unfortunately the demand side
3
technologies are usually studied and characterised with less detail than the supply
side in the studies reviewed in this paper.
4. Delaying emission reductions will increase the need for negative emission
technologies.
The more delayed the start of robust mitigation policies for a specific climate goal, the
higher the rate of emissions reduction after policy is implemented, and the more likely
there is to be a heavy reliance on negative emissions technologies to keep within the
cumulative carbon budget for the 21st century that is required to meet the climate
policy.
5. Mitigation costs increase when climate targets are more stringent and when action is
delayed.
Costs of mitigation vary largely between models; however, in most of the scenarios,
the more stringent the climate target, the higher the mitigation costs. Consumption
losses for the least cost pathways to reach the 2°C target in 2100, with the
assumption that all mitigation technologies are available (including all major
renewables, nuclear and CCS), are in the range of 2 to 6% in 2050 and 3 to 11% in
2100 relative to the no mitigation (or “business as usual”/baseline) scenarios.
Delayed action, in general, increases the global mitigation costs and also leads to a
possible fossil fuel lock-in of the electricity system, creating large and expensive
unusable assets. As such the most cost-effective scenarios to achieve the 2°C target
are characterised by early mitigation creating clear signals for low-carbon technology
investments and a near term peak in emissions (with the latest possible peaking
dates occurring between 2025 and 2030, and with peak GHG emissions in the range
of 30 to 50 GtCO2-eq).
6. There is scope for further analysis on the implications of mitigation scenarios.
There are a number of areas where there is variation between (or a lack of
transparency around assumptions for) the models used. These include: differing
assumptions on population and economic growth; a lack of detail on technology costs
and deployment constraints; a lack of detailed analysis of technologies and measures
to achieve energy efficiency and reduced end-use demand; a lack of detail on the
implications of lock-in and the consequences of early retirements of carbon-intensive
technologies; and a lack of monetisation of mitigation co-benefits such as energy
security and reduced air pollution.
These model limitations and variations imply that there remains a detailed research
agenda going forward, to help better understand the real-world drivers and
constraints behind the feasibility of deep emissions reduction scenarios.
4
Table of contents
1
Introduction ................................................................................. 6
2
Studies included in the report ...................................................... 7
3
Pathways description .................................................................11
4
Socio-economic assessment ......................................................18
5
Technology parameters ..............................................................19
6
Mitigation policies .......................................................................24
7
Technology and cost implications ...............................................26
8
Co-benefits and secondary impacts of mitigation options ...........34
9
Current knowledge and data gaps ..............................................36
10
Concluding points .......................................................................38
11
References .................................................................................39
5
1. Introduction
In March 1994 the UNFCCC entered into force and recognised that it is necessary to
stabilize atmospheric greenhouse gas (GHG) concentrations at a level that would prevent
dangerous anthropogenic interference with the climate system [1]. Defining the dangerous
level is a value judgement; however a consensus between stakeholders in Copenhagen in
2009 [2] concluded that to comply with this goal the warming achieved should be limited to
below 2°C compared with preindustrial times. The 2°C target is still included in the text
adopted in the 2011 UNFCCC in Durban [3] as well as an assessment of scenarios that
could achieve a possible temperature change below 1.5°C. To limit the long-term
temperature rise to this level, deep cuts in global GHG emissions are needed.
There has been a great deal of analysis to consider whether the mitigation commitments (or
‘Copenhagen pledges’) made to date are consistent with achieving a 50:50 chance of
limiting the surface temperature rise to 2°C (UNEP objective [4]) with the conclusion that the
scenarios including these near term pledges are not least-cost optimal pathways [5]. Many
authors argue that with further ambitious global policies, the target is reachable ([6]; [7]; [8]),
although others suggest it could be too late, as we are already locked into a fossil based
energy system under the weaker-than-optimal “Copenhagen pledges” ([9]; [10]; [11]).
Part of the reason for this dichotomy of views is that the complexity of the climate system, as
well as the extent of uncertainties embedded in it, gives rise to a wide variety of possible
emission trajectories that are consistent with a 2°C temperature rise. The additional
uncertainties and complexities with modelling the global energy system lead to an even
wider range of views on whether, or how, such cuts in emissions are possible.
In this report we present an assessment of studies examining the latest GHG emission
pathways that are compatible with limiting average global temperature rise to levels close to
2°C or at least below 2.5°C by the end of the 21st century. The majority of mitigation
scenarios assessed over recent years have focused on GHG pathways broadly consistent
with achieving atmospheric concentrations of GHGs between 450 ppm and 550 ppm.
Although, as already discussed, there remains uncertainty in the relationship between
atmospheric GHG concentrations and long-term temperature changes, broadly speaking the
450 scenarios are aimed at achieving an even or better chance of limiting surface warming
to 2°C, and the 550 scenarios to 2.5°C.
In the analysis part of this report we first examine the global pathways for greenhouse gas
emissions and the subsequent rate of reduction in annual emissions to comply with targeted
temperature rise below the 2°C and 2.5°C limit with at least 50% achievement chance (more
likely than not). Second, we examine the changes in technologies deployment and costs
between different mitigation options, focussing in particular on the implied rates of
deployment of a number of key electricity decarbonisation technologies. Third, we assess if
there are co-benefits reported from the mitigation and technology choices within the
pathways description. Finally the last chapter outlines the perceived gaps in data and
knowledge that need more scrutiny.
6
2. Studies included in the report
Modelling studies and inter-model comparisons
Energy Modelling Forum 27 Study (EMF27); https://emf.stanford.edu/projects/emf-27-globalmodel-comparison-exercise
The 18 Integrated Assessment Models (IAMs 1) participating in EMF27 are: GCAM, FARM,
MERGE, Phoenix, EC-IAM, TIAM-WORLD, IMACLIM, IMAGE, MESSAGE, POLES,
REMIND, WITCH, AIM-Enduse, BET, DNE21+, GRAPE, GCAM-IIM, and ENV-Linkages.
(See Table 1).
This study incorporates different levels of climate policy, including a 550 and 450 ppm target.
The 550 ppm target does not allow overshooting this level of CO2e concentrations at any
time until 2100 whereas in the 450 ppm case a temporary overshoot is allowed before
reaching the target in 2100 [12]. The study also considers two alternative policy regimes,
one fragmented policy case and a scenario based on the broadly defined goal articulated in
recent years by leaders in the G8 (now G7) [13].
Low climate IMpact scenarios and the Implications of required Tight emission control
Strategies (LIMITS); http://www.feem-project.net/limits/
LIMITS is funded under the EU FP 7 research program and compares seven IAMs
(MESSAGE, IMAGE, TIAM-ECN, REMIND, WITCH, AIM, GCAM – See table 1) over 12
scenarios. The scenario design encompasses atmospheric GHG concentrations reaching a
maximum of 450 or 500 ppm (two different scenarios), the assumed level of ambition in 2020
(a reference policy reflecting the unconditional Copenhagen Pledges and a more stringent
version based on conditional Copenhagen Pledges [14]), and the burden sharing scheme to
be adopted once the international treaty is signed (no sharing, per capita convergence and
equal effort in term of mitigation costs relative to economic output).
Assessment of Climate Change Mitigation Pathways and Evaluation of the Robustness of
Mitigation Cost Estimates (AMPERE); http://ampere-project.eu
Ampere encompasses 12 global IAMs: REMIND, MESSAGE-MACRO, WITCH, MERGEETL, IMACLIM, IMAGE/TIMER, POLES, {GEM-E3, WorldScan, AIM-Enduse, DNE21+,
GCAM scenarios until 2050 only} (See Table 1). Ampere has a strong focus on EU
technology and policy and concentrates its analyses on what action over the next 2 decades
implies for subsequent action to meet a 2°C goal. Following a path based on extrapolation of
the current international emission reduction commitments throughout the rest of the century,
model calculations suggest that global mean temperature would increase by 3.2 – 3.8°C by
2100[15] and [16].
Global Energy Assessment (GEA): Toward a Sustainable Future [17]
GEA uses two integrated assessment models (MESSAGE and IMAGE) in order to analyse
60 alternative transitions pathways towards a low-carbon future. The final GEA report
examines the major global challenges and their linkages to energy, the technologies and
1
In assessment of climate change, IAM refers to a mathematical tool that considers the social and
economic factors that drive the emission of greenhouse gases, the biogeochemical cycles and
atmospheric chemistry that determines the fate of those emissions, and the resultant effect of
greenhouse gas emissions on climate, ecosystems and human welfare.
7
resources available for providing adequate, modern and affordable forms of energy, the
plausible structure of future energy systems most suited to addressing the century’s
challenges, and the policies and measures needed.
The Roadmaps towards Sustainable Energy futures (RoSE); http://www.rose-project.org
The Roadmaps towards Sustainable Energy futures (RoSE) project aims to provide a robust
picture on energy sector transformation scenarios for reaching ambitious climate targets
involving 5 leading integrated assessment modelling teams (using REMIND, GCAM, WITCH,
IPAC, and China MARKAL). RoSE assesses the feasibility and costs of climate mitigation
goals across different models, different policy regimes, and different reference assumptions
[18]. The scenarios and policies within RoSE that are currently available and published are
similar to the AMPERE ones; however in the future RoSE will perform more scenarios with
the same climate policies (450 and 550 ppm) under different socio-economic global
development (different GDP, population growth, and fossil fuels prices for example). One
RoSE publication analyses the near-term role of fossil fuel usage under “late action”
regarding the potential carbon-intensive electricity generation capacity that will have to be
left idle when decarbonisation occurs [19].
