CDP–WWF Temperature
Scoring Methodology
A temperature scoring method for
targets, corporates, and portfolios
Open-source methodology to translate
the ambition of corporate greenhouse gas
(GHG) emission reductions into
temperature scores for corporates and
investment portfolios
CDP Worldwide and WWF International
Version 1.5.
May, 2024
1
© CDP Worldwide and WWF International 2024
ABSTRACT
This methodology is an open-source method to enable the translation of corporate greenhouse
gas (GHG) emission reduction targets into temperature scores at a scope, company, and portfolio
level. The methodology allows generating temperature scores for individual scope level targets
(e.g., Scope 1, Scope 2, Scope 3). It also provides a protocol to aggregate scope level scores into
a common intuitive metric reflecting the ambition of the company’s GHG reduction targets. Finally,
the method defines a series of weighting options that enable Financial Institutions (FI) and others
to aggregate the temperature scores of companies in a portfolio to a portfolio temperature score.
This is an update (version 1.5) of the initial publication of the methodology, which was published
in 2020.
The methodology provides a public, transparent, and science-based protocol to assess the
ambition of corporates and portfolios based on the ambition of GHG reduction targets. It enables
users to assess the ambition of public GHG emission reduction targets and can help users
compare the relative ambition of one company’s target versus another. Likewise, the method
allows comparing different portfolio ambitions and FIs to calculate their own portfolio temperature
score, which is a key starting point for aligning the portfolio with long-term temperature goals such
as 1.5C.
2
© CDP Worldwide and WWF International 2024
CDP Worldwide and WWF International
About CDP
CDP is a global non-profit that runs the world’s environmental disclosure system for companies,
cities, states, and regions. Founded in 2000 and working with more than 740 financial institutions
with over US$136 trillion in assets, CDP pioneered using capital markets and corporate
procurement to motivate companies to disclose their environmental impacts, and to reduce
greenhouse gas emissions, safeguard water resources and protect forests. Over 24,000
organizations around the world disclosed data through CDP in 2023, with more than 23,000
companies – including listed companies worth two thirds of global market capitalisation – and over
1,100 cities, states, and regions. Fully TCFD aligned, CDP holds the largest environmental
database in the world, and CDP scores are widely used to drive investment and procurement
decisions towards a zero carbon, sustainable, and resilient economy. CDP is a founding member
of the Science Based Targets initiative, We Mean Business Coalition, The Investor Agenda, and
the Net Zero Asset Managers initiative.
Visit https://cdp.net/en or follow us @CDP to find out more.
About WWF
WWF is an independent conservation organization, with over 30 million followers and a global
network active in nearly 100 countries. Our mission is to stop the degradation of the planet's
natural environment and to build a future in which people live in harmony with nature, by
conserving the world's biological diversity, ensuring that the use of renewable natural resources
is sustainable, and promoting the reduction of pollution and wasteful consumption. Find out more
at panda.org.
3
© CDP Worldwide and WWF International 2024
KEY TERMINOLOGY
Carbon dioxide removal, or “CDR”: a process in which carbon dioxide (CO
2
) is removed from the
atmosphere by deliberate anthropogenic activities and durably stored in geological, terrestrial, or
ocean reservoirs, or in products
1
.
Compound Annual Reduction, or “CAR”: the annualized emissions reduction rate over a specific
period of time, as implied by climate scenarios and corporate GHG emissions reduction targets.
Default temperature score: default °C-value applied to companies in the absence of valid
climate target data (here: 3.4°C, derived from Climate Action Tracker’s “policies & action emission
scenario”).
Warming function: a linear regression model used to project the impact of GHG emissions
reduction rates on global warming by the end of the century.
Global climate models, simulations, and scenarios: Climate models are a mathematical
description of the earth’s climate system. Global coupled climate models include physical
principles of the atmosphere, ocean, land surface, and sea ice. The results from running global
climate models are referred to as model simulations. The scenarios are primarily derived from
Integrated Assessment Models (IAMs) and serve as boundary conditions for global climate
models. They describe possible future pathways, covering a wide range of assumptions regarding,
e.g., GHG emission trajectories, socio-economic trends, and technological developments. For
simplicity, we refer to the model simulations (based on different scenarios) as “scenarios”.
Temperature scores, or “TS”: a forward-looking metric that expresses the GHG emissions
reduction targets of a company, portfolio, or fund with the associated annual global mean surface
temperature rise.
Total GHG emissions: refer to the total of a company’s emission scope 1, scope 2, and scope 3
emissions. Scope 1 refers to direct emissions from sources that are owned or controlled by the
company, scope 2 refers to indirect emissions from purchased energy, and scope 3 refers to
indirect emissions (not included in scope 2) that occur upstream and downstream of a company’s
value chain. In the regression models, the GHG considered depend on the variable used for each
respective scope (see Chapter 4 for more details).
1
See IPCC_AR6_WGIII_Factsheet_CDR.pdf for further information.
4
© CDP Worldwide and WWF International 2024
Contents
1. Fundamentals ..................................................................................................................... 7
1.1. Understanding temperature scores .............................................................................. 7
1.2. What a temperature score does not cover ................................................................... 7
1.3. Warming Function versus Single Scenario ................................................................... 8
2. Key changes to the previous version ................................................................................... 9
3. Introduction and methodological overview..........................................................................10
4. Step 1: Create benchmarks ...............................................................................................12
4.1. Underlying data of linear regression models ...............................................................12
4.2. Linear regression models ............................................................................................13
4.2.1. Sector-specific variables and benchmarks ...........................................................15
4.2.2. Analysis for scenario selection .............................................................................17
4.2.3. Regression results ...............................................................................................19
5. Step 2a: Target validation ..................................................................................................22
5.1. Assigning a temperature score to disclosed targets ....................................................22
5.2. Target validation .........................................................................................................23
5.3. Default temperature score for companies without valid targets ...................................24
5.3.1. Purpose of a default temperature score ...............................................................24
5.3.2. Default score approaches ....................................................................................24
5.4. Temperature floor .......................................................................................................27
6. Step 2b: Company scoring .................................................................................................29
6.1. Target timeframe ........................................................................................................29
6.2. Target quality criteria ..................................................................................................30
6.3. Target criteria waterfall ...............................................................................................31
6.3.1. Target vintage ......................................................................................................32
6.3.2. Boundary coverage ..............................................................................................32
6.3.3. Target types .........................................................................................................33
6.3.4. Scope coverage ...................................................................................................34
6.3.5. Reduction ambition ..............................................................................................34
6.3.6. Target year ..........................................................................................................34
5
© CDP Worldwide and WWF International 2024
6.4. Multiple scope 3 targets ..............................................................................................35
6.4.1. Aggregation of multiple scope 3 temperature scores............................................36
6.5. Temperature score aggregation ..................................................................................37
6.6. Using temperature scores ...........................................................................................37
7. Step 3: Portfolio scoring .....................................................................................................39
7.1. Weighting objectives and principles ............................................................................39
7.2. Weighting options .......................................................................................................40
7.3. Weighting method assessment ...................................................................................43
7.4. Additional notes on the portfolio scoring ......................................................................46
8. Limitations and outlook ......................................................................................................47
8.1. Methodology limitations ..............................................................................................47
8.1.1. Linear regression model ......................................................................................47
8.1.2. Formula to compute annual reduction rates .........................................................48
8.1.3. Sector granularity .................................................................................................49
8.1.4. Climate target metric(s) ........................................................................................50
8.1.5. Forward- versus backward-looking assessment ...................................................50
8.1.6. Carbon credits and carbon offsets .......................................................................51
8.1.7. Assurance of GHG and GHG reduction target data..............................................51
8.2. Outlook: Next steps and future research .....................................................................51
9. Version control ...................................................................................................................53
9.1. Change log .................................................................................................................53
10. Consultation questions .......................................................................................................55
10.1. Step 2b .......................................................................................................................55
10.1.1. Should older targets be considered in the absence of recent target data? ........55
10.1.2. Should TS aggregation based on GHG weighting be conducted based on base
year or current year data? ..................................................................................................55
10.1.3. Aggregation when missing GHG data ...............................................................56
10.1.4. Weighting of multiple scope 3 targets ...............................................................56
10.1.5. Scoring of portfolios including targets using the CDP–WWF Temperature Score-
or Temperature Rating methodology ..................................................................................57
10.1.6. Do you agree with the scoring approach taken in Step 2b, Section 6.2.3.1. (see
“Target type exception”)? If not, what alternative suggestion do you have? .......................57
6
© CDP Worldwide and WWF International 2024
10.1.7. Do you agree with the general approach taken in Step 2b, Chapter 6 “Company
scoring”? 57
10.2. Step 1 .........................................................................................................................57
10.2.1. Do you agree with the approach taken in Step 1, “Create benchmarks”? ........57
10.3. Step 2a .......................................................................................................................57
10.3.1. Do you agree with the approach taken in Step 2a, Chapter 5 “Target validation”?
57
10.4. Step 3 .........................................................................................................................57
10.4.1. Do you agree with the approach taken in Step 3, Chapter 7 “Portfolio scoring”?
57
10.5. General questions .......................................................................................................57
10.5.1. Do you have suggestions for future research that could inform the next version of
the CDP–WWF Temperature Score methodology? ............................................................57
10.5.2. Do you have any other comments? ..................................................................57
11. References ........................................................................................................................58
12. Annex ................................................................................................................................60
Annex 1: Warming Function versus Single Scenario .............................................................60
Annex 2: Details of sector variables .......................................................................................62
Annex 3: Result of linear regression model ...........................................................................63
Annex 4: Calculation examples .............................................................................................69
Annex 5: Summary of required data for applying the Temperature Scoring method ..............73
Annex 6: Tables, figures, and equations ................................................................................76
Tables ................................................................................................................................76
Figures ...............................................................................................................................76
Equations ...........................................................................................................................77
7
© CDP Worldwide and WWF International 2024
1. Fundamentals
Given the diversity in climate target scopes, timeframes, and metrics used by companies across
various sectors, understanding, and comparing the adequacy of companies’ decarbonization
goals can be challenging. Initially published in October 2020, this methodology document was
designed to serve as a protocol for assessing and comparing the ambition of companies’ GHG
emissions reduction targets. The key audience of this methodology are Financial Institutions (FIs)
and large corporates wishing to assess, steer, and set targets on their scope 3 emissions as well
as data providers offering ITR metrics. Other users of this methodology include academia and civil
society, as well as supervisory authorities and regulators.
1.1. Understanding temperature scores
The first version of this method introduced a scoring methodology which translates diverse GHG
reduction targets into an intuitive metric expressed in projected warming by 2100. This metric,
sometimes referred to as an Implied Temperature Rise (ITR), can be used to compare the
ambition of different companies’ decarbonization goals as expressed in their public GHG
emissions reduction targets. Henceforth, this output metric will be referred to in this document as
a temperature score (TS).
These scores should not be interpreted like IPCC results (i.e., global climate projections) nor does
the metric predict a certain temperature outcome. Temperature scores instead allow a relative
comparison of climate ambition with respect to the temperature goals of the Paris Agreement. The
scores should be interpreted as follows: This company’s GHG reduction target implies an annual
reduction rate that is consistent with an ambition heading towards X°C – under the assumption
that all companies behave the same.
Users of TS can use this metric to engage with companies to set targets (or improve existing
ones), compare the ambition of corporate GHG emissions targets, measure the alignment of their
own scope 3, and set targets accordingly (e.g., at the portfolio level for FIs, or at the supply chain
level for large companies).
1.2. What a temperature score does not cover
Temperature scores are not intended to serve as a comprehensive metric summarising a
company’s climate transition performance or overall “green credentials”. Temperature scores do
not provide insights into a company’s operational or financial performance relative to these
ambitions, the current trajectory of the company’s historical GHG emissions, or the existence of a
credible climate transition plan to achieve these goals.
The primary purpose of this methodology’s TS is to assess a company’s climate target ambition
through a broad benchmark of climate scenarios. It is therefore suggested that temperature scores
are used as a comparative tool for assessing the climate ambition across multiple companies (e.g.,
within a supply chain or financial portfolio) or, on an aggregated level, of portfolios and supply
chains and that other complementary metrics are consulted to obtain a complete picture of a
8
© CDP Worldwide and WWF International 2024
company’s climate profile. Also, TS should not be used as a predictive tool for estimating a precise
degree of global warming.
1.3. Warming Function versus Single Scenario
To calculate a temperature score, this methodology benchmarks companies’ committed GHG
reduction ambition against a statistical regression model based on all vetted modelled scenarios
of the IPCC’s Sixth Assessment Report (AR6) (see Chapter 4). This approach is referred to as the
“Warming Function” and establishes a linear statistical relationship between the rates of GHG
emissions reduction and the projected temperature outcomes these scenarios imply by the end of
the century.
The Warming Function differs from a Single Scenario approach. A Single Scenario approach relies
on one single scenario and assumes that the future will unfold as per the selected single scenario’s
underlying assumptions. Because the Warming Function approach relies on multiple scenarios, it
reduces the scenario selection bias inherent to the Single Scenario approach. It also provides
greater comparability between implementations of the CDP–WWF method from different data
providers and users. This is as comparability between temperature scores using single scenario
approaches rely on the use of the same scenario for TC computation. However, the robustness
the CDP–WWF warming function provides is traded against less transparency on the effects of
the different assumptions underlying the input models and scenarios of a Warming Function on
the temperature scores. Another difference is that the Single Scenario approach, by using only
one scenario, can more easily allow for more granular, sector-specific analysis compared to this
methodology’s Warming Function. Further research and development are needed to allow the use
of more sector-specific warming functions, in addition to the sectors currently covered with their
own warming functions: steel, aluminium, cement, and power generation.
CDP and WWF have considered the different approaches and concluded that the robustness of
the warming function is preferable, despite its trade-offs, as it provides a more comprehensive,
comparable, and unbiased view of potential temperature outcomes.