TIAM-UCL global modelling studies: The CCC report [20] and UKERC Global study [21]
In these studies the TIAM-UCL energy system IAM, using a simple climate module,
simulates GHG pathways consistent with 2°C, 2.5°C and 3°C targets (with and without
overshoot). Technology availability such as carbon capture and sequestration (CCS) or bioenergy with carbon capture and storage (BECCS) as well as different CO2 emission rate of
reduction limits have been applied to the scenarios. The targets feasibility is discussed and
technology development and system costs are reported.
Assessment reports
Climate Change 2014, Mitigation of Climate Change. Contribution of Working Group III to the
Fifth Assessment Report of the Intergovernmental Panel on Climate Change [22]
The IPCC assessment report builds upon more than 800 GHG emissions scenarios
extracted from pre-published material in peer reviewed literature, with a large number of the
most recent scenarios made up of the EMF 27, Ampere, RoSE and LIMITS studies already
described. The scenarios are collated in the AR5 scenario database (IAMC AR5 Scenario
Database, [23]). The scenarios are grouped into baseline scenarios and GHG mitigation
scenarios, which in turn are further grouped by levels of ambition to reduce GHG emissions:
the most stringent mitigation scenarios (category 1) correspond to long-term total radiative
forcing targets of 2.3 to 2.9 Wm-2 (425 to 475 ppm CO2e), which are compatible with limiting
global average temperature change to below 2°C. Category 2 scenarios correspond to
stabilization of total radiative forcing between 2.9-3.4 Wm-2, category 3: 3.4-3.9 Wm-2,
category 4: 3.9-5.1 Wm-2, and category 5: 5.1-6.8 Wm-2. Per category, the indicative
temperature change ranges are between 1.7-2.2°C, 2.0-2.5°C, 2.3-3.0°C, and 3.0-4.0°C
respectively. Scenarios in the highest category (Category 6) correspond to modest mitigation
efforts leading to radiative forcing levels greater than 6.8 Wm-2 with temperature outcomes in
2100 above 4°C. The AR5 report also presents the implications, in terms of GHG emissions
and temperature change in 2100, of the Representative Concentration Pathways (RCPs)
8
which were developed by the Integrated Assessment Modelling (IAM) community to provide
a standardised set of GHG pathways. These RCPs are also summarised in this paper.
The UNEP Emissions Gap Report 2012 [4]
For this assessment, the analyses of 12 research and/or modelling groups have been
reviewed (Peterson Institute for International Economics; Project Catalyst; AVOID research
programme, led by the Met Office Hadley Centre; UNEP Risoe; and the World Resources
Institute; Climate Action Tracker, Climate Analytics for the research projects and C-ROADS;
WITCH; GAINS; ENV-linkages; FAIR for the models). A total of 223 emission pathways
produced by 15 modelling groups have been analysed. Pathways for different temperature
targets (<2, 2-2.5, 2.5-3, 3-3.5, 3.5-4, 4-5, >5) with a likely (>66%) chance being below the
targets of temperature increases have been produced. The pathways include several near
term targets too.
The subsequent UNEP Emissions Gap Report 2013 [5] includes reviews of the results from
projects that are listed in this report (LIMITS, EMF and AMPERE) as well as some single
model studies.
9
Table 1: List of models involved in the projects included in the report.
Model
name
Model
category
Solution
Algorithm
AIM /
AIMEnduse
Partial
equilibrium
Recursive
dynamic
BET
General
equilibrium
Intertemporal
optimization
China
Partial
MARKAL equilibrium
Dynamic
linear
optimisation
DNE21+
Partial
equilibrium
Intertemporal
optimization
GCAM /
Partial
equilibrium
Recursive
dynamic
General
equilibrium
Intertemporal
optimization
GCAMIIM
EC-IAM
Reported
Land use sector
representation
Coverage of
greenhouse gases
2050
Marginal Abatement
Costs (MACs) for land
use emissions
All GHGs and other
radiative agents
2100
None (land use
emissions exogenous)
CO2
2050
MAC curves for souces
CO2
2050
MACs for land use
emissions
All GHGs and other
radiative agents
2100
Endogenous land use
dynamics, Afforestation
option
All GHGs and other
radiative agents
None
Kyoto gases from
fossil fuel
combustion and
industry
2050
MACs for land use
emissions
Kyoto gases
2100
Land is competed across
crops, pasture, forests,
and biomass
CO2 from fossil fuel
combustion and
industry
time
horizon
2100
ENVgeneral
Linkages equilibrium
Recursive
dynamic
FARM
general
equilibrium
Recursive
dynamic
GCAM
Partial
Equilibrium
Recursive
dynamic
2100
Endogenous land use
dynamics
full basket of
greenhouse gases,
precursors and
aerosols
GEM-E3
general
equilibrium
Recursive
dynamic
2100
best allocation between
food/ energy crops
production
All GHGs
GRAPE
General
equilibrium
Intertemporal
optimization
Endogenous land use
dynamics
All GHGs and other
radiative agents
IMACLI
M
general
equilibrium
Recursive
dynamic
None
CO2 from fossil fuel
combustion and
industry
IMAGE /
TIMER
Partial
equilibrium
Recursive
dynamic
2100
Endogenous land use
dynamics
All GHGs and other
radiative agents
IPAC
Multi-model
framework
links several
models
2100
Land equilibrium model
All GHGs
10
2100
2100
MERGE
General
equilibrium
Intertemporal
optimization
MACs for land use
emissions,
All GHGs and other
radiative agents
2100
No CO2 emissions from
land use
MESSA
GEMACRO
General
equilibrium
Phoenix
general
equilibrium
Recursive
dynamic
Partial
equilibrium
Recursive
dynamic
General
equilibrium
Intertemporal
optimization
POLES
REMIND
Intertemporal
optimization
2100
MACs for land use
emissions
All GHGs and other
radiative agents
Afforestation option
TIAMECN
Partial
Equilibrium
Intertemporal
optimization
TIAMWorld
Partial
equilibrium
Intertemporal
optimization
WITCH
General
equilibrium
Intertemporal
optimization
WorldSc
an
general
equilibrium
Recursive
dynamic
None
CO2 from fossil fuel
combustion and
industry
None
Kyoto gases from
fossil fuel
combustion and
industry
All GHGs and other
radiative agents
2100
Baseline land use
emissions from land use
model MAgPIE coupled
to MACs, endogenous
land use dynamics from
MAgPIE in some
scenarios
2100
None - exogenous
CO2, CH4, N2O.
2100
MACs for land use
emissions
Kyoto gases with
the exception of FGases
MACs for land use
emissions
Kyoto gases
2050
2100
2100
2100
3. Pathways description
Within Table 2 (physical climate parameters) we have listed the scenarios that are broadly
consistent with a 2°C (in green) and a 2.5°C (in amber) target. It should be noted that all
these scenarios represent ambitious goals with dramatic changes in anthropogenic GHG
emissions.
Limitations to comparing scenarios
Comparing the findings of different scenarios can be difficult, although not impossible, in part
due to the variety of ways the targets within different studies are set. Targets used by
studies include:
11
1. the maximum temperature in 2100 [4];
2. the Representative Concentration Pathway (RCP) describing the radiative forcing
[24] and [25];
3. the maximum concentration of GHGs (LIMITS, AMPERE, RoSE, EMF27);
4. the emissions pathways (EMF27 with Fragmented Policy and G8 commitment policy
[7]).
Another possible limitation in comparing the scenarios of different studies, noticeable in
Table 3 (Socio-economic parameters), is the diverse socio-economic storylines that supports
scenario developments for modelling. Within the LIMITS project alone, 2050 GDP varies
between USD 103 to USD203 trillion depending on the model used. In the same way
population in 2100 is assumed to range from 8.7 to 10 billion with possible stabilisation at the
end of the century. The second group of scenarios shows consistency toward a temperature
change in 2100 below the 2.5°C target and are all converging toward the 550ppm CO2
concentration in 2100. This group also allows a delayed peak year - until 2035 - with higher
peaking emission quantities in certain cases.
Pathways broadly consistent with meeting 2°C or below
Based on 2°C target studies, it seems that 500 ppm is the maximum permissible CO2-eq
concentration in 2100, with emissions peaking in 2030 at the latest; the later peaking dates
proving less cost-effective and relying heavily on CO2 removal technologies such as CCS.
Most of these scenarios will also have to exhibit negative CO2 emissions at the end of the
century.
A few of these scenarios indicate an expected temperature change below 2°C in 2100
(between 1.5 and 1.8°C): RCP2.6, EMF27-450 and UNEP_Gap_Report_1.5°C. Generally
the mean annual GHG emissions reduction rate, following the peak, is between 2 to 5%
when the peak year is around 2020. Later peaking pathways will lead to larger rates of GHG
emission reduction (from 6 to 8% per year) and require negative emissions at the end of the
period to comply with the target with a temporary overshoot. Although the rapid reduction in
global emissions in some scenarios is technically and economically feasible within the
modelling framework, political decisions, social acceptance and institutional factors will also
play a major role in the real world – elements which are not part of the modelling framework,
other than through the mechanism of delayed or regionally fragmented action. Focussing at
national level, some examples of very rapid emissions reductions can be found in the recent
past: during the 1980s France was reducing emissions at a rate of 4 to 5% per year as a
result of the large-scale deployment of new nuclear power plant facilities; the UK sustained a
reduction reaching 2% per year in the 1970s decade by a strong switch from coal to gas in
electricity production. These examples highlight the practical rates achievable through
technical changes; however the recorded reductions lasted only a decade or less, and were
at country levels as opposed to the global level required in the scenarios discussed in this
paper. In the literature maximum possible global reduction rate can be extracted; for example
maximum annual rates of 3.5% [2] and 4.3% [26] take into account assumptions about
technological development, economic costs, and/or socio-political factors. In some scenarios
analysed for this study the emission reduction rates can reach 8 to 10% per year are largely
exceeding these possible maximum values.
12
In all of these scenarios the only geoengineering technology explored to mitigate climate
change is bio-energy coupled with CCS (BECCS) that could produce negative emissions
during the end of the century. Particularly in scenarios with later peaking years, BECCS is
therefore a critical technology, and its absence often results in an inability for models to meet
the prescribed target where this is relatively stringent (i.e. consistent with 2 °C or less).
Pathways broadly consistent with remaining under 2.5°C by
2100
Pathways in agreement with relaxed target achieving between 2°C and 2.5°C show different
behaviours in term of physical climate factors. Those pathways are coloured in amber in the
Table 2. The concentrations of GHG in 2100 can reach the range 500 to 550 ppm. Mostly in
this case the peak of emissions is allowed to be delayed till 2035 with, after this peak,
emission reduction rates between 3 to 4%/y closer to the accepted maximum range
discussed above. Finally and crucially in these pathways overshoot in radiative forcing is
occurring rarely and the all scenarios still show a positive GHG emission in 2100 in the range
of 1 to 20 GtCO2.
13
Table 2: Physical climatic parameters
Study /
model /
scenario
GHG
concentration
/ ppm of
CO2eq in 2100
Temperature
GHG pathway peak
change by 2100 year and CO2 level
EMF27 450
450ppm
1.5-1.8°C
(target on
RF=2.6 Wm-2)
2025=20to35
GtCO2eq
LIMITS FP7 450ppm
450ppm
1.7±0.1°C
EMF27 G8
480 to 500ppm
LIMITS FP7 500ppm
500ppm
AMPERE
450
immediate
action
14
450ppm
CO2 reached by
2050, by 2100
Forcing
Wm-2 by
2100
2050=8to15 / 2100=20to0 GtCO2
2.5 to 2.9 Cumulative 2010(overshoot 2100=1750–2300
GtCO2-eq
3 to 3.5
around
2050)
2020-30=2.8,203040=5.3,204050=5.2%/y mean
ensemble
2020=53±1 GtCO2eq 2050=25±2 /
2100=0±1 GtCO2
2.8
2011-2050=
(overshoot 860to1270, 20113.0)
2100= 1720to2090
GtCO2-eq
2020-30=2.8,203040=5.3,204050=5.2%/y mean
ensemble
1.8-2.3°C
2020to2030=25to35
GtCO2eq
2050=10to20 /
2100=0to20 GtCO2
3 to 3.8
NA
1.9±0.15°C
2020=53±1
GtCO2eqGtCO2
2050=35±2 /
2100=1±1 GtCO2
3.2
2011-2050=
1500to1940, 20112100= 2100to2720
GtCO2eq
2020-30=2.7,203040=3.8,204050=4.2%/y mean
ensemble
1.9°C (1.7–2.5)
probablity
>2°C=36%
2020=45±5 GtCO2eq 2050=25±2 /
2100=2±2 GtCO2eq
2.8 (max
3.3)
2011-2050= 1689
GtCO2-eq (1084–
2261), 2011-2100=
2289 (1273–2708)
2030-50=4%/y mean
ensemble (3 to 4.5)
O
C above preindustrial
Cumulative CO2eq
budget 2011-2050,
2011-2100
Rate of emissions
reduction from peak
year to 2050
RCP 2.6
455ppm (2050) 1.9°C (0.9-2.3°C
427ppm (2100) min to max
range CMIP5)
2020, 37.6 GtCO2
12.4 (2050), -1.5
(2100) GtCO2
2.7
2011-2050= 5501,270, 2011-2100=
630-1,180 GtCO2
N/a
RoSE
Immediate
action
450ppm
2°C with 50%
chance
2010, at 45
GtCO2eq, although
one model (GCAM)
peaks in 2035, at
50Gt CO2eq
25-30 GtCO2eq by
2050 /
2.6 (peak
3.4)
Not specified
Not specified
RoSE
delayed
action to
2020
450ppm
2°C with 50%
chance
2020 peak at 50-57
GtCO2eq
20-25GtCO2eq in
2050 /
2.6 (peak
3.5)
Not specified
Not specified
RoSE
delayed
action to
2030
450ppm
2.6 (peak
3.8)
Not specified
10%/y between 30302040 (maximum rate)
3.4 (max
3.8)
2011-2050= 2155
GtCO2-eq (1500–
2289), 2011-2100=
2756 (1732–3071)
2030-50=7.5%/y
mean ensemble (6.5
to 8.5)
NA
2030-2050=2to4.5%/y
(late
action=6to8.5%/y)
5 GtCO2eq in 2100
15
2030 peak at 5565GtCO2eq
17-22GtCO2eq in
2050 /
0-5 GtCO2eq in 2100
AMPERE
450ppm
450 delayed
action
UNEP
Emissions
Gap Report
2013
2°C with 50%
chance
highest
450ppm for
CO2
2.1°C (2.0–2.5)
probability
>2°C=60%
2020to2030=45±5
GtCO2eq
2050=20±3 /
2100=0±2 GtCO2eq
Target 2°C with
likely chance
>66%
2010-2020 36to47
GtCO2eq (2020
median=40,42,44)
2050=20,21,22GtCO NA
2eq / 2100:1/3models
negative emissions
UNEP
Emissions
Gap Report
2013
highest
450ppm for
CO2
UNEP
Emissions
Gap Report
2013
highest
Target 1.5°C
450ppm for CO with medium
chance
(50to66%)
Target 2° C with
medium chance
(50to66%)
2010-2020 44to49
GtCO2eq (2020
median=46)
2050=25to32GtCO2e NA
q (28med) /
2100:1/3models
negative emissions
NA
NA
2010-2020 37to44
GtCO2eq
2050=5to18GtCO2eq NA
(28med) / 2100:few
scenarios negative
emissions
NA
2030-2050=2to4.5%/y
(late
action=6to8.5%/y)
30-70% reduction in
2050 compared to
2005. Negative in
2100.
NA
NA
NA
Global
NA
Energy
Assessment
2°C with at least
50% chance
2020 Peak year.
CCC2013UCL
440ppm for
CO2, in line
with 2°C in
2100
2° C
2016 peak 36 GtCO2. 13.9 GtCO2 in 2050 /
35.0 GtCO2 in 2020
5.0 GtCO2 in 2100
(Early action)
2.99 (peak 2010-2050= 1462 /
3.17 in
2010-2100= 2400
2050)
GtCO2-eq
4.00%
CCC2013UCL
450ppm for
CO2, in line
with 2°C in
2100
2.1° C
2025 peak 41GtCO2.
39.9 GtCO2 in 2020
(Later action)
13.9GtCO2 in 2050 /
5.0 GtCO2 in 2100
3.2 (peak
3.47 in
2050)
4.50%
CCC2013UCL
465ppm for
CO2, in line
with 2°C in
2100
2.15° C
2025 peak 41GtCO2.
39.9 GtCO2 in 2020
(Reduced action)
20.9GtCO2 in 2050 /
5.6 GtCO2 in 2100
3.44 (peak 2010-2050= 1775 /
3.61 in
2010-2100= 2880
2070)
GtCO2-eqGtCO2-e
16
2010-2050= 1710 /
2010-2100= 2650
GtCO2-eqGtCO2-e
2.00%
RCP 4.5
500ppm (2050) 1.7-3.2°C (min to 2035= 42 GtCO2
540ppm (2100) max range
CMIP5)
40 GtCO2 (2050) 12
GtCO2 (2100)
4.5
2011-2050= 1,2601,640, 2011-2100=
1,870-2,430 GtCO2
NA
EMF27 550
550ppm
2-2.3°C (target
RF=3.7 Wm-2)
2025to2030=25to40
GtCO2eq
2050=10to20 /
2100=0to20 GtCO2
3.4 to 3.8
Cumulative 20102100=2460–
3260GtCO2-eq
2020-30=2.7,203040=3.8,204050=4.2%/y mean
ensemble
AMPERE
550
immediate
action
550ppm
2.3°C (2.1 to
2.8) probability
>2°C=73%
2030=60 GtCO2eq
2050=20±2 /
2100=1±2 GtCO2eq
3.7 (max
4.0)
2011-2050= 2019
GtCO2-eq (1460–
2420), 2011-2100=
3251(2330–3643)
NA
AMPERE
550ppm
550 delayed
action
2.4°C (2.2–2.6)
probability
>2°C=76%
2030=60 GtCO2eq
only EU road map
2050=20±2 /
2100=1±2 GtCO2eq
2011-2050= 2198
GtCO2-eq (1695–
2384), 2011-2100=
3287 (2584–3619)
NA
RoSE 550
2.4 °C (2.2–2.6)
2020 (50 GtCO2eq) – 30-40 GtCO2eq
2030 (50-55
(2050)
GtCO2eq)
Not specified
2030-2050 = 3%/yr
(maximum rate
across the models)
17
550ppm
3.9-4.2
4. Socio-economic assessment
Globally, economic and population growth continue to be some of the most important drivers
in increasing CO2 emissions. Whilst between 2000 and 2010 the contribution of population
growth to GHG emissions has remained roughly identical to the previous three decades,
during this same period of time the contribution of economic growth has risen sharply [22].