Please refer to Annex 1: Warming Function versus Single Scenario for further details about the
rationale supporting that conclusion.
9
© CDP Worldwide and WWF International 2024
2. Key changes to the previous version
This version 1.5 of the CDP–WWF Temperature Scoring methodology (formerly known as CDP–
WWF Temperature Rating) marks the first update since its initial publication in October 2020
(version 1.0).
The primary objectives of this update revolve around refreshing this methodology’s benchmarks
with the latest climate science. It also implements adjustments and improvements on specific
aspects to ensure it remains relevant, fair, and effective. Finally, this version adds transparency
on methodological choices and implications through the presentation of additional analysis,
enhanced explanation, and dedicated sections.
Key changes introduced in this version include (please refer to the change log in Chapter 9 for
more details):
- Update of the input climate model simulations and scenarios of the linear regression
models by substituting the Intergovernmental Panel on Climate Change (IPCC) SR15
model simulations and corresponding scenarios with the IPCC’s AR6 set of climate
simulations with scenarios, including adjustments of scenario filtering decisions (see
Chapter 4 for more details).
- Revision of the calculation formula for the annual rate of GHG emissions reduction,
transitioning from a Linear Annual Reduction to a Compound Annual Reduction
approach (see Section 4.2 for more details).
- Introduction of a specific scope 2 benchmark relying on energy related variables. Scope
1 and scope 2 assessments are now carried out at the single scope level (see Section
4.2.1 for more details).
- Clarification and enhancement of the target selection process (“waterfall”) to prioritize
targets when a company reports multiple targets within the same scope category and
timeframe (see Section 6.3 for more details).
- Update of the target timeframe definition for short-, medium-, and long-term (see
Chapter 6 for more details).
- Update of the default score from 3.2°C to 3.4°C for companies without valid targets or
insufficient data disclosure, reflecting the latest projection based on real-world action and
current policies (see Section 5.3 for more details).
- Clarification on the best possible temperature score for a company, introducing a 1.5°C
temperature floor (see Section 5.4 for more details).
- Addition of a dedicated chapter detailing this methodology’s purpose and intended
outcomes (Chapter 1). Another new chapter focuses on the methodology’s key
limitations and a roadmap for further updates planned (Chapter 8). Finally, a
presentation of the rationale for relying on a warming function over single-scenario-
based benchmarks was added in Chapter 1 and in Annex 1: Warming Function versus
Single Scenario.
10
© CDP Worldwide and WWF International 2024
3. Introduction and methodological overview
Companies are directly responsible for a significant portion of global GHG emissions
(Intergovernmental Panel on Climate Change (IPCC) (Ed.), 2022). GHG emission reduction
targets are a partial but relatively crucial forward-looking marker of a company’s ambition to
mitigate its climate impact. In 2023, close to 5000 companies covering approximately 10 gigaton
(GT) of scope 1+2 GHG emissions, publicly reported GHG emissions targets through CDP (based
on CDP data 2023). However, assessing and comparing the ambition of corporate targets has
traditionally been complex as targets can be expressed with different metrics, over multiple
timeframes and cover various types of emission scopes.
The aim of a temperature score is to translate GHG emission targets into a single intuitive metric
that is linked to the long-term temperature projections associated with the ambition of the target.
In the initial publication of this methodology (version 1.0), a protocol for expressing (“scoring”)
climate targets in a temperature metric referring to projected warming by 2100, was presented.
This updated version builds on the original method and further develops that protocol.
The methodology is composed of three steps, represented in Figure 1:
Figure 1: Steps of the Temperature Scoring methodology
The benchmark creation step (Step 1, Chapter 4) consists in running linear regression models.
The outputs of these models are based on a warming function, derived from all vetted model-
based scenarios in the IPCC Sixth Assessment Report (AR6) Scenario Explorer and Database
2
hosted by the International Institute for Applied Systems Analysis (IIASA) (Byers et al., 2022).
The linear regression models allow the assessment of end-of-century temperature outcomes
expected from short, medium, and long-term projected changes in absolute GHG emissions or
GHG emissions intensity metrics. As such, the regression models are used to relate target
ambition (measured in committed rate of GHG emission reduction) to warming projections by the
end of the century (expressed in centigrade temperature change compared to preindustrial levels).
2
Accessible through this link: https://data.ece.iiasa.ac.at/ar6/#/login?redirect=%2Fworkspaces.
Create benchmarks:
IPCC scenarios and
linear regression
models for different
variables and
timeframes
Step 2a
Target validation:
Assess validity of
targets
Portfolio scoring:
Aggregation of multiple
company scores into
one portfolio score
Step 2b
Company scoring:
Select targets and
compute scope-level
and company-level
temperature scores
Step 1
Step 2 (Company-level assessment)
Step 2 (Company-level assessment)
2 (
Company-level assessmentStep 1assessment)
Step 3
11
© CDP Worldwide and WWF International 2024
As companies often have multiple climate targets, covering different emission scopes and
timeframes, and users may receive data from several sources, Step 2a (Chapter 5) defines the
process and criteria for validating the various company’s GHG reduction targets. This step can be
seen as an eligibility screening of targets allowed as input for temperature score computation. In
Step 2b (Chapter 6), scope-level targets are selected using the selection hierarchy (“waterfall”),
and scope-level temperature scores are calculated. Finally, these TS are then aggregated into
combined company-level scores. The target validation step (Step 2a) defines the minimum quality
criteria for determining the acceptability of a GHG emissions reduction target to be scored. The
company scoring step (Step 2b) specifies the process required to select target data to be used for
scoring and how to aggregate multiple targets to produce company-level scores.
The final step (Step 3, Chapter 7) is used to weight company scores when assessing an
aggregation of companies, such as a financial portfolio or a company value chain.
In addition to computing temperature scores for disclosed targets, the methodology also defines
an approach to address non-disclosing companies. Default scores are introduced, also to allow
TS aggregation for company-, portfolio- or supply chain-level TS (see Section 5.3 on default
scores).
12
© CDP Worldwide and WWF International 2024
4. Step 1: Create benchmarks
4.1. Underlying data of linear regression models
The linear regression models used in this methodology are based on underlying data from global
climate models with scenarios from the IPCC’s AR6 (see Section 4.2 for more information about
the linear regression models). The data is collected and downloaded from the AR6 Scenarios
Database hosted on a Scenario Explorer by the International Institute for Applied Systems
Analysis (IIASA), released in 2022 (Byers et al., 2022)
3
.
Climate models are a mathematical description of the earth’s climate system. Global coupled
climate models include physical principles of the atmosphere, ocean, land surface, and sea ice.
The results from running global climate models are referred to as model simulations. The
scenarios, on the other hand, are primarily derived from Integrated Assessment Models (IAMs)
and serve as boundary conditions for global climate models. They describe possible future
pathways, covering a wide range of assumptions regarding, e.g., GHG emission trajectories,
socio-economic trends, and technological developments. The AR6 Scenarios Database contains
a large number of model simulations with different scenarios which in turn could be classified into
categories based on the projected change in temperature by the end of this century and the
respective probability
4
. The IPCC undertook a vetting process for all model simulations and
scenarios reporting global data to ensure that key indicators such as GHG emissions and energy
are within reasonable ranges
5
(Intergovernmental Panel on Climate Change (IPCC) (Ed.), 2023).
In total, approximately 1,200 different model simulations and scenarios related to GHG emissions
passed this vetting process (Byers et al., 2022). To allow for higher reliability when generating the
linear regression models within this methodology, only model simulations with scenarios that
passed the IPCC’s vetting process are considered. For simplicity, we refer to the model
simulations (based on different scenarios) as “scenarios”.
3
Copyright 2022 IIASA, Publication date: 09/11/2022. Downloaded from this link:
https://data.ece.iiasa.ac.at/ar6/#/data-download.
4
Scenarios are classified into the following categories (C): C1: limit warming to 1.5°C (>50%) with no or
limited overshoot, C2: return warming to 1.5°C (>50%) after a high overshoot, C3: limit warming to 2°C
(>67%), C4: limit warming to 2°C (>50%), C5: limit warming to 2.5°C (>50%), C6: limit warming to 3°C
(>50%), C7: limit warming to 4°C (>50%), and C8: exceed warming of 4°C (>=50%).
5
For more information regarding the IPCC’s vetting process, please read Annex III: Scenarios and
Modelling Methods.
13
© CDP Worldwide and WWF International 2024
4.2. Linear regression models
The underlying data described in Section 4.2.1 is processed and modelled into linear regression
models in two R scripts: CDP–WWF_ITR_preparation_of_data.R and CDP–
WWF_ITR_Regression.R processes AR6 data from two different data file publications
6
, while the
linear regression models are created in CDP–WWF_ITR_Regression.R. The R scripts are open
source and are available upon publication of this methodology https://github.com/WWF-
Sweden/ITR-regression.
One linear regression model is created for each of the following variables and time horizon
7
.
Table 1: Linear regression models associated with different time horizons and variables
Emissions | Kyoto
Gases
Emissions | CO
2
|
Energy | Supply
Emissions | CO
2
|
Energy and Industrial
Processes
Emissions | Kyoto
Gases /
GDP | PPP
Emissions| CO
2
|
Energy | Supply /
Secondary energy
Regression on 5-year
horizon
Regression on 5-year
horizon
Regression on 5-year
horizon
Regression on 5-year
horizon
Regression on 5-year
horizon
Regression on 10-year
horizon
Regression on 10-year
horizon
Regression on 10-year
horizon
Regression on 10-year
horizon
Regression on 10-year
horizon
Regression on 30y
horizon
Regression on 30-year
horizon
Regression on 30-year
horizon
Regression on 30-year
horizon
Regression on 30-year
horizon
This gives a total of 15 linear regression models that we refer to as benchmarks. The regression
coefficients are outlined in Table 3 in Section 4.2.3.
Equation 1: Linear regression formula
where:
= error term of the regression model.
= dependent variable, derived from MAGICC v7.5.3 (Model for Assessment
of Greenhouse Gas Induced Climate Change)
8
(Byers et al., 2022), and collected from the AR6 Scenario Database.
6
Two published data files are used: AR6_Scenarios_Database_World_v1.1 containing time series of different variables, and
AR6_Scenarios_Database_metadata_indicators_v1.1 containing metadata related to the climate models and scenarios. Copyright
2022 IIASA, Publication date: 09/11/2022. Downloaded from this link: https://data.ece.iiasa.ac.at/ar6/#/data-download.
7
Unique variables: Emissions|Kyoto Gases, Emissions|CO2|Energy|Supply, Emissions|CO2|Energy and
Industrial Processes, Emissions|CO2|Energy|Supply / Secondary energy and Emissions|Kyoto Gases /
GDP|PPP.
Three different time horizons: 5y, 10y and 30y.
8
Based on the variable AR6 climate diagnostics | Surface Temperature (GSAT) | MAGICCv7.5.3|50.0th
Percentile.
14
© CDP Worldwide and WWF International 2024
This variable is suitable for this methodology’s aim – to translate GHG emission reduction targets into a single common
and intuitive metric – as it returns a single unambiguous value expressed in projected temperature change in 2100.
= independent variable, is the annualized
reduction rate implied by the variable’s absolute change between two points in time, starting in 2020. This is expressed
as the compound annual reduction rate (CAR) for different time horizons – from five to 30 years (in five-year intervals).
Equation 2: Compound Annual Reduction Rate
where:
= compound annual reduction
= in the context of scenarios the year of reference is 2020. In the context of corporate targets, the year the
target was set.
= in the context of scenarios, the 5-year interval period after 2020 analyzed. In the context of corporate targets,
the year the target should be met.
= the percentage change of emissions between the base year and
end year (e.g., if the scenario’s GHG emission pathway shows a reduction in emissions by 50% between the two
periods, then the value should be −0.5). Note that when targeted reduction is 100%, the equation for CAR does not give
a meaningful value. Therefore, in the context of corporate targets, the method assigns the temperature floor (see Section
5.4) as the TS (see more under limitations in Section 8.1.2)
In this methodology, the time frames used for benchmarking targets are 5 years (short term), 10
years (medium term) and 30 years (long term). Year 2020 represents the base year. For example,
to run a regression model of projected temperature change in 2100 and GHG emissions’ evolution
in the next 30 years, the GHG emissions CAR from all scenarios’ emissions pathways between
2020 and 2050 are first calculated.
For an absolute reduction in emissions between two intervals, CAR will be negative as per the
formula above. To run the regression models the sign of CAR is flipped (a reduction in emissions
will be counted as a positive CAR) so that slopes in the regressions are negative and a reduction
in emissions can be more intuitively interpreted as an improvement in temperature outcomes when
calculating targets’ temperature scores. This transformation has no impact on the outcome of the
methodology and similar temperature scores would be observed without it, but it should be noted
if applying the regression coefficients provided in this paper.
15
© CDP Worldwide and WWF International 2024
Box 1: from Linear Annual Reduction (LAR) to Compound Annual Reduction (CAR).
The change in the annual reduction rate formula from linear annual reduction (LAR), used in the
initial publication of the methodology (version1.0), to compound annual reduction (CAR) has an
expected effect on the fit of the linear model as calculated by R
2
. LAR expresses changes in the
independent variable over the period in terms of percentage points (i.e., GHG emissions reduced
by two percentage points each year indicates that they are reduced from 100% to 98% to 96%
from a given baseline value), as opposed to annualized percentages with CAR (reduction of 2%
from the 2020 baseline value, and reduction of 2% from the 2021 value the following year, and so
on). LAR is normalized by nature with a significantly smaller variance, which mechanically
increases R
2
of the regressions, especially for long-time horizons. CAR represents a cumulative
reduction rate that is more accurately modelled by an exponential function as highlighted in Figure
2 (see Section 4.2.3). Investigating the possibility of using another statistical model will be subject
for a next update of this methodology.
The main reason for changing from LAR to CAR is to improve the interpretation of the model
results with respect to real economy changes in GHG emissions that are usually expressed in
annualized percentage changes as opposed to reduction in percentage points from a base year.
This change also incentivizes earlier action, as the CAR implies a higher absolute reduction in
GHG emissions in the short term given the higher baseline.
4.2.1. Sector-specific variables and benchmarks
Table 2 outlines which regression models that are used to benchmark sector specific targets in
this methodology. Some benchmarks are common across sectors and scopes.
Table 2: Sector variables and associated linear regression models for each target type and scope category.
9
Sector
10
Target
type
Scope 1 benchmarks:
AR6 regression
model variable
Scope 2 benchmarks:
AR6 regression
model variable
Scope 3 benchmarks:
AR6 regression
model variable
All sectors
(except for the
ones listed below)
Absolute
Emissions | Kyoto
Gases
Emissions | CO
2
|
Energy | Supply
Emissions | Kyoto
Gases
Intensity
Emissions | Kyoto
Gases /
GDP | PPP
Emissions | CO
2
|
Energy | Supply /
Secondary energy
Emissions | Kyoto
Gases /
GDP | PPP
9
Some of the variables in the table are expressed in CO2 rather than CO2e or Kyoto Gases. However, this
is the nature of the specific AR6 variable, and CO2 is considered the best proxy for CO2e/Kyoto Gases in
this methodology.
10
In this paper, sectors are defined following the CDP-Activity classification system.
16
© CDP Worldwide and WWF International 2024
Power
Generation
Absolute
Emissions | CO
2
|
Energy | Supply
The regression model
for all sectors applies
The regression model
for all sectors applies
Intensity
Emissions| CO
2
|
Energy | Supply /
Secondary energy
The regression model
for all sectors applies
The regression model
for all sectors applies
Cement/
Steel/
Aluminium
Absolute
Emissions | CO
2
|
Energy and Industrial
Processes
The regression model
for all sectors applies
The regression model
for all sectors applies
Intensity
The regression model
for all sectors applies
The regression model
for all sectors applies
The regression model
for all sectors applies
These are used to assess corporate GHG targets according to their type (using absolute or
intensity metrics), sectors (All sectors, Power Generation, and Cement/Steel/Aluminium), and
emission scope category (scope 1, 2, and 3). The sector variables used to benchmark absolute
targets are directly available from the AR6 Scenario Database. The intensity variables are
calculated from AR6 scenarios data for the purpose of this methodology (this calculation is
included as part of the CDP_WWF_ITR_Regression.R script mentioned in Section 4.2).
The benchmarks are selected and allocated to a sector based on a combination of criteria: data
availability, suitability of the AR6 variable to proxy the specific sector emissions, and the fit of the
linear regression model when using this benchmark.
The following updates are introduced in this methodology compared to the initial version (1.0):
- A specific benchmark for scope 2 targets based on the following variables to proxy indirect
emissions from energy consumption: Emissions | CO
2
| Energy | Supply, and Secondary
energy output for the intensity targets
11
. This change relies on the assumption that energy
consumption’s absolute emissions and intensity should follow a similar path as the supply
of energy. This benchmark is now also used to assess Power Generation scope 1 targets.
Previously, one common benchmark is used to assess both scope 1 and scope 2 targets
for all sectors, and the Power Generation scope 1 emissions are assessed according to a
different benchmark
12
.
- The sector-specific benchmark for Primary Energy is removed in this methodology since
the variable used in the initial version (version 1.0) is considered too generic and
11
These two variables are related to GHG emissions from fuel combustion and fugitive emissions from
fuels: electricity and heat production and distribution, other energy conversion (e.g. refineries, synthetic fuel
production, solid fuel processing) including pipeline transportation fugitive emissions from fuels and
emissions from transport and storage (Byers et al., 2022).
12
The variable Emissions | CO
2
| Energy and Industrial Processes is used for absolute targets and
Emissions | CO
2
| Energy | Supply | Electricity / Secondary Energy | Electricity for intensity targets
17
© CDP Worldwide and WWF International 2024
alternative suitable variables yielded low linear fits in regression models. Therefor the “all
sectors” benchmarks are now used to assess companies’ targets in the fossil fuel sector.
See Annex0 for a more detailed description of the variables. Future updates of this methodology
aim to conduct further research to inform the possible inclusion of additional sector specific
regression models.
4.2.2. Analysis for scenario selection
In this version of the methodology, all scenarios that passed the vetting process from the IPCC
are selected to feed into the linear regression models. After conducting analysis on the potential
impact of excluding certain scenarios based on additional considerations detailed below, the
overall conclusion is not to perform any exclusionary filters.
In the initial publication of this methodology (version 1.0), the IPCC’s SR15
13
scenarios were first
filtered before the linear regression models were generated. This was done by creating different
scenario sets that matched normative and precautionary preferences concerning overshoot and
the level of plausible carbon dioxide removal (CDR above 10 Gt CO
2
/year considered as not
plausible)
14
,
15
. Baseline scenarios, i.e., scenarios where no deliberate mitigation action was taken,
were also removed from the initial SR15 dataset. Besides these normative, precautionary
preferences, the best model fit over medium- and long-term horizons
16
was the basis for selecting
and applying scenario set 4 in the version 1.0 methodology. This scenario set applied a CDR limit
to maximum 10 Gt CO
2
/year as well as excluding baseline scenarios.
When applying a similar CDR-related precautionary consideration to the AR6 dataset, the results
are contradictory to the ones using data from the SR15. While intuitively, excluding scenarios
based on high CDR (defined as >10 Gt CO
2
/year) is expected to increase the GHG emission
reduction ambition
17
reflected in the regression models, the opposite outcome was found. After
investigation, the potential reasons identified are the following: firstly, and most significantly, the
underlying dataset of AR6 is fundamentally different compared to the one of SR15 (e.g., different
report purpose, other and updated climate models and scenarios are used, etc.), hence why
similar normative precautionary preferences and filter options will not necessarily lead to
13
Global Warming of 1.5°C, IPCC Special Report on the impacts of global warming of 1.5°C above pre-
industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the
global response to the threat of climate change, sustainable development, and efforts to eradicate poverty.
Accessible through this link: Download Report — Global Warming of 1.5 ºC (ipcc.ch).
14
56 unique scenario sets were generated in the initial methodology (version 1.0).
15
For more information about the scenario filtering process of the initial methodology (version 1.0), please
review https://www.cdp.net/en/investor/temperature-ratings/CDP–WWF-temperature-ratings-methodology
16
Represented by R^2.
17
The ambition is defined as the GHG emission annual reduction rate required to reach 1.5°C warming in
2100 based on the regression coefficients. The lower estimate rate, then the lower the ambition.
18
© CDP Worldwide and WWF International 2024
regression outcomes as in the initial methodology. Secondly, in AR6, the levels of yearly CDR
tend to be high and increase by the end of the century compared to the first half. This is also the
case for scenarios with an ambitious GHG emissions reduction rate that project to limit warming
to 1.5°C by the end of the century. As a result, excluding scenarios based on such CDR criteria
(i.e., CDR>10 Gt CO
2
/year) also means excluding ambitious 1.5°C-scenarios, leading to an overall
decrease of the GHG emission reduction ambition reflected in the linear regression models (rather
than in an increase). Following these findings, no CDR filter is applied to the AR6 dataset.
Further analysis shows that excluding baseline scenarios from the AR6 dataset has no significant
impact on the annual GHG emission reduction ambition reflected in the regression models. This
insight, in combination with the possibility that these baseline scenarios are potential future
trajectories of the world’s development, leads to the conclusion to keep baseline scenarios in the
dataset used to derive the benchmarks.
Additional analysis to inform the filter choices was conducted. The effect on the projected
temperature change at the end of the century (dependent variable in the linear regression models)
and the annual GHG emissions reduction rate (independent variable in the linear regression
models) was tested for two additional filters: 1) the exclusion of scenarios with a GHG emission
peak year before year 2024, and 2) outlier analysis using statistical analysis of Cook’s D, leverage,
and residuals.
The results from filter test 1) shows that the GHG emission reduction ambition significantly
decreases from a short time frame perspective. However, the effect is less noticeable in the longer
term. One possible explanation for this outcome could be that very few scenarios with a projected
1.5°C temperature outcome by the end of the century
18
assume a peak year of GHG emissions
after 2020. Thus, removing scenarios based on such condition will decrease the overall GHG
emissions reduction ambition, leading to steeper regression lines. In addition, as demonstrated in
the linear regression models (see Figure 2 in Section 4.2.3) the variability of the scenarios’ GHG
emission reduction is, in general, larger in the shorter time frames compared to the longer time
frames, which could explain why the effect is less noticeable in the longer term. This result leads
to the conclusion not to remove any scenarios from the dataset based on the results from filter
test 1.
The outlier analysis (filter test 2) highlights that the number of scenarios considered as outliers
according to the three approaches (Cook’s D, leverage, and residuals) is limited. The result also
shows that no scenario alone is deemed to influence the linear regression model to a significant
extent, and hence, no scenario is removed from the data set based on the results from filter test
2.
18
Scenarios classified as Category 1: limit warming to 1.5°C (>50%) with no or limited overshoot, and
Category 2: return warming to 1.5°C (>50%) after a high overshoot.
19
© CDP Worldwide and WWF International 2024
The overall conclusion from the scenario selection analysis is that the vetting process from IPCC
is considered sufficient criteria for selecting scenarios for the linear regression models in this
methodology. As a result, the linear regression models computed and used in this methodology
include all scenarios that passed the IPCC’s vetting process.
4.2.3. Regression results
Figure 2 provides an illustration of the linear regression model applied to the variable Emissions |
Kyoto Gases on the 30-year horizon
19
. The remaining linear regression results for different time
horizons are found in Annex 3: Result of linear regression model.
Figure 2: Example of the result of the linear regression model for Emissions | Kyoto Gases for a 30-year time horizon
Table 2 summarizes the details of the linear regression models for 5-, 10- and 30-year horizons
representing short-, medium- and long-term targets, for each variable used in this methodology.
The fit of the regression line, represented by R
2
,
and the intercept increase as the time horizon
increases (intercept of 2.4, R
2
of 0.40 for 5 years; intercept of 2.8, R
2
of 0.74 for 30 years). This is
logical since the degree of variability between scenarios decreases over longer horizon and the
range of possible annualized reduction rates leading to a given temperature outcome is lower for
19
Time horizons used in this methodology: 5-year (short term), 10-year (medium term), and 30-year (long
term). Base year is 2020.
20
© CDP Worldwide and WWF International 2024
longer time horizons. In addition, no action (zero year-on-year reduction) will lead to higher
temperature outcomes if observed over longer time horizons.
Compared to the initial methodology (version 1.0) where the scenarios were grouped into different
scenario sets based on the filtering process (see Section 4.2.2), this methodology only contains
one scenario set including all the vetted scenarios from the IPCC AR6 database (see Section
4.2.24.2.2).
Table 3: Summary of linear regression results (note that rounding differences may occur).
Regression
model
variable
5-year horizon
10-year horizon
30-year horizon
Sample
size
Intercept
Slope
R
2
Sample
size
Intercept
Slope
R
2
Sample
size
Intercept
Slope
R
2
Emissions |
Kyoto Gases
1,115
2.40
−0.21
0.40
1,115
2.46
−0.24
0.46
1,112
2.81
−0.30
0.74
Emissions | CO
2
| Energy |
Supply
1,160
2.35
−0.12
0.32
1,156
2.40
−0.11
0.36
851
2.85
−0.15
0.63
Emissions | CO
2
| Energy and
Industrial
Processes
1,171
2.27
−0.22
0.37
1,170
2.33
−0.24
0.43
1,155
2.58
−0.19
0.56
Emissions |
Kyoto Gases /
GDP | PPP
986
3.13
−0.21
0.38
986
3.21
−0.23
0.44
983
3.57
−0.30
0.70
Emissions| CO
2
|
Energy | Supply
/ Secondary
energy
764
2.19
−0.08
0.15
764
2.22
−0.07
0.17
575
2.61
−0.11
0.41
To calculate the temperature score of a company’s climate target, the target is first mapped to the
most representative benchmark available in the methodology (see Table 2). Depending on the
target’s time-horizon, the appropriate regression parameters (see Table 3) can be applied
according to the following formula:
Equation 3: Temperature Score
where:
= intercept of the regression model for a given time horizon and variable
21
© CDP Worldwide and WWF International 2024
= slope of the regression model for a given time horizon and variable
= compound annual reduction of the variable over time horizon implied by the target set by the company.
Using the formula expressed above, if CAR is 0 (i.e., the company’s absolute reduction of
emissions over the next 10 years is zero) then the projected temperature outcome of the target
will amount to the intercept of the linear regression.
The next section provides further details on the validation and assessment of targets (Step 2).
22
© CDP Worldwide and WWF International 2024
5. Step 2a: Target validation
5.1. Assigning a temperature score to disclosed targets
The methodology assumes that there is a linear relationship between the change in common
climate target metrics (e.g., absolute or intensity GHG emissions) for a specific timeframe and the
projected global warming in 2100. This assumption is applied for corporate targets, for target
horizons set out in the Section 4.2. Limitations to this assumption of linearity are covered in Section
8.1.1.
The first step in assigning temperature scores to disclosed corporate GHG targets is to assess
which types of targets could be adequately matched to a scenario variable, and consequently,
which associated linear regression model should be applied. Disclosed corporate GHG targets
refer to either absolute GHG reduction targets and/or GHG intensity reduction targets. Table 2 in
Section 4.2.1 shows the sector variables and the associated linear regression model applicable in
this methodology. Table 4 showcases examples of climate target wording applicable for the
respective variables.
Table 4: Target class, wording, and scenario variables
Target Class
Example of target wording
AR6 benchmark variable
Absolute GHG reduction
targets
Company X commits to reduce
absolute scope 1 GHG emissions
60% by 2030 from a 2022 base
year.