Further, without additional efforts to reduce GHG emissions beyond those in place today,
emissions growth is expected to persist driven by growth in global population and economic
activities.
The studies examined assumed a range of increases in global population sizes and
economic growth, with a high variation noted in the range of economic growth estimates
between studies. This is, in part, due to there being more uncertainties in estimating
economic growth than population increase.
Estimated population growth
In most studies, 2050 population is estimated at around 9 billion while 2100 population
assumptions vary from 9.1 billion (LIMITS and RCP 2.6) to 10 billion (AMPERE). Within the
EMF 27 project variation in population growth exists between models as no socioeconomic
harmonisation was required. One of the RCP scenarios (RCP 4.5) assumes global
population peaking in 2065 (9 billion) and then sees a fall to 8.7 billion in 2100 in contrast
with estimates in other scenarios, in which population continues to rise until the end of the
century. Most of these studies did not explicitly discuss future urbanisation rates, which is
one of the key drivers that contribute to increasing per capita emissions, especially in the
emerging economies in the near and medium term and in developing countries in the longterm. As of 2011, more than 52% of the global population lives in urban areas whilst in 2006,
urban areas accounted for 67–76% of energy use and 71–76% of energy‐related CO2
emissions; by 2050, the urban population is expected to increase to 5.6–7.1 billion, or 64–
69% of world population [22].
Estimated economic growth
Global studies such as UNEP Emissions Gap Report (UNEP, 2013) and Global Energy
Assessment 2012 assume per capita GDP growth of 2% per year to 2050, mostly driven by
developing countries, while the UCL-CCC 2013 [20] and AMPERE studies assume slightly
higher growth rates of 2.4% and 2.7% respectively. EMF 27 assumes an average growth rate
of 1% per year to 2100.
18
Table 3: Socio-economic parameters
Study / model /
scenario
Economic growth rate to 2050, to Population growth
2100
AMPERE
global average growth=2.7% per
year
2100=10billions / medium
fertility variant of the 2010
version of the UN World
Population Prospects
CCC2013-UCL
2.4% to 2050
9.3 billion 2050
EMF27
annual average 1% in 2100
9 billions 2050 to 9.6 2100
(stabilisation)
Global Energy
Assessment
2%/year to 2050
about 9 billion in 2050 (Urban
population 6.4 billion in 2050)
LIMITS FP7
GDP 2010=54 (mean)
2050=103to203 (model dependant)
2100=730(TIAM-ECN only)
9.2, 9.1 billion (TIAM_ECN
only)
RCP 2.6
2000=$34, 2030=$84, 2100=$229
trillion (1995 $US)
8.2 billion (2030),9.1 billion
(2100)
RCP 4.5
2000=$40, 2030=$75, 2095=$330
trillion (2005 $US) (2005)
8 billion (2030), 9 billion
(2065=peak),8.7 billion (2100)
UNEP Emissions
Gap Report 2013
per capita GDP growth rate is 2% to NA
2050
5. Technology parameters
This section highlights the technologies included in the different scenarios. The usual
approach in the majority of the projects included in this report is to use a business-as-usual
or reference scenario to compare to a series of mitigation scenarios (of different levels of
stringency) involving a large portfolio of technologies available at specified costs. These “full
technology portfolio” scenarios usually allow strong and rapid developments in renewable or
other low-carbon technologies in the power sector as well as the deployment of new
technologies such as CCS. Energy efficiency improvement options in final energy demand
sectors are also included, but treated as separate from the group of energy generating
technologies. In these “full portfolio” modelling exercises technology costchanges over time
can occur through two channels: learning-by-doing (experience gained during development)
or learning-by-searching (research and development activities). These improvements can
induce new dynamics between technology adoptions and create divergences among model
results (LIMITS and EMF27). However in most model scenarios, technology costs are
specified as a set of input assumptions.
Technological challenges are studied in a series of further scenarios including limitations on
the availability of specific technologies or groups of technologies. The usual technology
19
restrictions (which are assumed to follow from technical limits or political decisions to restrict
technology deployment) in these alternative scenarios are: a nuclear phase out, no CCS
development, reduced deployment of wind and solar because of intermittency limits, and
finally reduced availability of biomass as an energy feedstock. Some scenarios are
calculated with a combination of these restrictions.
In the baseline scenarios the energy demand increase is met primarily with carbon intensive
fossil fuels; generally the CO2 emissions for 2050 reach 2 to 3 times the 2010 levels. Within
the full technology portfolio scenarios decarbonising the electricity generation sector is one of
the main approaches to achieving the targets. The share of low-carbon generation in the
electricity supply increases from 30% today to 80-100% decarbonisation in 2050 (depending
on the stringency of the climate target). Within the total primary energy sources, low-carbon
sources represent only between 60% and 70% in these scenarios highlighting the more
difficult decarbonisation of other sectors (such as transport) compared to electricity
generation. Renewable technologies such as wind and solar commonly take the largest
share of low-carbon electricity generation in 2050. The share of CCS in electricity generation
in 2050 varies from 8% to 32%. This is partly driven by assumption on CCS deployment,
which starts in or after 2025 in all scenarios, and assumptions deployment renewable
technologies. However this fact changes during the second half of the century when CCS
technology becomes more mature and renewable generation reaches saturation; this is
particularly the case in scenarios with stringent targets or overshoots when BECCS in the
second part of the century is needed to create negative emissions.
20
Table 4: Technology parameters
Study / model /
scenario
Key technology transitions
Technologies Deployment constraints
excluded
AMPERE 450ppm
immediate action
Decarbonisation power generation, substitution of
fossil fuel with electricity in stationary, transport
electrification, acceleration of energy efficiency
improvements: renewable+storage+gas-fired
capacity.
Nuclear, CCS
AMPERE 450ppm
delayed action
Nuclear and CCS critical only if delayed action 2030 None
CCC 2013/ TIAM-UCL Decarbonisation occurs foremost in the power
sector and, with the CO2 intensity decreasing from
over 500 g/kWh in 2010 to 54 g/kWh in 2030 and to
1.5 g/kWh in 2050.
EMF 27
Near-term: decarbonisation of electricity
Long-term: increase in use of electricity in end-use and
appropriate mix of technologies for decarbonizing
electricity depends on assumptions about supply
technology characteristics and availability.
21
Sensitivity
analysis has
been carried
out without
CCS
technology
Constraints added to solar, wind and
biomass: E.g. biomass 100 EJ/year,
maximum 20% electricity from solar and wind
together.
As above
Build rate constraints: 15%/year for CCS
(coal, gas and biomass); 15%/year for PV
and 10%/year for wind. New nuclear capacity
addition per year is constrained to 30 GW in
2030 and 60 GW in 2050. Biomass potential
(energy crops and solid biomass) is limited to
10,600 TWh/year.
No CCS and
Solar & wind 20% max share. Biomass is
nuclear phase constrained to 100EJ/y
out
GEA 2012 (analysed
60 alternative
transitions) /
MESSAGE and
IMAGE
Cumulative CO2 capture up to 250 Gt in 2050.
None
Not reported
None
Not reported
None
Not reported
Complete phase-out of coal power without CCS by
2050.
Strong bioenergy growth from 45 EJ in 2005 to 80–
140 EJ by 2050 (co-firing with coal or gas with
CCS).
Low carbon electricity ranges from 75%-100% in
2050
LIMITS FP7 450 ppm
Electricity generation annual new capacity:
2010-30: wind 60, Solar 30, Nuclear 30, and CCSFF 15
GW (6 models average)
2030-50: wind 145 / solar 175 / Nuc 45 / CCSFF 80 /
BECCS 25 GW (6 models average)
RCP 2.6
11 GtCO2 stored annually
350 EJ of bioenergy
[Van Vuuren 2007]
RCP 4.5
Electricity sector becomes carbon negative by 2100
(40% nuclear, 20% coal w/CCS, 10% gas w/CCS,
5% hydro, 5% biomass w/CCS, 20% other)
None
Not reported
RCP 6.0
By 2100 CCS deployed on 74% of thermal fossil
power plants, share of non-fossil power plants
exceeds 30%
None
Not reported
22
RCP 8.5
Fossil fuels dominate (coal use multiplies tenfold
throughout century) and renewables learning rates
are low (10% price reduction for each doubling of
capacity).
None
Not reported
RoSE
All three models (WITCH, REMIND, GCAM) show
significant reduction in final energy demand by
2100, between 20% (REMIND) and 55% (WITCH)
lower than in the baseline.
None
In GCAM, biomass availability is limited to
210 EJ/year. In REMIND, existing capital can
be prematurely retired, but only up to
4%/year of installed capacity.
None
Not reported
UNEP-Emissions Gap Net negative emissions in 2100.
Report 2013
Bio-energy with CCS in the second half of the
centaury
23
6. Mitigation policies
There are three ways for pricing carbon to reduce CO2 emissions [28].