Company X commits to reduce
absolute scope 2 GHG emissions
60% by 2030 from a 2022 base
year.
Company X commits to reduce
absolute scope 3 emissions GHG
with 50% by 2030 from a 2022
base year.
• Emissions | Kyoto Gases
• Emissions | CO
2
| Energy | Supply
• Emissions | CO
2
| Energy and
Industrial Processes
GHG economic intensity
target
Company X commits to reduce
scope 1 GHG emissions 60% per
unit of added value by 2030 from
a 2022 base year.
Company X commits to reduce
scope 2 GHG emissions 60% per
unit of added value by 2030 from
a 2022 base year.
Company X commits to reduce
scope 3 GHG emissions 50% per
unit of added value by 2030 from
a 2022 base year.
• Emissions | Kyoto Gases / GDP |
PPP
• Emissions | CO
2
| Energy | Supply /
Secondary energy
23
© CDP Worldwide and WWF International 2024
The second step is to convert the corporate target into a corresponding annual reduction rate to match the format of the
independent variable of the regression models. These annualized reduction rates are calculated using the formula for
CAR (see Equation 2 in Section 4.2)4.2. As per
Equation 3, the CAR of a target is used in the linear equation specified by the model’s parameters
to convert the target’s ambition into a temperature score. For example, an absolute GHG reduction
target of 30% between the base year 2020 and the target year 2035, mapped to the scenario
variable Emissions | Kyoto Gases, would result in a compound annual reduction rate of 2.3%.
Applying the Emissions | Kyoto Gases linear equation for the long-term timeframe (corresponding
to the 2035 target year), this target’s ambition would translate to a 2.1°C temperature score
(calculated as 2.81 − 0.30 * 2.3).
5.2. Target validation
All targets are subjected to a validation procedure to assure their usefulness in generating
company-level temperature scores. The method also attempts to split targets covering multiple
scopes into single-scope targets. The target validation process checks that each target is
accompanied by the data required to calculate temperature scores in later steps.
The following criteria need to be met for a target to be valid:
1) The scope coverage of the target, any single scope or combination of scopes 1, 2, or 3.
2) Target type must be absolute or intensity with valid metric.
3) Target progress < 100% on the date the relevant target was first published.
4) Base year < target year.
5) Target year ≥ current year.
6) Base year GHG data must be available for the emission scope of the target, i.e., scope
1 GHG data for a scope 1 target, etc.
7) Boundary coverage of the target is required for the emission scope(s) of the target, for
instance, 60% of the scope 1 emissions. If this number is missing, the value will be set
to zero.
8) Target reduction ambition must not be negative.
Targets which do not meet the criteria are removed from further calculation.
Temperature scores are calculated on the most disaggregated level of targets that the provided
data allows. A scope 1+2+3 target is split into one scope 1+2 target and one scope 3 target, and
a scope 1+2 target (including those split from a scope 1+2+3 target) will be split into one scope 1
target and one scope 2 target. Targets for the three scopes will then be scored separately and a
combined score will later be calculated in the target aggregation procedure, see Section 6.5. It
should be noted, however, that temperature scores are aggregated using the company’s current
GHG emissions. Therefore, the presence of current GHG data must be verified before a scope
1+2 target is split. If the data is missing, the scope 1+2 target is kept and scored using the provided
target data.
When splitting a valid scope 1+2 target, the scope 1 target is assigned the scope 1 boundary
coverage from the original target and the scope 2 target is assigned the scope 2 boundary
24
© CDP Worldwide and WWF International 2024
coverage from the original target. Both targets are assigned the reduction ambition of the original
target.
The procedure for splitting targets thus means that a target covering scopes 1+2+3 ideally results
in three targets for the individual scopes, where each target consists of its respective boundary
coverage and the reduction ambition from the original target. However, even if current GHG data
is missing, a scope 1+2+3 target will be split into a scope 1+2 target and a scope 3 target for
separate scoring. For the scope 1+2 target the boundary coverage is calculated as follows:
Equation 4: Boundary coverage
where bc = boundary coverage
The reduction ambition is copied from the original scope 1+2+3 target. The resulting scope 3 target
keeps the scope 3 boundary coverage and the target reduction ambition from the original target.
Note that if reduction ambition is missing or has a value of zero, the TS will be calculated as the
intercept of the applicable regression function.
Finally, the boundary coverage is used to adjust each target’s reduction ambition according to the
procedure described in Section 6.3.2.
5.3. Default temperature score for companies without valid targets
5.3.1. Purpose of a default temperature score
The purpose of assigning a default temperature score is to provide a means for scoring a full
portfolio or value chain although some portfolio constituents lack publicly disclosed targets or fail
to meet specific criteria for target coverage or quality.
In instances where companies do not have valid targets, it is assumed that they are following a
business-as-usual trajectory, as they have not publicly articulated their GHG emissions reduction
strategies through GHG targets. Therefore, default scores represent the anticipated business-as-
usual GHG emissions trajectory. In essence, the default score reflects the pathway expected to
be followed if companies continue operating under existing governmental policies, and thus
adhering to the minimum requirements of current regulation.
5.3.2. Default score approaches
Business-as-usual trajectories can be defined at a company, sector, and/or economy-wide level.
This version 1.5 of the methodology still focuses on uniform default scores at an economy-wide
level but will aim to provide more sectoral granularity in version 2.0 (see Section 8.2). While
economy-wide default scores assume the company’s temperature score is aligned with that of the
global economy, sector-specific approaches define business as usual pathways at a sector level
and assume the company’s trajectory is consistent with that of the sector.
25
© CDP Worldwide and WWF International 2024
5.3.2.1. Economy-wide default scores
An economy-wide default score applies the score uniformly to all companies, regardless of sector
or current performance.
The first version of this methodology relied on the December 2019 end-of-century warming
projections from the Climate Action Tracker (“CAT”) to establish a 3.2°C economy-wide default
score. At the time, this value corresponded to the upper bound of the range of median
temperatures expected from the continuation of current policies being implemented by
governments (i.e., real-world action based on current policies).
Using the same source (updated in December 2023, see Figure 3) and the same “policies & action”
projections, a range of warming between 2.2°C and 3.4°C is expected by the end of the century,
with a median projection of 2.7°C (50% probability).
26
© CDP Worldwide and WWF International 2024
Figure 3: Summary and breakdown of 2100 warming projections based on a range of future scenarios (Source: Climate
Action Tracker, December 2023)
This aligns with the conclusions of the 2023 UNEP Emissions Gap Report (United Nations
Environment Programme, 2023), which finds that a continuation of the level of mitigation effort
under current governmental policies would result in a warming of 2.7°C at the end of the century
(range: 1.8–3.5°C, with a 50% probability).
27
© CDP Worldwide and WWF International 2024
Table 5: End-of-century warming projections based on a range of future scenarios (UNEP Emissions Gap Report, 2023)
When presenting temperature estimates with a 66% probability, CAT’s median projections of
3.0°C also align with the UNEP Emissions Gap Report’s warming estimates.
This updated version of the methodology uses a 3.4°C value (i.e., the upper bound of the range
of temperature outcomes from CAT’s “policies & action” projections) to derive temperature scores
for companies with no valid forward-looking targets. This implies that these companies are
expected to decarbonize along a 3.4°C pathway, consistent with global policies implemented to
ensure the reduction of GHG emissions at this rate.
While it could be argued that the ambition of current policies has improved somewhat over recent
years (as reflected in the UNEP Emissions Gap Report 2023’s projections, for instance), their
impact at the company level remains uncertain. Bearing in mind the purpose and objectives of the
default temperature score outlined in Section 5.3.1, the shift from 3.2°C to 3.4°C in this version
aims primarily at reinforcing the continued need for a conservative approach when considering
companies without valid targets.
Chapter 8 outlines the plan for future methodological development, where we aim to provide more
granular and sector specific default scores for companies with no valid GHG emission reduction
targets.
5.4. Temperature floor
The best score applicable to a company’s climate ambition under this methodology is 1.5°C (i.e.,
in the case where a calculated score results in a temperature that is lower than 1.5°C, the applied
score shall be 1.5°C).
The reasoning behind flooring all temperature scores to 1.5°C is grounded in the prevailing
scientific consensus
20
which, at the time of drafting this version of the methodology, remains that
1.5°C represents the lower bound of feasible outcomes within the most optimistic climate
scenarios. In addition, this aligns with the conclusions of the Climate Action Tracker’s latest
20
Intergovernmental Panel on Climate Change (2021) – AR6 Synthesis Report: Climate Change 2023
29
© CDP Worldwide and WWF International 2024
6. Step 2b: Company scoring
Companies set targets across different timeframes, emissions scopes, and units. Target data is
also often collected from several sources and a target database may include multiple targets for
the same company, scope, and timeframe. This section describes how the method selects and
prioritizes which relevant valid targets to be scored.
6.1. Target timeframe
The timeframe sorting first defines the range of target timeframes as applied in this methodology.
Targets are scored depending on which year the target ends in relation to the current year. The
target timeframes are divided into the following three buckets:
- Short-term: targets ending in up to 5 years from the current year, e.g., −2029 (current
year 2024)
- Mid-term: target ending 5–10 years from the current year, e.g., 2030–2034 (current year
2024)
- Long-term: target ending in more than 10 years from the current year, e.g., 2035–2050
(current year 2024)
Targets can then be scored across these three different timeframes, providing insights on the
short-, medium-, and long-term ambition of companies’ GHG emissions reduction targets.
Target timeframe also defines how long a target is used for scoring in a target timeframe bucket,
i.e., when a mid-term target becomes a short-term target and for how long a target is considered
as valid for scoring by this methodology. For instance, targets with end dates during the calendar
year (CY) 2024 would be valid throughout 2024 and would become invalid on January 1, 2025. A
mid-term target with an end date during CY 2030 would become a short-term target on January
1, 2026, as mid-term targets are defined as targets with 5–10 years left to run. There are
exceptions to this rule set out in Section 6.3.6.1 below.
Figure 4 displays a summary of the company protocol steps, including the waterfall, leading to the
output of a matrix of temperature scores for each timeframe and scope combination.
30
© CDP Worldwide and WWF International 2024
Figure 4: Step 2b – generating temperature scores at a company level, based on valid publicly disclosed targets or a
default approach for companies with no valid targets.
6.2. Target quality criteria
Targets are classified in terms of seven key criteria, presented in Table 6. Key requirements for
valid targets across these seven key criteria are further detailed in the different paragraphs of this
31
© CDP Worldwide and WWF International 2024
section. A description of how these criteria are prioritized in the selection process is provided in
Section 6.3.
Table 6: Target quality criteria
Criteria
Description
Target vintage
Defines the age of the target, based on the date the target was last
publicly communicated.
Boundary coverage
Within a given GHG emissions scope, companies define how much of
that scope will be included in the boundary of the target, e.g., 50% of
scope 1 or 95% of combined scope 1+2 is covered by the target.
Target Type
Defines whether the target ambition is based on an absolute or
intensity GHG emissions reduction.
Target scope
coverage
Defines the GHG emissions scope(s) covered by the target. Targets
can be set across individual or combined GHG emission scopes, as
defined in the GHG Protocol (World Business Council for Sustainable
Development & World Resources Institute, 2015), e.g., scope 1,
scope 2, scope 3, scope 1+2, scope 1+2+3, etc.
Reduction ambition
Defines the GHG emissions reduction ambition over the target
duration, e.g., absolute emission reduction of 30% by 2030.
Target timeframe
Classifies targets according to the duration between the current year
and the target end date. Targets can cover different timeframes.
Target progress
Describes the rate of achievement at the time the target was first
published. To be scored, target achievement has to be less than
100% at the time the target is published.
6.3. Target criteria waterfall
As mentioned in the introduction to this and displayed in Figure 4, a company’s targets are sorted
by scope coverage and according to three timeframes.
Some target quality criteria may be conflicting. Therefore, the method has established a waterfall
for these criteria, to determine which criteria should take precedence and be used for scoring.
For instance, companies may be reporting multiple targets within the same scope and timeframe,
e.g., two mid-term targets covering scope 1+2, covering different parts of their operation. Further,
users of this method may use different databases to collect target and emissions data which may
be updated with different frequencies, resulting in a combined user database with multiple targets
that may be conflicting. The below waterfall approach is used to select a single target for each
timeframe and scope category for scoring.
32
© CDP Worldwide and WWF International 2024
Table 7: Target waterfall criteria
Rank
Criteria
Priority
1
Vintage
More recently published targets
2
Boundary coverage
Highest coverage
3
Type
1) Absolute
2) Intensity
3) Other
4
Scope coverage
Prefer single scope S1, S2, and S3 targets before combined
scope targets that cannot be broken down into single scope
components.
5
Reduction ambition
Higher reduction ambition preferred
6
Timeframe
Longer-term targets are preferred within each timeframe
bucket (i.e., short-, medium-, long-term). If there are several
targets with same target year the more recent base year is
preferred.
6.3.1. Target vintage
A company’s latest statement of emission reduction ambition is preferred to older communicated
targets. Thus, a target to reduce GHG emissions by 30% by 2030 that was communicated last
month will take priority and be scored, instead of a target of 40% GHG emissions reduction by
2032 communicated last year, ceteris paribus. If the latest statement date is not available, the
method uses the target start year.
6.3.2. Boundary coverage
How much of companies’ emissions are covered in the GHG emissions reduction target, i.e., the
boundary coverage, often differ between companies and targets. Therefore, the ambition of the
target is normalized with the boundary coverage of the target. The boundary coverage of a target
can be expressed in percentage terms (e.g., 80% of the company’s scope 1 emissions in the base
year) or in emissions terms (e.g., 800,000 tons of scope 1 emissions out of a total 1,000,000 tons
of scope 1 emissions in the base year, i.e., 80%).
For combined GHG emissions scope targets, e.g., targets covering more than one scope, the
boundary coverage of the target is defined as the weighted average of the boundary coverage of
each of the scopes included in the target, using base year GHG as weights.
For all targets with less than 100% boundary coverage, the ambition of the target is adjusted by
the boundary coverage percentage. For instance, consider an absolute target of 30% reduction in
scope 1 and 2 GHG emissions, but that this target only covers 20% of the company’s scope 1 and
2 emissions. The ambition would then be adjusted to 6% (30% × 20% = 6%).
Without the quantified boundary coverage data, the method assumes 0% boundary coverage.
Given the adjustment to ambition is done based on the level of boundary coverage, as explained
33
© CDP Worldwide and WWF International 2024
above, any reduction ambition without boundary coverage data will be reduced to 0%. The
temperature score will then become the intercept of the regression.
6.3.3. Target types
Only GHG emission reduction targets are currently acceptable for scoring, i.e., absolute and
intensity GHG targets. Other targets, such as procurement, renewable electricity targets, or
engagement are currently not scored (see exception in Section 6.3.3.1). Long-term ambitious or
aspirational targets that are not quantitative (e.g., climate neutral/net-zero in 2050) are not scored
currently as these types of targets are difficult to objectively translate to a rate of change.
GHG emission reduction target types can broadly be divided into absolute and intensity targets.
All types of absolute targets based on GHG emissions and intensity targets based on GHG
reductions per unit of X are valid for scoring in the method, such as:
- Physical intensity targets: based on GHG emissions per unit of production – e.g. CO2eq /
kWh for Power Generation
- Economic intensity targets: based on GEVA (GHG emissions per unit of value added) or
revenue.
- Intensity targets where the conversion to absolute GHG emissions is disclosed.
6.3.3.1. Target type exception
The exception to this rule is scope 3 targets set using the CDP–WWF Temperature Rating method
version 1.0 or CDP–WWF Temperature Scoring method version 1.5 (this method).
The CDP–WWF methodology is an engagement method. Financial institutions (and corporates in
other sectors) can use this methodology to assess the temperature score of their portfolio (or value
chain) based on the portfolio (value chain) constituents’ current ambition. Users of this
methodology can use it to derive a temperature trajectory for their scope 3 emissions. It is the
underlying constituent of the portfolio or value chain that needs to reduce emissions and adapt its
ambition accordingly. To give constituents time to implement emissions reduction action, the
method needs to allow for implementation time. A common mid-term target setting period is 10
years, which for instance is used by the SBTi in its target setting framework (Science Based
Targets initiative, 2024). So, to allow for all constituents in a portfolio or value chain to complete
their emissions reduction action before 2050, the target must have been set and communicated
before this, i.e., by 2040 at the latest, to allow for 10 years of implementation. Thus, for a financial
institution, or other corporate, to be able to reach 1.5°C by 2050, all the constituents in the portfolio
or value chain must have set their 1.5°C-targets no later than 2040.
Therefore, financial institutions’ scope 3 targets set with the CDP–WWF method are scored by
linear extrapolation of the targeted TS reduction ambition until 2040, using Equation 5.
For instance, consider financial institution A with a target to reduce its portfolio’s TS to 2.5°C by
2030 from a 2025 base year TS of 3.0°C. This would equate to a TS of 1.5°C by 2040 [3.0 − (2040
− 2025) × ((3.0 − 2.5) / (2030 − 2025) = 1.5]. Thus, financial institution A’s scope 3 target set using
34
© CDP Worldwide and WWF International 2024
this method would then get a TS of 1.5°C. To better understand the practical implications of this
target type exception, assume that FI A’s scope 1+2+3 TS is also 1.5°C, and consider FI B whose
portfolio is entirely composed of FI A stock. Applying this target type exception, FI B’s scope 3 TS
would then be 1.5°C. In another example, consider FI C with a target to reduce its portfolio TS to
2.5°C by 2029 from a 3.0°C base in 2022. This would equate to a TS of 1.71°C [3.0 − (2040 –
2022) × ((3.0 − 2.5) / (2029 – 2022)], when scored as a constituent in another FI’s portfolio.
Equation 5: Scoring of targets using CDP–WWF Temperature rating method v 1.0 and CDP–WWF Temperature Scoring
method v1.5
To clarify, absolute and intensity targets take precedence over the CDP–WWF-method-based
targets in the waterfall.
6.3.4. Scope coverage
Single scope targets covering only scope 1, 2, or 3 emissions are assessed and scored separately.
Targets covering several scopes (combined scope targets), are first disaggregated to compute
single score targets, as described in Section 5.2, and later aggregated into a company level
temperature score. If data allows (see Annex 5: Summary of required data for applying the
Temperature Scoring method), single scope and combined scope targets are also scored as
combined and single scopes respectively.
6.3.5. Reduction ambition
More ambitious targets are preferred, as measured by CAR. Thus, a target A) to reduce GHG
emissions by 50% by 2030, from a 2020 base year, with a CAR of 6.7% [(1 − 50 / 100) ^ (1 / (2030
– 2020)) − 1] will be preferred before a target B) with GHG emissions reduction of 60% by 2034
from a 2020 base year, giving a CAR of 6.4%. Therefore, target A will be used for TS calculation,
ceteris paribus. In cases where reduction ambition is missing or is equal to zero, the value will be
set to zero which implies that the resulting temperature score will be equal to the intercept of the
applicable regression model.
6.3.6. Target year
Within each timeframe bucket, targets with a later target year are preferred as these are more
forward-looking. If the target years are the same, the more recent base year is preferred.
6.3.6.1. Target timeframe exception
Under the Financial Sector Science-Based Targets Guidance (Science Based Targets initiative,
2022), engagement targets are set for a maximum of five years. This currently includes targets
set with this CDP–WWF Temperature Rating method. This would mean that targets set using this
method would be treated as short-term targets in the CDP–WWF Temperature Scoring method,
which in many FI portfolios would make it impossible to use this method for setting targets under
the SBTi framework (Science Based Targets initiative, 2022).
35
© CDP Worldwide and WWF International 2024
Therefore, when financial institutions have assets in their portfolios that have set targets with this
method it prohibits, e.g., an FI from setting a meaningful medium target themselves, as the
portfolio constituents don’t have valid mid-term targets, according to the target validation laid out
in Chapter 5.
However, as this method relies on other companies setting or improving their targets, as explained
in Section 6.3.3.1, the engagement targets will naturally take some time to deliver GHG emissions
reduction. Therefore, an engagement target of up to five years could then be seen to be similar in
timeframe to a medium-term non-engagement target.
Therefore, the CDP–WWF Temperature Scoring method adds five years to all targets based on
SBTi engagement methods to enable this method to treat them as medium-term targets and
therefore allow companies with assets and activities with engagement targets in their portfolios to
set relevant medium-term targets.
Currently, the CDP–WWF Temperature Scoring method only makes this exception for this
engagement method.
6.4. Multiple scope 3 targets
Some companies publish multiple targets for scope 3 GHG emissions. This can take the shape of
an overall scope 3 target alongside separate targets for certain scope 3 categories, e.g., where
the company has significant GHG emissions, or the company wants to set a specific target for a
specific category.
There are also cases where there are multiple targets for a single scope 3 category. This is more
common for scope 3 category 15, investments. Targets by FIs validated by the SBTi often use this
approach, sometimes without a headline scope 3 target that includes all S3 targets. Often
boundary coverage for the individual categories is not published, nor is current or baseline GHG
emissions for the categories. Scope 3 category 15 targets are also often engagement targets,
instead of based on GHG emissions. This creates several issues for the company scoring.
- It becomes difficult to use the target criteria waterfall to select a target for scoring, as there
are several targets of the same vintage for the same scope and even for the same
category.
- As boundary coverage and GHG data often is not available it becomes difficult to weight
multiple scope 3 targets to one headline scope 3 target.
- Engagement targets are often stated as aligning portfolio coverage or temperature score
of a part of a portfolio to a certain coverage or temperature score targets. This is very
different compared to the GHG emissions reduction targets mostly used for scope 1 and 2
targets, both absolute and intensity targets.
Therefore, when companies set multiple targets for scope 3 targets, the method treats these
targets differently, based on the following principles:
36
© CDP Worldwide and WWF International 2024
1. Any target should be assessed with the current temperature scoring method, as if it was a
single scope 3 target.
2. If a headline scope 3 target is available, this will take priority and be scored as a single
scope 3 target, provided it can be scored under the current method.
3. If a headline target is not available or it cannot be scored and multiple scope 3 targets are
available, these targets will be scored and aggregated to a single companywide scope 3
target.
4. The selection of scope 3 targets to be scored follows the target criteria waterfall, but allows
for several targets in scope 3 and in each scope 3 category. This is common in scope 3
category 15 targets for investments, where parts of an investment portfolio may be
addressed with different targets. In these cases the waterfall should be applied to the
individual asset class and activity targets, within the scope 3 category, before aggregation.
All scope 3 targets must be of the same vintage to be scored as separate scope 3 targets.
Older scope 3 targets will be excluded from scope 3 scoring, when aggregating several
scope 3 TS to a headline TS.
5. This means that for instance for a target set with the sectoral decarbonization approach
(SDA) (Science Based Targets initiative, 2022), the target will be assessed based on the
reduction ambition of the target and mapped to the scenarios that the Temperature Scoring
methodology uses for the particular industry, as laid out in Table 4 in Section 5.1.
6. Engagement type targets such as portfolio coverage and supplier engagement cannot be
scored unless these targets can be converted to GHG emissions reduction ambition. If no
such conversion is possible, these targets will be given default scores.
7. Temperature score targets using the CDP–WWF temperature score method will inherit its
current or most recently calculated temperature score.
a. If underlying assets for this temperature score is available, these assets are
reprocessed and be given an up-to-date temperature score, as described in
Section 6.4.1.
6.4.1. Aggregation of multiple scope 3 temperature scores
As GHG emissions data for the individual scope 3 targets are often not available, the method
cannot apply the same approach as described in target aggregation, where share of GHG
emissions is used as weights. Therefore, the method also allows for an equal weight average to
aggregate multiple scope 3 targets:
- GHG emissions, or
- Equal weight average.
This means that if GHG data is available, the method weights multiple scope 3 targets using
Equation 6:
37
© CDP Worldwide and WWF International 2024
Equation 6: Aggregation of multiple S3 TS
and,
Note that current GHG emissions data must be present for all scope 3 categories for which there
are targets, if weighting of TS with GHG is to be performed. If GHG data is lacking, the weighting
will be done using the equal weight mean of TS for the timeframe in question.
6.5. Temperature score aggregation
The method accepts targets that cover either a single scope or the combinations of scopes 1+2,
scopes 1+3, scopes 2+3, and scopes 1+2+3. Temperature scores are calculated for each
validated target. The temperature scores are then aggregated using the company’s current GHG
data into a combined score for each scope and scope combination and time frame.
To combine single scope TS to combined scope TS, the single scope TS are aggregated by the
scope’s weighting in the company’s GHG profile, for instance:
where:
S = scope
S123 = scope 1, 2, and 3
TS = temperature score
GHG = greenhouse gas emissions in the current year
Combined TS, e.g., scope 1 and 2, can also be further aggregated into a full combined scope 1,
2, and 3 TS, as in this example:
Equation 7: Temperature score aggregation
where:
S12 = scope 1 and 2
6.6. Using temperature scores
Depending on the option chosen for timeframe coverage, up to 15 temperature scores can be
calculated per company based on target timeframe and scope coverage.
38
© CDP Worldwide and WWF International 2024
The mid-term timeframe is considered the key timeframe as it currently represents the main period
for corporate ambition and aligns with the SBTi’s target setting criteria of 5–10 years from the
reporting year. The short and long-term scores can be used to better understand if companies
have more immediate and longer-term goals in place.
See Annex 4: Calculation examples, which illustrates how these scores are calculated.
39
© CDP Worldwide and WWF International 2024
7. Step 3: Portfolio scoring
The final step of the temperature scoring method describes the portfolio scoring step, including
the different weighting options for aggregating the temperature scores of companies at an index
or portfolio level.
Portfolio scores are calculated by aggregating TS of the same timeframe. Several weighting
options are provided, that may be used in different applications.
7.1. Weighting objectives and principles
Before developing weighting approaches, a set of objectives were first developed to help evaluate
proposed weighting options (Table 8).
Table 8: Default weighting method objectives
In addition to meeting these objectives, the default weighting method should best adhere to a set
of weighting principles, presented in Table 9.
Table 9: Default weighting principles
Objective
Description
Enable Net-zero /
Paris alignment
The method should emphasize climate impact and support
investors in accurately assessing the °C temperature score of
an index or a portfolio and in aligning their investments with a
1.5°C pathway.
Support better disclosure
of GHG emissions by
corporations
The method should foster more and higher quality disclosure
of GHG emissions along the entire value chain (scope
1+2+3) by global corporations.
Support standardization of
methods
The method should be aligned with existing portfolio GHG
accounting methods.
Principle
Description
Comparability
Results should be comparable across different asset classes,
where applicable, and investment products.
Applicability
Investors should be able to perform the aggregation at a
reasonable cost with public / accessible data.
Reliability
The weighting method should produce results which are reliable
and verifiable.
Clarity
The weighting method should be understandable and practical to
implement.
40
© CDP Worldwide and WWF International 2024
7.2. Weighting options
Seven potential options for aggregating individual company temperature scores at the index /
portfolio are currently included in the method. These include:
- Option 1: Weighted average temperature score (WATS)
- Option 2: Total emissions
21
weighted temperature score (TETS)
- Option 3: Market Owned
22
emissions weighted temperature score (MOTS)
- Option 4: Enterprise Owned
23
emissions weighted temperature score (EOTS)
- Option 5: EV + Cash emissions weighted temperature score (ECOTS)
- Option 6: Total Assets emissions weighted temperature score (AOTS)
- Option 7: Revenue owned emissions weighted temperature score (ROTS)
Table 10 provides a description and formula for calculating the portfolio temperature scores using
each of these options.