1. Carbon taxation (tax);
2. Carbon trading on the basis of trade in rights to emit which are allocated or auctioned
(cap-and-trade);
3. Implicit pricing via regulations and standards, which impose constraints on
technologies that can be necessary where irremovable or unavoidable market
imperfections exist.
The interactions between these mitigation policies may be synergistic, or may have no
additive effect on reducing emissions [22]. For instance, a carbon tax can have an additive
environmental effect to policies such as subsidies for the supply of renewable energy. By
contrast, if a cap and trade system has a binding cap (sufficiently stringent to affect
emission‐related decisions), then other policies such as renewable energy subsidies have no
further impact on reducing emissions within the time period that the cap applies. In either
case, additional policies may be needed to address market failures relating to innovation and
technology diffusion.
Modelling studies either use (in the case of bottom-up least-cost optimisation models) a cap
and generated “shadow” prices for carbon (i.e. carbon prices which if imposed would
generate the same least-cost solution as that reported) or (in the case of top-down macroeconomic models) apply a carbon price in order to analyse impacts on emissions, as a result
of changing the relative prices of carbon-intensive energy and alternative energy. Additional
policies or constraints implemented in the modelling exercises, which are important in
determining the feasibility of meeting global climate change mitigation targets, are near-term
mitigation pathways (including caps or targets for specific near-term dates, often to simulate
levels of effort commensurate with the Cancun pledges), the choice of emissions peaking
year, and the method by which effort is shared between regions (e.g. through equal per
capita emissions, equal marginal abatement cost, or some other burden-sharing principle),
as well as whether regions can freely trade carbon permits with each-other.
Global studies almost invariably suggest that early action is preferred (Table 5: Policy
parameters), as it will reduce overall mitigation cost and mitigation uncertainty. It will also
avoid the risk of locked in investment in high carbon technologies, which will increase overall
mitigation costs through early capital scrapping. Though most of the studies suggest early
action is preferred, the peaking year and peaking emissions level vary across the studies.
For example, the Global Energy Assessment (IIASA, 2012) reports that limiting global
temperature increase to less than 2°C over pre-industrial levels (with a probability of >50%)
is achieved in the pathways through rapid reductions of global CO2 emissions from the
energy sector, peaking around 2020 and declining thereafter to 30–70% below 2000
emissions levels by 2050, ultimately reaching almost zero or even net negative CO2
emissions in the second half of the century. Higher CO2 prices in 2050 are generated the
longer the delay in implementing global emissions reductions (i.e. the later the date at which
global emissions peak) and the greater the required level of emissions reductions [20].
24
Table 5: Policy parameters in mitigation scenarios
Study / model /
scenario
Effort sharing / carbon trading
assumed?
Delayed action?
AMPERE
extrapolation current
policies
Fragmented policy
Peak: 2060-270
AMPERE 450ppm
immediate action
Strong global action
No
CCC 2013-UCL: Core Equal per capita emissions of
1.5tCO2/person in 2050
Peaking in 2016
CCC 2013-UCL: Later Equal per capita emissions of
action
1.5tCO2/person in 2050
Peaking in 2025
CCC 2013-UCL:
Reduced action
Equal per capita emissions of
2tCO2/person in 2050
Peaking in 2025
EMF 27
None
Harmonised policy=least cost
mitigation options, globally uniform
carbon price / Fragmented policy:
Group I (no Russia) -50% emission
in 2050; Group II slower reduction +
limited emission trading with Group
I; Group III: no policy (fossil fuel
producers + Russia)
LIMITS FP7 450 ppm
Burden sharing after 2020 (2
policies: per capita convergence
and equal effort)
Peak 2020
RCP 2.6
N/a
Emissions reduction begins in
2020
RCP 4.5
Yes - assumes a common global
price is shared across regions.
No – carbon price rises at a fixed
rate annually until the forcing level
is reached, then remains roughly
constant.
RCP 6.0
Yes – assumes a global emissions
permit market
Yes – no carbon price until 2050.
However relative to baseline
carbon intensity per unit energy
dos decrease in period to 2050.
RCP 8.5
N/a
N/a
RoSE
Carbon trading established
following any delays in action
RoSE explores immediate action,
delayed action until 2020, and
delayed action until 2030.
25
7. Technology and cost implications
The pathways indicate that the energy transformations need to be initiated without delay,
gain momentum rapidly, and be sustained for decades. This requires the rapid introduction of
policies and fundamental governance changes toward integrating climate change into local
and national policy priorities. A range of measures is required in each sector. For example,
as discussed in the GEA [17], rather than aiming for buildings that use zero fossil fuel energy
as quickly as possible, an economically sustainable energy strategy would implement a
combination of the following: reduced demand for energy; use of available waste heat from
industrial, commercial, or decentralized electricity production; on-site generation of heat and
electricity; and off-site supply of electricity. Some assessments, notably the IPCC Special
Report on Renewable Energy Sources and Climate Change Mitigation [27] and the GEA [17]
emphasize the great importance of accelerating demand-side efficiency and conservation
measures for future reductions of greenhouse gas emissions.
Fossil fuel consumption in higher-income regions consequently falls (as shown in Table 6:
technology and cost implications), and there is resultant downwards pressure on fossil fuel
prices. Cheaper resources are therefore available to the middle and low-income countries
and so there is almost no additional cost to these regions in 2020 (while marginal, the
change in cost is still positive, however.) The assumption of perfect foresight means that
some of the middle and low-income regions do implement some emissions reduction (albeit
at a much lower level of ambition than in high-income regions) [21].
Long infrastructure lifetimes mean that it takes decades to change energy systems; so
immediate action is needed to avoid lock-in of invested capital into existing energy systems
and associated infrastructure that is not compatible with longer climate targets [17].
Infrastructure developments and long‐lived capital stocks that lock societies into
GHG‐intensive emissions pathways may be difficult or very costly to change, reinforcing the
importance of early action for ambitious mitigation. This lock‐in risk is compounded by the
lifetime of the infrastructure, by the difference in emissions associated with alternatives, and
the magnitude of the investment cost. As a result, lock‐in related to infrastructure and spatial
planning is the most difficult to reduce [22].
The largest threat of lock-in of technologies within the energy system regards electricity
production from fossil fuel (for coal and gas) and has been considered in the case of a 2°C
target following scenarios with different short term emission reduction goals within the
AMPERE and the RoSE projects. Currently, about 90% of global primary energy supply
comes from coal, oil and gas. Climate policy and pricing of CO2 emissions are likely to make
some of the fossil installations unprofitable, thus resulting in pre-mature retirement of fossil
capacities before the end of their technical lifetimes. In the AMPERE project, a range of GHG
emissions targets are specified (from 50 to 60GtCO2eq), with the long term target in all cases
fixed below 2°C. Scenarios with higher short term targets have to rely heavily on negative
emissions at the end of the century to achieve the long term goal. More importantly, with
these less stringent short term targets the phase-out of coal (and gas) capacity in electricity
production is delayed until after 2030 and as a consequence fossil fuel generation capacity
continues to be built in the period to 2030; in this case the phase-out of coal and gas
production, to totally decarbonise the electricity system in 2050, creates stranded capacity in
the electricity generation sector. In the worst case (the highest 2030 target) the stranded
investment reaches globally $60billion for the 2010-2030 period and almost $450billion for
26
the 2030-2050 period with a particularly large contribution from China and South Asia, which
will have invested heavily in coal generation during the first period. The 2030-2050 costs for
these two regions represent more than 10% of their total investment in electricity generation
during the period. To avoid these large stranded capacities fixing lower short term targets are
effective (targets below 53GtCO2eq in 2030 reduce the above costs by two thirds). Other less
effective options available to avoid high costs from stranded capacity are reducing energy
demand (increasing efficiency), retrofitting old coal and gas capacity with CCS (if available)
and increasing the lifetime of existing coal capacity (instead of building new ones).
The RoSE results for similar scenarios show that in 2030, between 600 and 1400 GW of
fossil power generation capacity are idle in the best policy case with immediate action. Early
retirements peak later and at a higher level (up to 3,500 GW) in the delayed scenarios.
There is some variation in total energy system costs at a regional level. In 2020, overall costs
are higher in high-income regions since these are required to meet their Copenhagen Accord
emissions reductions. This switches after 2020, with a greater proportional cost in middle and
low-income regions [21]. For middle-income regions, the rapid increases in energy-service
are predominantly met through an increase in coal consumption. Given that coal
consumption needs to be severely restricted in 2°C scenarios until CCS is fully available,
these regions require a greater level of investment to meet the emissions reductions
required. In low-income regions it is assumed that the unit capital costs of many of the lowcarbon technologies (such as CCS) are higher than in the high-income regions. These are
consequently more expensive to deploy and overall costs are higher.
According to IIASA [17], in order to achieve the 2°C target, at least a 60–80% share of global
primary energy will need to come from zero-carbon options by 2050; the electricity sector in
particular will need to be almost completely decarbonized by mid-century (low carbon shares
of 75–100%). Getting to that point requires a complete phase-out of coal power without CCS
by 2050 with strong bioenergy growth in the medium term (co-firing with coal or gas with
CCS). Natural gas acts as a bridging or transitional technology in the short to medium term
and provides ‘virtual’ storage for intermittent renewables. In these scenarios nuclear energy
is a choice, not a requirement.