21
The total of a company’s scope 1, 2, and 3 reported and modelled GHG emissions of the latest reporting
period.
22
Based on a company’s market capitalisation, i.e., the total euro market value of a company’s outstanding
shares of stock. Commonly referred to as “market cap”, it is calculated by multiplying the total number of a
company’s outstanding shares by the current market price of one share.
23
Based on Enterprise value (EV). EV is a measure of a company’s total value and includes in its calculation
the market capitalisation of a company but also short-term and long-term debt.
Timeliness
The weighting method should produce results that are timely and
current.
Completeness
The weighting method should allow for complete portfolio
assessments.
41
© CDP Worldwide and WWF International 2024
Table 10: Details of portfolio aggregation methods
Weighting Option
Method
Temperature score formula
(where TS = Company temperature score)
Weighted average
temperature score
(WATS)
Temperature scores are allocated based
on portfolio weights
Total emissions weighted
temperature score (TETS)
Temperature scores are allocated based
on historical emission weights using total
company GHG emissions
Market Owned emissions
weighted temperature
score (MOTS)
Temperature scores are allocated based
on an equity ownership approach
Enterprise Owned
emissions weighted
temperature score
(EOTS)
Temperature scores are allocated based
on an enterprise ownership approach
Enterprise Value + Cash
emissions weighted
temperature score
(ECOTS)
Temperature scores are allocated based
on an enterprise value (EV) plus cash &
equivalents ownership approach
Total Assets emissions
weighted temperature
score (AOTS)
Temperature scores are allocated based
on a total assets ownership approach
42
© CDP Worldwide and WWF International 2024
Revenue owned
emissions weighted
temperature score
(ROTS)
Temperature scores are allocated based
on the share of revenue
The denominators in the formulas presented in Table 10 are defined as follows:
TETS: Portfolio GHG emissions are the sum of the portfolio company GHG emissions.
MOTS: Portfolio market value owned GHG emissions is the sum of portfolio company owned GHG
emissions weighted on the market cap of investee companies.
EOTS: Total enterprise owned GHG emissions is the sum of portfolio company owned GHG
emissions weighted on the enterprise value of investee companies.
ECOTS: Total EV + Cash owned GHG emissions is the sum of portfolio company owned GHG
emissions weighted on the enterprise value + cash of investee companies.
AOTS: Total Assets owned GHG emissions is the sum of portfolio company owned GHG
emissions weighted on the total assets of investee companies.
ROTS: Revenue owned GHG emissions is the sum of portfolio company owned GHG emissions
weighted on the share of revenue of investee companies.
43
© CDP Worldwide and WWF International 2024
7.3. Weighting method assessment
The analysis presented in this section is the result of assessments developed for version 1.0 of
this methodology (CDP–WWF Temperature Rating Methodology). In-depth revision of weighting
options was not part of the update to version 1.5. Future versions may consider providing more
guidance on weighting options and related analysis. Therefore, this section is unchanged from
version 1.0.
In Table 11, each proposed weighting method is compared against the objectives outlined in
Section 7.1 (Table 8).
44
© CDP Worldwide and WWF International 2024
Table 11: Assessment of options against weighting objectives
Table 12 provides an assessment of each option against the principles outlined above.
24
TCFD (Task Force on Climate-related Financial Disclosures, 2017): Implementing the Recommendations
of the Task Force on Climate-related Financial Disclosures
25
PCAF (Partnership for Carbon Accounting Financials, 2019): Accounting GHG emissions
and taking action: harmonised approach for the financial sector in the Netherlands
Objective
WATS
TETS
MOTS
EOTS
ECOTS
AOTS
ROTS
Comment
Enable Net-zero
/ Paris
alignment
✓
✓✓✓
✓✓
✓✓✓
✓✓✓
✓✓✓
✓✓✓
Exposure to high impact companies is best
reflected under TETS; exposure under the
ownership methods could be masked by high
market cap / EV / revenue, etc., of these
companies.
Support better
disclosure of
GHG emissions
by corporations
✓
✓✓✓
✓✓✓
✓✓✓
✓✓✓
✓✓✓
✓✓✓
WATS does not take current GHG emissions into
account, therefore the incentive for companies to
report is lower.
Support
standardisation
of methods
✓✓✓
✓
✓✓✓
✓✓
✓✓✓
✓✓
✓✓
WATS aligned to TCFD’s
24
main recommended
WACI method for measuring the carbon intensity
of a portfolio. MOTS aligned to TCFD’s approach
for carbon footprinting. ECOTS aligned to
PCAF
25
method for carbon footprinting of listed
equities and corporate debt.
45
© CDP Worldwide and WWF International 2024
Table 12: Assessment of options against weighting principles
The AOTS and ROTS methods best support the stated objectives whereas WATS is the least
supportive method. In contrast, WATS is better aligned to the principles compared to the
ownership approaches. Yet, some of the related disadvantages of EOTS/ECOTS/AOTS/ROTS
would be less significant with better corporate reporting of GHG emission inventories.
Objective
WATS
TETS
MOTS
EOTS
ECOTS
AOTS
ROTS
Comment
Comparability
✓✓✓
✓✓✓
✓
✓✓
✓✓
✓✓✓
✓✓✓
MOTS cannot be applied to corporate bonds. EOTS
and ECOTS are not always meaningful as e.g. EV is
not widely used for banks.
Applicability
✓✓✓
✓✓
✓✓
✓✓
✓✓
✓✓
✓✓
TETS requires GHG data, the ownership methods
require GHG and additional corporate financial data.
Specific corporate financial data may be difficult to
obtain for non-listed companies.
Reliability
✓✓✓
✓✓
✓✓
✓✓
✓✓
✓✓
✓✓
All options besides WATS are based on self-reported
and modelled GHG data.
Clarity
✓✓✓
✓✓✓
✓✓
✓✓
✓✓
✓✓
✓✓
Ownership based methods reduces transparency /
results are somewhat less intuitive.
Timeliness
✓✓✓
✓✓
✓✓
✓✓
✓✓
✓✓
✓✓
All options besides WATs are dependent on timely
GHG data.
Completeness
✓✓✓
✓✓
✓
✓
✓
✓
✓
TETS dependent on GHG data for all portfolio
companies; The ownership approaches (MOTS,
EOTS and ECOTS) require additional corporate
financial data.
46
© CDP Worldwide and WWF International 2024
7.4. Additional notes on the portfolio scoring
Double counting: Potential double counting of GHG emissions and their respective targets
when weighting and aggregating temperature scores should not impact the validity of this
metric. A TS is reflective of the climate target ambition of a company and/or a portfolio. This
methodology provides temperature scores at the most disaggregated, single-scope level over
different target timeframes, allowing flexibility for users to consider all scopes of a given
company or portfolio as relevant. Yet, the company-level score is representative of a company’s
total ambition, across all scopes.
47
© CDP Worldwide and WWF International 2024
8. Limitations and outlook
While temperature scores computed according to this methodology provide a robust assessment
of a company’s GHG emissions reduction ambition, they also come with inherent limitations. The
following section will elaborate on this methodology’s main limitations.
8.1. Methodology limitations
8.1.1. Linear regression model
The use of a linear regression model to evaluate the relationship between annual reduction rates
of GHG emissions and projected temperature outcomes is a simplification choice that has inherent
limitations.
First, the AR6 scenarios that form the basis of the linear regression models are different in their
respective underlying assumptions, and the purpose of the scenarios is not necessarily to combine
them in one linear regression, aiming to explain such as relationship. It is expected that this
general limitation will be true for any approach that uses several scenarios to evaluate GHG
targets, the reason being that benchmarks must either be based on a single scenario or some
statistical averaging of scenario results.
Second, a linear regression model is applied to AR6 scenarios’ variables that, as illustrated in
Figure 5, do not follow a linear pathway. To capture part of that curvature, the regression models
are applied to various time horizons. For short term horizons, however, the model fit is lower
because a wider range of annual reduction rates could lead to similar end-of-century temperature
outcomes.
48
© CDP Worldwide and WWF International 2024
Figure 5: Example of GHG and CO2 emissions pathways for a subset of scenarios, for illustrative purposes only (Source:
Figure 3.10 IPCC AR6 Chapter 3)
Finally, other models – e.g., an exponential decay model – might provide a better fit to the CAR,
as observed in Figure 2. The trade-off is an additional complexity in the application of the
methodology and coefficients that can be less intuitive to interpret (see Box 1: from LAR to CAR
in Section 4.2).
8.1.2. Formula to compute annual reduction rates
The calculation of annual reduction rates is a central part of this methodology. However, it has
inherent limitations.
There are different definitions of annual rate of change, and this choice has implications on the
model. Both linear annual reduction (LAR) and compounded annual reduction (CAR) present their
own advantages and disadvantages. The LAR has a lower variance compared to the CAR,
49
© CDP Worldwide and WWF International 2024
especially for long time horizons, which improves the fit of the linear model. As an example, 99%
over 30 years and 90% over 30 years have a respective LAR of 3.3% and 3.0%. For similar
percentage changes, the respective CAR are −14.2% and −7.4%. Additionally, the LAR can
handle targets of −100% which is not feasible with the CAR.
On the other hand, the linearity implied by the LAR is in contradiction with the scientific
understanding that early action (reducing emissions in the next several years) is crucial to achieve
1.5°C with limited overshoot. A constant annual reduction in absolute tons of GHG emissions
suggests that the efforts should be spread equally over the years. In relative terms, for LAR, the
reductions in early years would represent a much smaller share of current year emissions than at
the end of the target period. On the other hand, the CAR implies a constant share of reductions
out of the current year’s emissions, but a larger initial reduction in absolute terms in the early years
of the target period. While the LAR and the CAR should both lead to the same GHG emissions
level at the end of the target period, the impact on cumulated carbon budget can be significantly
different. It is worth noting that neither of these two applications can accurately capture the carbon
budget associated with the different scenarios.
Alternative solutions to address the challenges of assessing 100% and near 100% reduction
targets with the CAR need to be explored. It is worth noting, however, that this specific case will
occur only when companies plan to reduce 100% of emissions and covering 100% boundary.
While the model currently does not treat gross and net targets separately (as per Section 8.1.6),
the plausibility of a case where companies can achieve zero emissions should be further
discussed. In the AR6 database of scenarios, there are 1,115 scenarios with Kyoto gases data
available. Out of those 1,115 scenarios, three scenarios imply that GHG emissions (Kyoto gases
variable) reach zero tons or less in 2050.
8.1.3. Sector granularity
Currently, most companies’ target ambition is assessed against a cross-sectoral benchmark
(except for power generation and minor exceptions for cement/steel/aluminum companies, see
Section 4.2.1). Likewise, the applied default score of 3.4°C to companies with no or no valid target
data is applied consistently, without differentiating by a company’s sector affiliation.
Developing warming functions for sectors requires a minimum of available sectoral climate
scenarios. Currently, the AR6 Scenarios Database does not provide sufficient sector-specific
variables that fit a linear regression model, nor are they designed with the main aim to project a
sectoral GHG trajectory useful for benchmarking. Whether the inclusion of scenarios with more
sector granularity provided by other institutions (e.g., International Energy Agency) will result in
enough input data for constructing sectoral warming functions has to be investigated.
The choice of applying a uniform default score is an interim solution to enable the generation of
portfolio-level temperature scores by also weighting companies that do not have valid, forward-
looking targets. However, it can be argued that companies with an already low-emission profile
are unfairly scored with a 3.4°C TS, even in the absence of any climate target, unjustifiably
50
© CDP Worldwide and WWF International 2024
worsening the portfolio temperature score. This potential flaw could be addressed with a sector-
specific default temperature score for each sector. Companies operating in these sectors that
have no valid targets would then be assumed to decarbonize along the sector averages. Whilst
reasonable from a climate performance and contribution perspective, the temperature score – as
introduced by this methodology – aims to score a company’s climate target ambition and set
incentives accordingly. Moreover, new types of target metrics are being introduced by relevant
climate frameworks to equip those low-carbon companies with more suited target metrics (e.g.,
“maintenance targets”). Finally, this methodology scores climate targets across all emission
scopes (scope 1–3). This means, the emission performance or contribution of all emission scopes
of a company must inform a potential sector-specific default score. Future solutions will need to
balance adequacy target incentivisation.
8.1.4. Climate target metric(s)
This methodology allows scoring absolute and intensity GHG emission reduction targets. Other
metrics used for climate target setting, such as procurement, engagement, renewable electricity
targets, etc. as well as long-term qualitative commitments cannot be scored currently – partly due
to a lack of suitable variables in AR6 scenarios. Temperature scores might therefore not capture
a company’s whole climate ambition. Yet, one could argue that any climate target set using other
metrics should materialize in absolute and / or intensity GHG emission reduction.
Relying only on GHG emission reduction targets might come with further limitations: the
assessment of one metric type, GHG emission reductions, does not necessarily provide the full
picture of a company’s alignment with long-term or structural changes needed to meet the
temperature goals of the Paris Agreement. For example, two approaches to reducing power
related GHG emissions by 30% in 10 years (e.g., 2020–2030) may correspond to very different
outlooks for the subsequent 10 years (e.g., 2030–2040) based on the lifespan of assets, etc.,
which are not captured by GHG emissions targets. This uncertainty can be reduced by assessing
the temperature alignment of all short-, mid-, and long-term GHG emission reduction targets for a
company in cases where they have been disclosed and to consider further metrics informing a
company’s actions taken to transition.
8.1.5. Forward- versus backward-looking assessment
Providing companies with a temperature score based on the ambition of stated targets implicitly
assumes that the targets will be met. If the targets are not met, companies may be given unfairly
low temperature scores. The converse is also true; if companies exceed GHG reduction targets,
their scores are biased high.
Moreover, the methodology only scores companies based on their forward-looking ambition as
indicated by GHG targets without considering prior actions the company has taken to reduce GHG
emissions. This might penalise companies that have already reduced GHG emissions
considerably and whose cost of emissions reductions will likely increase as low-cost / high-return
51
© CDP Worldwide and WWF International 2024
options are already exhausted. Besides “early movers”, companies that, by nature of their
business model, are already operating at low emissions might similarly be disadvantaged.
Due to those limitations, FIs are well advised to consider further complementary climate metrics
of companies (e.g., metrics tracking past and current emissions and climate performance). Further
research will explore approaches combining forward-looking and backward-looking indicators.
8.1.6. Carbon credits and carbon offsets
Currently, temperature scores based on this methodology do not capture whether and to what
extent carbon credits and offsets are built in a company’s climate target. This means, companies
that have a higher ambition in numeric terms but relying on carbon credit and / or offsets to achieve
this higher GHG reduction commitment might be unfairly rewarded with a better temperature score
compared to companies with lower numeric GHG reduction ambition, though building on own
mitigation efforts along the value chain only. One main constraint to solve this issue is the lack of
data around carbon credits and offsets usage by companies. With the expected increase in
transparency of climate related disclosure by companies worldwide, future updates of this
methodology will aim to explore the treatment of carbon credits and offsets in climate targets.
Specifically, whether and how temperature scores can be adjusted to reflect carbon credits and
offsets usage will be part of future research.
8.1.7. Assurance of GHG and GHG reduction target data
GHG emission data considered for computing TS according to this methodology is not required to
have some level of assurance. This means reported GHG emission data is taken at face value.
There is evidence that non-assured carbon accounting underestimates actual emissions and that
assurance has an influence on a company reducing future emissions (Berg et al., 2024
26
). Future
research might explore how to account and potentially adjust for assurance in this context.
8.2. Outlook: Next steps and future research
This version represents an updated version of the CDP-WWF Temperature Rating methodology
(version 1.0, released in October 2020) (CDP & WWF, 2020). This methodology will continue to
evolve over time to include latest climate science in addition to further improvement to address
current methodological limitations.
Future versions may consider further development and research on the following issues:
- Expand the assessment scope to include backward-looking indicators (in terms of past
emission performance tracking and/or progress against targets).
- Adequacy and possibility of developing non-linear models, which might give a better
fit.
26
Accessible through this link: http://dx.doi.org/10.2139/ssrn.4734240.
52
© CDP Worldwide and WWF International 2024
- Exploring the possibility of developing sector-specific warming functions for more
adequate benchmarking.
- Enhance sector representativity of default scores (sector-specific default scores),
including for low-emitting companies (e.g., renewable energy producers, clean
transport solutions, energy storage, climate solutions, etc.).
- Further guidance on portfolio aggregation approaches for different applications (e.g.,
scoring an equity portfolio vs. an index).
- Increase the assessment scope of climate target metrics (e.g., renewable energy
consumption targets), including exploring a potential default treatment of SBTi targets.
- Account for the usage of carbon credits in company climate targets.
53
© CDP Worldwide and WWF International 2024
9. Version control
Table 13: Document version history
9.1. Change log
Main changes from version 1.0 to version 1.5 are summarized in Table 14: Change log.
Table 14: Change log
Section
Version 1.5
Version 1.0
1.2.
Method purpose and limitations
n/a
4.2.
Changed from LAR (linear annual reduction) to CAR
(compound annual reduction)
LAR formula (1.3.)
5.
AR6: Update model simulations with scenarios to latest
IPCC Sixth Assessment report
SR15: Based on IPCC Special Report
1.5°C
5.2.
Emissions reduction ambition of zero or less gets a TS
equal to intercept
n/a
5.2., 5.3.2
All climate model simulations with scenarios that passed
IPCC'
s vetting process are considered sufficient criteria for
selecting the scenarios for the linear regression models
SR15 model simulations with scenarios
were filtered before the linear
regression models were generated,
based on a set of normative
precautionary preferences concerning
overshoot and the level of CDR
5.3.
Default score 3.4°C to reflect CAT (Climate Action
Tracker) projection December 2023
Default score 3.2°C to reflect CAT
projections December 2019
5.3.1.
Update sector variables/benchmarks and associated
regression models and clarification of sector benchmarks.
The sector variables were selected and allocated to a
sector based on a combination of data availability, how
well the AR6 variable was suitable to the specific sector
and the fit of the linear regression model. Sector
variables/benchmarks available: Power Generation
(absolute and intensity), Cement/Steel/Aluminium
(absolute)
Limited sector variables/benchmarks
based on SR15 data
5.3.1.
Introducing a linear regression model for Scope 2 targets
(for all sectors). The two variables applied for absolute
n/a
Version Name
Description
Date published
Consultation method
Draft method published to coincide with the method
consultation period which ran from April 30 – May 22,
2020.
April 30, 2020
Beta method
Beta version to be used for testing.
June 30, 2020
Version 1.0
Updated methodology incorporating feedback from beta
testing process.
October 1, 2020
Version 1.5
consultation
Updated methodology. See Table 14: Change log.
May, 2024
Version 1.5
Updated methodology incorporating feedback from public
consultation.
July, 2024
54
© CDP Worldwide and WWF International 2024
and intensity targets: Emissions | CO2 | Energy | Supply
and Emissions | CO2 | Energy | Supply / Secondary
energy
5.4.
Introduction of a temperature score floor at 1.5°C
n/a
6.1.
Introducing target vintage and reduction ambition criteria
for target selection for scoring
Target quality criteria (2.1.)
6.1.2.
All targets are now calculated on a pro-rata basis, based
on boundary coverage and CAR is adjusted accordingly
Target boundary of 67% and 95%
respectively were treated as full targets.
LAR was adjusted if boundary coverage
was below these thresholds
6.1.6.
- Short term for targets with target year (TY) < 5 years
- Mid-term for targets with 5 ≤ TY < 10 years
- Long term for targets with TY ≥ 10 years
Broader mid-term definition including
targets with target years in 5–15 years
6.1.6.
Targets are now valid during the calendar year when the
target expires
Targets were not valid in end year
6.2.
New target criteria waterfall with target vintage as first
consideration
Target vintage was not considered for
target selection. Boundary coverage
was a more important criterion
6.3.
Introducing explicit target criteria waterfall to prioritize
target selection
Less detailed target filtering
6.3.
Introducing aggregation for single scope targets for all
scopes
Less flexible target aggregation based
on combined scope 1 and 2 targets and
separate scope 3 targets
8.2.
Method roadmap
n/a
12.1
Warming function vs single scenario
n/a
12.5
Data requirements
n/a
55
© CDP Worldwide and WWF International 2024
10. Consultation questions
Please fill in the survey here: https://forms.office.com/e/rMbZEYFnba
10.1. Step 2b
10.1.1. Should older targets be considered in the absence of recent target data?
In the target criteria waterfall (Section 6.3) target vintage is determined as the most important
criteria for target selection. Following this hierarchy, the methodology would use whatever the
company has communicated publicly most recently about their targets as input in the target
validation process.
At times, companies communicate targets for different scopes at different times and they may
update and restate parts of their targets at different times. For example, company A has set a
combined scope 1 and 2 targets in 2021. At the same time, they also set a scope 3 target. Two
years later, in 2023, the company updates its scope 1 and 2 targets. For the scope 3 target should
the methodology:
A) Be conservative and assume that the scope 3 target is no longer valid, as the company
has not confirmed that the old scope 3 target is still active, or
B) Assume that the scope 3 target is still active, even though the company has only
communicated about scope 1 and 2 and not actively restated the old target.
One rationale supporting option A is to better capture potential corporate strategy changes, e.g.,
due to CEO changes, or changes in political context, which could imply a withdrawal, or an
ambition pause. The assumption that companies would likely report active targets in their updated
target, further supporting option A.
On the contrary, option B would allow to cover cases where companies do not expressly reiterate
targets from one year to the next in their annual reports – a company could set a scope 1+2 target
in year N and announce a scope 3 target in year N+1 without formally reiterating that the recently
set scope 1+2 target remains active. Option B gives companies more leniency by looking at the
latest available target per scope component, while option A incentivizes consistent, thorough
target communication.
10.1.2. Should TS aggregation based on GHG weighting be conducted based on
base year or current year data?
For combined TS, version 1.0 of this methodology is used to weigh scope 1 and 2 TS using base
year emissions. The resulting scope 1+2 TS would then be aggregated with the scope 3 TS to
calculate a scope 1+2+3 TS, and that aggregation was carried out using current year GHG data
(mostly because scope 1+2 and scope 3 targets might have very different base years, as well as
due to the volatility of scope 3 data).
This version 1.5 currently envisages to aggregate all single-scope TS (as opposed to combined
1+2 TS with scope 3) using only current year GHG data. The benefit is that GHG weighting is
56
© CDP Worldwide and WWF International 2024
consistent across all scopes, and reflects the latest available data as well as the scope breakdown
evolution over the target’s lifetime. It also provides better consistency in terms of GHG accounting.
The drawback is that this creates a discrepancy between the way other target parameters are
calculated (namely ambition and boundary coverage, which rely on base year data). This also
means that a combined scope TS will evolve over the target’s lifetime, as the breakdown between
scope 1 and scope 2 changes (in some cases, this can result in scope 1+2 TS marginally
increasing if the company delivers on the scope component with the higher ambition – i.e., the
part of the target that had the most positive influence on the TS in the first place).
A) I agree with the approach proposed for Version 1.5.
B) I think there are alternative aggregation approaches that are more appropriate. Please
explain.
10.1.3. Aggregation when missing GHG data
There may be cases where a company’s targets produce valid TS for single scopes, but where
GHG data is missing for the aggregation step. In such cases, there are two options the method
can use:
i) Assign a default score to the combined TS (for example scope 1 TS = 1.8 and scope
2 TS = 1.5 would be combined to a scope 1+2 TS of 3.4). This option accentuates the
necessity for companies to improve their reporting of GHG emissions.
ii) Use a function such as MAX(TS
1
, TS
2
), whereby the example above would give a
scope 1+2 TS of 1.8. This is mathematically reasonable as the average must be
somewhere between TS
1
and TS
2
. Furthermore, this option would contribute to a
portfolio score which is closer to the actual ITR performance of the portfolio and it would
also give recognition to the fact that valid targets have been set.
In your opinion, which alternative is the most reasonable?
A) Alternative i) is the more reasonable approach, or
B) Alternative ii) is the more reasonable approach.
10.1.4. Weighting of multiple scope 3 targets
As GHG emissions data is often not available for individual scope 3 targets, an alternative
aggregation approach is needed, to be able to score a companywide scope 3 target. In Chapter
6, the method allows for the use of an equal weight average as described in Section 6.4.1.
A) I agree with this approach
B) I think there are alternative aggregation approaches that are more appropriate. Please
explain.
57
© CDP Worldwide and WWF International 2024
10.1.5. Scoring of portfolios including targets using the CDP–WWF Temperature
Score- or Temperature Rating methodology
Do you agree with the scoring approach taken in step 2b, section 6.2.3.1. (see “Target type
exception”)?
10.1.6. Do you agree with the scoring approach taken in Step 2b, Section 6.3.3.1.
(see “Target type exception”)? If not, what alternative suggestion do you have?
10.1.7. Do you agree with the general approach taken in Step 2b, Chapter 6
“Company scoring”?
Yes or No. If no, please explain.
10.2. Step 1
10.2.1. Do you agree with the approach taken in Step 1, “Create benchmarks”?
Yes or No. If no, please explain.
10.3. Step 2a
10.3.1. Do you agree with the approach taken in Step 2a, Chapter 5 “Target
validation”?
Yes or No. If no, please explain.
10.4. Step 3
10.4.1. Do you agree with the approach taken in Step 3, Chapter 7 “Portfolio
scoring”?
Yes or No. If no, please explain.
10.5. General questions
10.5.1. Do you have suggestions for future research that could inform the next
version of the CDP–WWF Temperature Score methodology?
10.5.2. Do you have any other comments?
58
© CDP Worldwide and WWF International 2024
11. References
Berg, F., Oliver Huidobro, J., & Rigobon, R. (2024). On the Importance of Assurance in Carbon
Accounting (S&P Global Market Intelligence Research Paper Series). S&P Global Market
Intelligence. https://ssrn.com/abstract=4734240
Byers, E., Krey, V., Kriegler, E., Riahi, K., Schaeffer, R., Kikstra, J., Lamboll, R., Nicholls, Z.,
Sanstad, M., Smith, C., van der Wijst, K.-I., Al Khourdajie, A., Lecocq, F., Portugal-Pereira,
J., Saheb, Y., Strømann, A., Winkler, H., Auer, C., Brutschin, E., … van Vuuren, D. (2022).
AR6 Scenarios Database hosted by IIASA. International Institute for Applied Systems
Analysis. https://doi.org/10.5281/zenodo.5886911
CDP Worldwide & WWF International (2020, April 24). Temperature Rating Methodology A
temperature rating method for targets, corporates, and portfolios. https://cdn.cdp.net/cdp-
production/comfy/cms/files/files/000/003/741/original/Temperature_scoring_-
_beta_methodology.pdf.