In IPCC 2014 [22] substantial reductions in emissions would require large changes in
investment patterns. Mitigation scenarios in which policies stabilize atmospheric
concentrations (without overshoot) in the range from 430 to 530 ppm CO2eq by 2100 lead to
substantial shifts in annual investment flows during the period 2010–2029 compared to
baseline scenarios. Over the next two decades (2010 to 2029), annual investment in
conventional fossil fuel technologies associated with the electricity supply sector is projected
to decline by about USD 30 (2–166) billion (median: ‐20% compared to 2010) while annual
investment in low‐carbon electricity supply (i.e., renewables, nuclear and electricity
generation with CCS) is projected to rise by about USD 147 (31–360) billion (median: +100%
compared to 2010).
Carbon prices are also assessed for the mitigation scenarios. The carbon price tends to rise
over time when emissions mitigation effort increases and as a consequence becomes more
expensive to achieve. The inter-model spread in carbon price increases as the required
emissions reduction effort increases, because the models have different technical
capabilities and costs for deep levels of mitigation. The targets analysed in this report
corresponding to temperature goals lower than 2°C and 2.5°C are stringent and as
27
consequences the carbon price reported can diverge amongst models within the same
project. The carbon prices for the 2°C target pathways, in the optimum case scenarios of
early adoption of mitigation policy and availability of all key low-carbon technologies in the
models, range between $15-$115 per ton of CO2 in 2020 and increase to $30-$140 in 2030,
$80-$900 in 2050 and $110-$4000 in 2100 (prices in US$2010). The less stringent 2.5°C
pathways display lower prices mostly before 2050, then after the mid-century the carbon
prices reach similar levels as those for the more stringent target. These values are presented
for the optimal case scenarios; any delays in policy adoption or failure in one of the lowcarbon technologies assumed will rapidly change and increase these prices.
Few studies (for example AMPERE) reported macroeconomic costs of climate change
mitigation policies. Generally macroeconomic costs increase with the stringency of the target
and are higher for the pledge pathways. Delayed action, in general, increases the global
mitigation costs and also leads to a possible fossil fuel lock-in of the electricity system,
creating large and expensive unusable assets. As such the most cost-effective scenarios to
achieve the 2°C target are characterised by early mitigation creating clear signals for lowcarbon technology investments and a near term peak in emissions (with the latest possible
peaking dates occurring between 2025 and 2030, and with peak GHG emissions in the range
of 30 to 50 GtCO2-eq). In the optimal policy scenarios (AMPERE), consumption losses
(2010-2100) are between 0.6-6.5%. The comparable range of the early action scenarios is
between 0.6-6.8 %, and for the late action scenarios between 0.7-6.8 %. Consumption
losses for the least cost pathways to reach the 2°C target in 2100, with the assumption that
all mitigation technologies are available (including all major renewables, nuclear and CCS),
are in the range of 2 to 6% in 2050 and 3 to 11% in 2100 relative to the no mitigation (or
“business as usual”/baseline) scenarios.
28
Table 6: Technology and cost implications
Study / model Share of
/ scenario
fossil fuels
in primary
energy by
2050, 2100
GW deployed of key
technologies (solar, wind,
CCS, nuclear)
Change in
Shadow carbon
Consumption
energy intensity price in 2020, 2030, losses in 2020,
of GDP, 20112050, 2100
2030, 2050, 2100
2050
For each sector:
% reduction in
primary energy
relative to
baseline by
2050/2100
N/A
AMPERE
extrapolation
current
policies
N/A
N/A
2030: up to
30$/tCO2
N/A
N/A
AMPERE 450
immediate
action
N/A
Renewable share of primary N/A
energy 2050=30%,
2100=55%
2050=150 $/tCO2
0.6 to 4 % loss
(cumulative 20102100 with 5%
discount rate)
N/A
AMPERE 450
delayed action
N/A
Renewable share of primary N/A
2050=18% (≈2010) /
2100=65%
2050= 450 to 750
0.5 to 3.7 %
$/tCO2 (high or low
2030 targets)
N/A
AMPERE 550
immediate
action
N/A
N/A
N/A
2050=80 $/tCO2
0.25 to 2% (in 1
model -1.2%)
N/A
AMPERE 550
delayed action
N/A
N/A
N/A
2050= 90 to 100
$/tCO2 (high or low
2030 targets)
0.2 to 1.9%
N/A
29
CCC 2013UC: core
N/A
N/A
EMF 27 BAU
N/A
EMF 27 550
EMF 27 450
Core: in
2005$/tCO2, 219 in
2030, 889 in 2050
N/A
N/A
S=3 (1to12), W=15 (5to20), N/A
N=39 (0to66) EJ/y
Renewable electricity: 15%
(10 to 20) / 18% (10 to 25) &
Non-Electricity RE=2% (1 to
3) / 4% (1 to 6) & Nuclear
electricity =9%
N/A
N/A
N/A
N/A
10 to 30% rel. to
S=15 (5to25), W=25
(15to35), N=45 (8to68) EJ/y BAU
Renewable electricity: =25%
(22 to 32) / 40% (32 to 45) &
Non-Electricity RE =4% (1
to 5) / 12% (5 to 16) &
Nuclear electricity = 17%
time-average
carbon price 5 to 50
$/tCO2
0.3 to2.2 % loss
(cumulative 20102100 with 5%
discount rate) +
one model 8%
N/A
N/A
S=30 (10to35), W=33
(20to35), N=67 (11to214)
EJ/y
time-average
carbon price 12 to
92 $/tCO2
0.8 to3.2 % loss
(cumulative 20102100 with 5%
discount rate) +
one model 11.7%
N/A
Renewable electricity: =32%
(28 to 36) / 55% (40 to 60) &
Non-Electricity RE =6% (4
to 8) / 22% (10 to 26) &
Nuclear electricity = 18%
30
N/A
15 to 50% rel. to
BAU
1.5-2.2%
annually to 2050
GEA 2012
N/A
Nuclear 75-1850 GW in
2050. Renewable (164-651
EJ) 30-75% of primary
energy in 2050.
LIMITS FP7
450 ppm
N/A
2010-30(25,60,15,25);2030- N/A
50(170,140,105,45)GW/y
RCP 2.6
30% (2050)
10% (2100)
RCP 4.5
40% (2050)
N/A
N/A
CO2 price range in
2020 is 15110$tCO2 (200
prices) and it
increases at a rate
of 5% per year.
N/A
N/A
N/A
Cumulative 20102050: $45trillion
comp to base
N/A
$60 tCO2 (2020)
0.3% (2020)
$80 tCO2 (2030)
0.9% (2030)
$160 (2050)
1.7% (2050)
Primary energy
reduced 20%
compared to
baseline in 2100
$200-250 (2100)
0.8% (2100)
$85 tCO2 (2100)
Primary energy
down about 10%
in 2100 cf baseline
60% (2100)
RCP 8.5
31
80% (2100)
N/A
-0.5% per year
N/A
N/A
N/A
RoSE 450
immediate
action
54-61%
(2050)
Up to 240 EJ/year of
nuclear in 2100
18-23%
(2100)
Up to 106 EJ/year of fossil
with CCS in 2100
2.0-2.6% per
year to 2050
35-46 $/tCO2 (2020) 1.4-2.5% GDP
loss cumulatively
$62-105 (2030)
over the period to
$170-403 (2050)
2100
32-52% below
BAU (2100)
$1,200-4,900 (2100)
Up to 300 EJ/year biomass
(of which 195 EJ/year with
CCS) in 2100
23-45% below
BAU (2050)
Up to 285 EJ/year of solar in
2100
RoSE 450
delayed action
to 2020
50-61%
(2050)
Up to 240 EJ/year of
nuclear in 2100
15-22%
(2100)
Up to 97 EJ/year of fossil
with CCS in 2100
Up to 290 EJ/year biomass
(of which 195 EJ/year with
CCS) in 2100
Up to 275 EJ/year of solar in
2100
32
2.0-2.7% per
year to 2050
0-26 $/tCO2 (2020)
$67-116 (2030)
$177-472 (2050)
$1,500-5,600 (2100)
1.5-2.5% GDP
loss cumulatively
over the period to
2100
23-47% below
BAU (2050)
36-53% below
BAU (2100)
RoSE 450
delayed action
to 2030
42-58%
(2050)
Up to 130 EJ/year of wind in 2.0-2.9% per
2100
year to 2050
10-20%
(2100)
Up to 257 EJ/year of
nuclear in 2100
Up to 280 EJ/year of
biomass in 2100 (of which
up to 196 EJ/year with CCS)
0-26 $/tCO2 (2020)
$8-203 (2030)
$212-630 (2050)
1.7-2.9% GDP
loss
cumulatively over
the period to 2100
$2,300-7,500 (2100)
25-50% below
BAU (2050)
34-54% below
BAU (2100)
Up to 309 EJ/year of solar in
2100
RoSE 550
59-77%
(2050)
Up to 118 EJ/year nuclear in 1.9-2.2% per
year to 2050
2100
37-58%
(2100)
Up to 260 EJ/year biomass
in 2100 (of which up to 192
EJ/year with CCS)
Up to 213 EJ/year of solar in
2100
Up to 255 EJ/year of fossil
with CCS in 2100
33
13-30 $/tCO2 (2020) 1.5% GDP loss
cumulatively over
$25-49 (2030)
the period to 2100
$88-137 (2050)
$170-1495 (2100)
18-31% below
BAU (2050)
15-45% below
BAU (2100)
8. Co-benefits and secondary impacts of mitigation
options
Co-benefits and risks are intrinsic to mitigation options chosen for the global transformation
pathways implemented. The large reduction in GHG emissions necessary to fulfil the
stringent targets presented in the report has a significant effect on the energy system (from
primary energy mix to final demand levels). These changes to the energy system induce
side-effects including possible health benefits, changes to energy security or impacts on
biodiversity. These effects are challenging to weight against the costs of mitigation as they
apply to different systems (economic, social and environmental) and are measured in
different units. Some integrated assessment models such as PAGE [29] or FUND [30]
amalgamate some of these side-effects in a relatively simple manner to the global economic
impact of the pathways but debates arise from the materialisation of such side-effects into
monetary quantities. Within a small number of the studies included in the report two cobenefits to mitigation scenarios are reported: impacts on air pollution and energy security.
Impacts on global air quality
The impact on air pollution is reported in AMPERE and the RCPs scenarios as avoided
emissions of NOx and SO2, two important precursors to air quality pollutants: ozone and
particulate matters. These two pollutants have negative impacts on human health, crop
production and building preservation. The reduction in emitted quantities is reported due
exclusively to climate considerations – no air quality policies are included. For mitigation
scenarios achieving a concentration of 450ppm CO2e, the models show strong reductions in
NOx (for example 60% below the baseline in 2050 for RCP 2.6) and more modest reductions
for SO2 (for example 20% below the baseline for RCP2.6). However, no direct effects on
health and morbidity or on crop production are directly reported within the studies.
Impacts on energy security
Within the EMF27 project energy security is analysed for the mitigation scenarios compared
to the baseline energy system. The reported information is orientated toward the qualitative
analysis of energy system security. The study shows that mitigation policies in general
increase national energy sufficiency and resilience via an increase and diversification of
energy sources and carriers. In LIMITS the policy scenarios are combined with large
reductions in fossil fuel dependence (and as a consequence imports) that increases energy
security after 2030. The RoSE project reports that 450ppm scenarios show radical reductions
in energy trade (to almost zero in one for the models in the ensemble - WITCH), whilst
energy diversity rises to mid-century, then declines as renewables start to dominate.
However, some regions’ dependence on imported oil could increase if unconventional
sources are not exploited in mitigation scenarios.
34
Other impacts of mitigation options
Other potential side-effects from the climate mitigation scenarios have been highlighted in
the assessments reports such as IPCC 2014 [22]; these include biodiversity and land use
changes, water consumption, employment. No precise assessments have been found in the
projects analysed.
Table 7: Secondary impacts of mitigation options
Study /
model /
scenario
Air quality details
Energy security details
Other
EMF 27 BAU
Not specified
oil&gas: growing dependence
on import + strong market
concentration / primary
energy: low diversification
N/A
EMF 27 450
Not specified
oil&gas: reduced dependence N/A
on import + marginal impact
on diversity of suply / primary
energy: increased
diversification
LIMITS FP7
Base
Not specified
large global trade FF / FF
extract exceeds reserves /
low diversity (transport)
LIMITS FP7
450 ppm
Not specified
N/A
low global trade FF (present
level) / some FF left in ground
/higher diversity (even in
transport)
RCP 2.6
NOx levels 60% of
baseline (2050)
Not given
Land used for
bioenergy increases
from 3.9 million km2 in
baseline to 9.3 million
km2.
Not given
Bioenergy of >50
EJ/year by 2100
deemed possible by
shifting towards food
products with a smaller
carbon footprint and
reducing beef
consumption.
SO2 levels 20% of
baseline (2050)
RCP 4.5
Not specified
N/A
35
RCP 6.0
NOx about 5-10%
below baseline to 2060,
then declines to about
25% below baseline by
2100.
Not given
No significant change
in bioenergy usage in
the scenario relative to
the baseline.
SO2 about 5-10%
below baseline to 2060,
then declines to about
50% below baseline by
2100.
RoSE
Not specified
N/A
450ppm scenarios show
radical reductions in energy
trade, energy diversity rises to
mid-century, then declines
However, some regions’
dependence on imported oil
could increase if
unconventional sources are
not exploited
9. Current knowledge and data gaps
The goal of this report is to summarise the characteristics of possible long-term
transformation pathways of GHG emissions toward stabilisation of climate change below
long-term temperature targets of around 2°C and 2.5°C. A large number of scenarios
assessed in this report share some common themes:
1. mitigation to 2°C or below is possible but requires early action to keep costs down;
2. at least a 60–80% share of global primary energy will need to come from zero-carbon
options by 2050;
3. the electricity sector in particular will need to be almost completely decarbonized by
mid-century (low carbon shares of 75–100%);
4. achieving a complete decarbonisation of the electricity sector will require a full
portfolio of technologies. In particular, BECCS will be needed to achieve negative
emissions later in the decade and coal power without CCS will need to be completely
phased out by 2050;
5. delays or removal of key technologies makes the requisite levels of mitigation harder
to achieve and / or more costly.
There are a number of areas where there are significant variations between models used
and the comparisons between different studies are therefore challenging. Variations
between models include differences in socio economic assumptions, technology constraints,
for example build rate constraints, which are not explicitly reported in most cases despite
36
being crucial, and assumed level of energy efficiency improvements which are in most cases
not explicitly modelled. It is not clear why some models do not solve when the target is
stringent and what technology constraints are binding. It is noticeable that there are
inconsistencies between models related to methodologies, availability and costs for
technology development. The technical potential and costs for renewables or nuclear power
are uncertain and hard to quantify. In the case of CCS, studies have focused on its
availability as a technology, rather than on the specific challenges and processes to use CCS
effectively, for example the feasibility of storage within regions, or transport to storage sites in
other regions.
Long infrastructure lifetimes mean that it takes decades to change energy systems; so
immediate action is needed to avoid lock-in of invested capital into existing energy systems
and associated infrastructure that is not compatible with longer climate targets. Lock-in is
sometimes explicitly identified as a challenge but the consequences of having massive
capacities of idle assets (such as political feasibility) have not as yet been fully explored. In
general the studies analysed do not account for political and institutional constraints, which is
a challenge for future modeling studies.
Further, within the range of results analysed, scenarios are defined by either concentration,
radiative forcing or temperature targets. Where possible it could be helpful for scenarios to
focus more explicitly on temperature stabilisation rather than concentration or radiative
forcing, as for example temperature stabilisation pathways may have a different emissions
profile than the concentration stabilisation pathways particularly in allowing temporary
overshoots in GHG concentrations.
Global warming impacts including those on energy and land use systems can be critical for
mitigation policy decisions (primary energy availability, energy demand under climate
change, crop production, and afforestation) and should also be included in the model
calculations producing the pathways where possible. GHG emission reductions are a source
of possible co-benefits, which has not been quantified in monetary terms. These co-benefits
include for example air quality, health benefits, sustainable development, and green
employment. Finally, under higher levels of warming, systems such as the climate or the
natural environment may be affected by large positive feedbacks that could trigger tipping
points and extreme events. Taking into account these strong feedbacks and possible high
damages from changing climate could be important for high-end scenarios such as, in
general, the “business as usual” case used as a reference to calculate consumption losses
due to climate mitigation policies. In most studies the baseline or BAU is assumed to see
GDP grow continuously to 2100, unaffected by climate damages.
There are also discrepancies in geographical representation between models used in the
pathway studies. There is a potential large variation in efficiency and cost of mitigation
policies between different countries within the same regional block. Representing in greater
detail the intra-regional differences in economic or technical development within integrated
assessment model regions would be beneficial for future studies.
Even though introduced in some projects (in AMPERE and EMF27), energy efficiency
improvements are usually treated as a single extra scenario without details in technology,
costs and deployments in the same way that can be found for the supply side changes. For
achieving stringent climate targets low energy intensity on the demand side is important
enough to bring more focus on it within the pathway descriptions.
37
The final suggestion from this report would recommend considering the difference between
spending resources on abatement today and investments today in new technology that will
allow lower-cost abatement later, such as renewable investment versus research in CCS and
geo-engineering.
10. Concluding points
This report presents an assessment of recent studies examining greenhouse gas emission
pathways that are compatible with limiting average global temperature rise to levels close to
2°C or at least below 2.5°C by 2100.
Whilst the studies examined show some variation due to differing underlying assumptions in
areas such as technology constraints and socioeconomic factors, strong commonalities
could be found between the studies and the following key conclusions drawn:
1. Mitigation to 2°C or below still is possible; 2.5°C is easier to achieve and can
accommodate delay in decision.
2. Early action and a full portfolio of low-carbon technologies are needed to keep costs
down.
3. Delays or removal of key technologies makes mitigation harder and / or more costly,
and in the case of some models and scenarios, results in more stringent temperature
targets becoming unacahievable.
4. Strong decarbonisation of the electricity sector combined with an increase in
electrification of energy end use are required to achieve stringent targets.
5. Rapid increases in energy efficiency significantly improve the feasibility, and reduce
the cost, of mitigation.
There are over 20 global energy systems/IAM models used in the studies reported in this
report. Each model has its own strength and weakness. Each model has its own
assumptions on various parameters relevant to the models, such as technology costs,
degree of technology availability and constraints on technology deployment rates. Though all
the models are focused on climate change mitigation policy, in order to meet targets of either
2°C or below 2.5°C, there are a number of areas where there is (sometimes significant)
variation between models used, which makes comparisons between the different studies
challenging:
1. Different studies use different socio-economic assumptions on population and GDP
growth
2. Technology constraints are very often not explicitly stated in scenario descriptions,
despite being crucial for the achievability of targets.
3. Assumed levels of energy efficiency not explicitly modelled, at least not to the detail
of which technologies and measures are required on the demand side to achieve the
stated rates of efficiency improvement.
38
4. Costs are often reported in different ways, with some studie focused on carbon
prices, some on consumption or GDP losses in particular years, and some on
cumulative discounted GDP losses.
5. The degree of lock-in of carbon-intensive energy assets under mitigation scenarios is
sometimes identified, but the consequences of idle assets, in terms of economic
costs as well as political economy effects, is most often not fully explored.
6. Important co-benefits of GHG emissions reductions are not quantified in monetary
terms, which makes it difficult to understand the degree to which these might
effectively offset mitigation costs.
These model limitations and variations imply that there remains a detailed research agenda
going forward, to help better understand the real-world drivers and constraints behind the
feasibility of deep emissions reduction scenarios.
11. References
[1] UNFCCC, 1992, United Nations Framework Convention on Climate Change, United
Nations Office, New York, 1992
[2] UNFCCC, 2009. Report of the Conference of the Parties on its fifteenth session, held in
Copenhagen from 7 to 19 December 2009. Part Two: Action taken by the Conference of
the Parties at its fifteenth session, Copenhagen: UNFCC.
[3] UNFCCC, 2012. Report of the Conference of the Parties on its seventeenth session, held
in Durban from 28 November to 11 December 2011. Addendum. Part two: Action taken
by the Conference of the Parties at its seventeenth session, Durban.
[4] UNEP 2012. The Emissions Gap Report 2012. United Nations Environment Programme
(UNEP), Nairobi
[5] UNEP 2013. The Emissions Gap Report 2013. United Nations Environment Programme
(UNEP), Nairobi
[6] Riahi, K. et al., 2011. RCP 8.5 - a scenario of comparatively high greenhouse gas
emissions. Climatic Change, 109(1), pp.33–57.
[7] Den Elzen, M.G.J., Hof, A.F. & Roelfsema, M., 2011. The emissions gap between the
Copenhagen pledges and the 2°C climate goal: Options for closing and risks that could
widen the gap. Global Environmental Change, 21, pp.733–743. Available at:
http://linkinghub.elsevier.com/retrieve/pii/S0959378011000070.
[8] Rogelj, J. et al., 2011. Emission pathways consistent with a 2°C global temperature limit.
Nature Climate Change, 1(8), pp.413–418.
[9] Adam, D., 2009. World will not meet 2C warming target, climate change experts agree.
The
Guardian,
Tuesday
14
April
2009.
<http://www.guardian.co.uk/environment/2009/apr/14/global-warming-target-2c>.
[10] Joshi, M. et al., 2011. Projections of when temperature change will exceed 2°C above
pre-industrial levels. Nature Clim. Change, 1(8), pp.407–412.
[11] New, M. et al., 2011. Four degrees and beyond: the potential for a global temperature
39
increase of four degrees and its implications. Philosophical transactions. Series A,
Mathematical, physical, and engineering sciences, 369, pp.6–19.
[12] Kriegler E., N. Petermann, V. Krey, J. Schwanitz, G. Luderer, S. Ashina, V. Bosetti, A.
Kitous, A. Méjean, L. Paroussos, F. Sano, H. Turton, C. Wilson, and D. van Vuuren
(2013). Diagnostic indicators for integrated assessment models of climate policies.
Technological
Forecasting
and
Social
Change
(DOI:
http://dx.doi.org/10.1016/j.techfore.2013.09.020).
[13] Blanford G., E. Kriegler, and M. Tavoni (2014). Harmonization vs. Fragmentation:
Overview of Climate Policy Scenarios in EMF27. Climatic Change (DOI: DOI:
10.1007/s10584‐013‐0951‐9).
[14] Kriegler E., M. Tavoni, T. Aboumahboub, G. Luderer, K. Calvin, G. DeMaere, V. Krey, K.
Riahi, H. Rosler, M. Schaeffer, and D.P. Van Vuuren (2013c). What does the 2°C target
imply for a global climate agreement in 2020? The LIMITS study on Durban Platform
scenarios. Climate Change Economics. 4, (DOI: 10.1142/S2010007813400083).
[15] Kriegler E., K. Riahi, N. Bauer, V.J. Schanitz, N. Petermann, V. Bosetti, A. Marcucci, S.
Otto, L. Paroussos, S. Rao, and al. (2014). Making or breaking climate targets: The
AMPERE study on staged accession scenarios for climate policy. Technological
Forecasting and Social Change (DOI:http://dx.doi.org/10.1016/j.techfore.2013.09.021).
[16] Riahi K., E. Kriegler, N. Johnson, C. Bertram, M. Den Elzen, J. Eom, M. Schaeffer, J.
Edmonds, and et al. (2014). Locked into Copenhagen Pledges ‐ Implications of
short‐term emission targets for the cost and feasibility of long-term climate goals.
Technological Forecasting and Social Change (DOI: 10.1016/j.techfore.2013.09.016).
[17] Global Energy Assessment – Toward a Sustainable Future, 2012, Cambridge University
Press, Cambridge UK and New York, NY, USA and the International Institute for Applied
Systems Analysis, Laxenburg, Austria.
[18] Kriegler, E., Mouratiadou, I., Luderer, G., Bauer, N., Calvin, K., DeCian, E., Brecha, R.,
Chen, W., Cherp, A., Edmonds, J., Jiang, K., Pachauri, S., Sferra, F., Tavoni, M.,
Edenhofer, O. (2013). Roadmaps towards Sustainable Energy futures and climate
protection: A synthesis of results from the RoSE project (1st edition). Potsdam Institute
for Climate Impact Research, Potsdam.
[19] Luderer, G., Bertram, C., Calvin, K., De Cian, E., Kriegler, E. (2013) Implications of weak
near-term climate policies on long-term mitigation pathways. DOI 10.1007/s10584-0130899-9
[20] Anandarajah, G. and Dessens, O., (2013) Modelling of global energy scenarios under
CO2 emissions pathways with TIAM-UCL, report for the CCC.
[21] McGlade, C., O. Dessens, G. Anandarajah, and P. Ekins (2014) Chapter 24: Global
scenarios of greenhouse gas emissions reduction, Ekins, P. and Watson, J. (eds.) UK
Energy in a Global Context (in preparation for UKERC: London).
[22] Climate Change 2014, Mitigation of Climate Change. Contribution of IPCC AR5 WG3
2014: Working Group III to the Fifth Assessment Report of the Intergovernmental Panel
on Climate Change [Edenhofer, O., R. Pichs-Madruga, Y. Sokona, E. Farahani, S.
Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B. Kriemann, J.
Savolainen, S. Schlömer, C. von Stechow, T. Zwickel and J.C. Minx (eds.)]. Cambridge
40
University Press, Cambridge, United Kingdom and New York, NY, USA.
[23] IAMC AR5 Scenario Database, 2014, hosted by IIASA at https://secure.iiasa.ac.at/webapps/ene/AR5DB/dsd?Action=htmlpage&page=about#intro
[24] Van Vuuren D., E. Stehfest, M.G.J. den Elzen, T. Kram, J. van Vliet, S. Deetman, M.
Isaac, K. Klein Goldewijk, A. Hof, A. Mendoza Beltran, R. Oostenrijk, and B. Ruijven
(2011). RCP2.6: exploring the possibility to keep global mean temperature increase
below 2°C. Climatic Change. 109, 95-116 pp. (DOI: 10.1007/s10584‐011‐0152‐3)
[25] Thomson, A. M., K. V. Calvin, S. J. Smith, G. P. Kyle, A. Volke, P. Patel, S. DelgadoArias, B. Bond-Lamberty, M. A. Wise, L. E. Clarke, J. A. Edmonds (2011) RCP4.5: a
pathway for stabilization of radiative forcing by 2100, Climatic Change 109:77–94 (DOI
10.1007/s10584-011-0151-4)
[26] Davis, S., K. Caldeira nad H. Matthews, 2010, future CO2 emissions and climate change
from existing energy infrastructure, Science 10, 329, DOI:10.1126/science.1188566
[27] IPCC Special Report on Renewable Energy Sources and Climate Change Mitigation,
2011. Prepared by Working Group III of the Intergovernmental Panel on Climate Change
[O. Edenhofer, R. Pichs-Madruga, Y. Sokona, K. Seyboth, P. Matschoss, S. Kadner, T.
Zwickel, P. Eickemeier, G. Hansen, S. Schlömer, C. von Stechow (eds)]. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA, 1075
[28] Stern (2008). The Economics of Climate Change. American Economic Review: papers
and
proceedings
2008,
98:2,
pp.1-37.
http://www.atyponlink.com/AEAP/doi/pdf/10.1257/aer.98.2.1
[29] Hope, C., 2011, The PAGE09 integrated assessment model: A technical description,
Cambridge Judge business School Working Paper.
[30] Tol, R.S.J. (2009), Climate Feedbacks on the Terrestrial Biosphere and the Economics
of Climate Policy: An Application of FUND, ESRI Working Paper 288.
41