Climate Action Tracker (2023). Warming Projections Global Update.
https://climateactiontracker.org/documents/1187/CAT_2023-12-
05_GlobalUpdate_COP28.pdf
Intergovernmental Panel on Climate Change (IPCC) (Ed.). (2023). Annex III: Scenarios and
Modelling Methods. In Climate Change 2022 - Mitigation of Climate Change: Working Group
III Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate
Change (pp. 1841–1908). Backmatter, Cambridge: Cambridge University Press.
https://dx.doi.org/10.1017/9781009157926
Intergovernmental Panel on Climate Change (IPCC) (Ed.). (2022). Climate Change 2022:
Mitigation of Climate Change. Contribution of Working Group III to the Sixth Assessment
Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.
https://www.ipcc.ch/report/ar6/wg3/chapter/chapter-2/
Liljedahl, I., & Rondahl, E. (2022). Evaluation of warming function vs single scenario [Master’s
thesis, KTH]. DIVA. https://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-318130
Science Based Targets initiative. (2022, August). Financial Sector Science-Based Targets
Guidance version 1.1. https://sciencebasedtargets.org/resources/files/Financial-Sector-
Science-Based-Targets-Guidance.pdf
Science Based Targets initiative. (2024, March). SBTi Corporate Net-Zero Standard version 1.2.
https://sciencebasedtargets.org/resources/files/Net-Zero-Standard.pdf
United Nations Environment Programme (2023). Emissions Gap Report 2023: Broken Record –
Temperatures hit new highs, yet world fails to cut emissions (again).
https://doi.org/10.59117/20.500.11822/43922.
59
© CDP Worldwide and WWF International 2024
World Business Council for Sustainable Development & World Resources Institute. (2015). The
Greenhouse Gas Protocol. A Corporate Accounting and Reporting Standard. Revised
Edition. https://ghgprotocol.org/sites/default/files/standards/ghg-protocol-revised.pdf
60
© CDP Worldwide and WWF International 2024
12. Annex
Annex 1: Warming Function versus Single Scenario
When it comes to temperature alignment assessments, there are two related approaches to
benchmark construction: the Warming Function approach and the Single Scenario approach.
The Warming Function approach (upon which this methodology relies) involves choosing a set
of suitable climate scenarios based on the company’s sector and target scope(s). A function of
the scenario set is constructed using a regression, which is then used as a benchmark for
assessing company alignment. The Warming Function relies on the Contraction Approach (or
Rate of Change approach), whereby all companies are expected to decarbonize at the same pace
(within a sector – this methodology provides four sector-specific benchmarks – or the wider
economy) along a linear path. The company’s annual rate of GHG emissions reduction (as implied
by its targets) is then compared to the Warming Function’s benchmark.
This approach reduces the risk of scenario selection bias in the benchmark construction
(essentially by taking a diversification approach to scenario selection, akin to portfolio
management). The warming function approach also makes it easier to compare temperature
scores from different sources, as scenario selection has less impact on the resulting temperature
score. This way, the Warming Function approach helps harmonizing the market and provides
necessary conditions for standardized temperature scores. This could help reducing some of the
criticism ESG metrics have received recently and drive acceptance in the marketplace for ESG
metrics in general and temperature scores in particular.
However, the warming function approach has drawbacks. It requires a higher number of varied
climate scenarios and tends to be less case-specific, which can lead to less accurate results. For
example, assessing a cement company and a wind power company based on the same
benchmark would yield an unfair result due to the different emission reduction levels required by
these sectors. In addition, the warming function approach has been criticized for its lack of
transparency. It has been perceived as a “black box”, as the implications of and dependencies of
the ITR results on the different underlying scenario assumptions are difficult to understand.
In the Single Scenario approach, a single scenario is chosen as a benchmark, which can be
adapted to a specific company and target. This allows for a granular assessment of the target
alignment to the chosen scenario. However, it also opens the possibility for bias, as the scenario
can be chosen to benefit the company and the targets assessed. It also involves an inherent
overreliance on the assumptions of one scenario, which is all the more problematic if a scenario
is chosen relying on assumptions which ultimately might not be plausible.
The Single Scenario method can either rely on:
- the Contraction/Rate of Change Approach (as does the Warming Function),
61
© CDP Worldwide and WWF International 2024
- the Convergence Approach, whereby the physical GHG emissions intensities of
companies within a sector are expected to converge towards the same sectoral value over
time, and
- the Fair Share Budget Approach, which combines a Rate of Change with a Carbon
Budget Approach. To compute a temperature score, the company’s rate of change and its
cumulative GHG emissions are compared to their assigned budget, often calculated
through its market share.
While scenarios chosen for a Single Scenario analysis are generally more detailed in terms of
sector granularity, regions, and units available for analysis, they can infer a false sense of security
for users, and lead to misinterpretation due to the significant variations in available mitigation
pathways. In other words, even though the motivation for selecting certain assumptions might
seem robust, the alignment result would be highly dependent on those assumptions to materialize.
It is also worth mentioning that scenarios chosen for a Single Scenario analysis can also be used
in a scenario sample applied in a Warming Function.
Recent research conducted at KTH, the Royal Institute of Technology, Stockholm, Sweden
(Liljedahl & Rondahl, 2022), concluded that the influence of scenario selection on the company-
level outcome was very high, with some temperature scores varying by 0.95°C depending on the
scenario used (which represents up to 50% of the range of temperature scores expected from the
CDP–WWF Temperature Score methodology, i.e. from the temperature floor of 1.5°C to the 3.4°C
default score.
According to Liljedahl and Rondahl (2022), because scenarios present only a certain trajectory
based on a set of assumptions and not a range of outcomes with associated probabilities, a
scenario cannot replace a thorough analysis of different outcomes.
The research by Liljedahl and Rondahl (2022) concludes that Warming Functions enable less bias
in the choice of scenarios, but also finds that under such methodologies, companies within
different sectors are often assessed against the same cross-sectoral benchmark. Ultimately,
Liljedahl and Rondahl recommend introducing more sectoral granularity in Warming Function
methodologies. The CDP–WWF Temperature Scoring Methodology currently includes sector
warming functions for power generation, steel, aluminium and cement. For version 2.0 CDP and
WWF intend to introduce more sector warming functions, as described in Section 8.1.3.
62
© CDP Worldwide and WWF International 2024
Annex 2: Details of sector variables
Table 15: Details of sector variables
Name of AR6 variable
Name of AR6 variable
in code
Description of variable
Emissions | Kyoto
Gases
Emissions|Kyoto
Gases
Emissions including the seven GHG gases under
the Kyoto Protocol; Carbon dioxide (CO
2
),
Methane (CH
4
), Nitrous oxide (N
2
O),
Hydrofluorocarbons (HFCs), Perfluorocarbons
(PFCs), Sulphur hexafluoride (SF
6
), Nitrogen
trifluoride (NF
3
).
Emissions | Kyoto
Gases /
GDP | PPP
INT.emKyoto_gdp
Emissions including the seven GHG gases under
the Kyoto Protocol (see above) divided by gross
domestic product.
Emissions | CO
2
|
Energy | Supply
Emissions|CO
2
|
Energy|Supply
CO
2
emissions from fuel combustion and fugitive
emissions from fuels: electricity and heat
production and distribution, other energy
conversion (e.g., refineries, synthetic fuel
production, solid fuel processing, incl. pipeline
transportation, fugitive emissions from fuels and
emissions from carbon dioxide transport and
storage (Byers et al. 2022).
Emissions| CO
2
|
Energy | Supply /
Secondary energy
INT.emCO
2
energysupp
ly_SE
CO
2
emissions from fuel combustion and fugitive
emissions from fuels: electricity and heat
production and distribution, other energy
conversion (e.g., refineries, synthetic fuel
production, solid fuel processing incl. pipeline
transportation, fugitive emissions from fuels and
emissions from carbon dioxide transport and
storage divided by the total secondary energy –
(the sum of all secondary energy carrier
production)) (Byers et al., 2022).
Emissions | CO
2
|
Energy and Industrial
Processes
Emissions|CO
2
| Energy
and Industrial
Processes
CO
2
emissions from energy use on supply and
demand side and from industrial processes (Byers
et al., 2022).
63
© CDP Worldwide and WWF International 2024
Annex 3: Result of linear regression model
The figures below show the result of the linear regression model for the variable Emissions | Kyoto
Gases (applied as a default variable for all sectors for scope 1 and scope 3 targets) for the 5- to
30-year time horizon (with a 5-year interval). The time frames that are used in this methodology
are 5-year (for short term targets), 10-year (for medium term targets) and 30-year (for long term
targets). To demonstrate the pattern of the linear regression model across time, all time frames
are shown in figures 6–11.
Figure 6: Result of the linear regression model for Emissions | Kyoto Gases for a 5-year time frame
64
© CDP Worldwide and WWF International 2024
Figure 7: Result of the linear regression model for Emissions | Kyoto Gases for a 10-year time frame
65
© CDP Worldwide and WWF International 2024
Figure 8: Result of the linear regression model for Emissions | Kyoto Gases for a 15-year time frame
66
© CDP Worldwide and WWF International 2024
Figure 9: Result of the linear regression model for Emissions | Kyoto Gases for a 20-year time frame
67
© CDP Worldwide and WWF International 2024
Figure 10: Result of the linear regression model for Emissions | Kyoto Gases for a 25-year time frame
68
© CDP Worldwide and WWF International 2024
Figure 11: Result of the linear regression model for Emissions | Kyoto Gases for a 30-year time frame
69
© CDP Worldwide and WWF International 2024
Annex 4: Calculation examples
Table 16: Company Alpha (combined scope 1+2 target)
Label Variable Company Alpha Calculation details
Company activity Retail
Target scope(s) Scope 1; Scope 2
Target type Absolute
A Base year 2019
B Base year total emissions (Scope 1, tCO2e) 5,000,000
C Base year total emissions (Scope 2, tCO2e) 2,500,000
D Base year value covered (Scope 1, tCO2e) 3,000,000
E Base year value covered (Scope 2, tCO2e) 2,000,000
F Current Year total emissions (Scope 1, tCO2e) 4,500,000
G Current Year total emissions (Scope 2, tCO2e) 2,250,000
H Target year 2034
Target timeframe Mid-term (10-year horizon)
I Targeted reduction from base year (%) 50.0%
J Boundary coverage (Scope 1) 60.0% =D/B
K Boundary coverage (Scope 2) 80.0% =E/C
L Boundary coverage (Scope 1+2) 66.7% =(D+E)/(B+C)
M Normalized reduction ambition (Scope 1) 30.0% =J*I
N Normalized reduction ambition (Scope 2) 40.0% =K*I
O Normalized reduction ambition (Scope 1+2) 33.3% =L*I
P Normalized CAR S1 -2.3% =(1+M)^(1/(H-A))-1
Q Normalized CAR S2 -3.3% =(1+N)^(1/(H-A))-1
R Normalized CAR S12 -2.7% =(1+O)^(1/(H-A))-1
Scope 1 benchmark Emissions | Kyoto Gases
S Scope 1 TS (°C) 1.90 =2,46-0,24*(-P)*100
Scope 2 benchmark Emissions | CO2 | Energy | Supply
T Scope 2 TS (°C) 2.03 =2,40-0,11*(-Q)*100
Scope 1 + Scope 2 TS 1.94 =(S*F+T*G)/(F+G)
70
© CDP Worldwide and WWF International 2024
Table 17: Company Beta (Scope 1 intensity target + no Scope 2 target)
Label Variable Calculation details
Company activity
Target scope(s) Scope 1 No Scope 2 target disclosed
Target type Intensity N/A
A Base year 2020 N/A
B Base year total emissions (Scope 1, tCO2e)
C Base year total emissions (Scope 2, tCO2e)
D Base year value covered (Scope 1, tCO2e) 18,000,000 N/A
E Base year value covered (Scope 2, tCO2e) N/A N/A
F
Current Year total emissions (Scope 1, tCO2e)
G
Current Year total emissions (Scope 2, tCO2e)
H Target year 2026 N/A
Target timeframe Short-term (5-year horizon) N/A
I Targeted reduction from base year (%) -25.0% N/A
J Boundary coverage (Scope 1) 90.0% N/A =D/B
K Boundary coverage (Scope 2) N/A N/A =E/C
L Boundary coverage (Scope 1 + Scope 2) N/A N/A =(D+E)/(B+C)
M Normalized reduction ambition (Scope 1) -22.5% N/A =J*I
N Normalized reduction ambition (Scope 2) N/A N/A =K*I
O Normalized reduction ambition (Scope 1+2) N/A N/A =L*I
P Normalized CAR S1 -4.2% N/A =(1+M)^(1/(H-A))-1
Q Normalized CAR S2 N/A N/A =(1+N)^(1/(H-A))-1
R Normalized CAR S12 N/A N/A =(1+O)^(1/(H-A))-1
Scope 1 benchmark Emissions| CO2 | Energy | Supply / Secondary energy N/A
S Scope 1 TS (°C) 1.78 N/A =2,11-0,08*(-P)*100
Scope 2 benchmark N/A Default score
T Scope 2 TS (°C) N/A 3.40
Scope 1 + Scope 2 TS =(S*F+T*G)/(F+G)
1.99
Company Beta
Power generation
2,500,000
20,000,000
3,000,000
19,500,000
71
© CDP Worldwide and WWF International 2024
Table 18: Company Gamma (single Scope 1 + single Scope 2 target)
Label Variable Calculation details
Company activity
Target scope(s) Scope 1 Scope 2
Target type Absolute Absolute
A Base year 2022 2021
B Base year total emissions (Scope 1, tCO2e)
C Base year total emissions (Scope 2, tCO2e)
D Base year value covered (Scope 1, tCO2e) 5,000,000 N/A
E Base year value covered (Scope 2, tCO2e) N/A 1,000,000
F
Current Year total emissions (Scope 1, tCO2e)
G
Current Year total emissions (Scope 2, tCO2e)
H Target year 2040 2035
Target timeframe Long-term (30-year horizon) Long-term (30-year horizon)
I Targeted reduction from base year (%) -35.0% -75.0%
J Boundary coverage (Scope 1) 55.6% N/A =D/B
K Boundary coverage (Scope 2) N/A 100.0% =E/C
L Boundary coverage (Scope 1 + Scope 2) =(D+E)/(B+C)
M Normalized reduction ambition (Scope 1) -19.4% =J*I
N Normalized reduction ambition (Scope 2) -75.0% =K*I
O Normalized reduction ambition (Scope 1+2) =(M*B+N*C)/(B+C)
P Normalized CAR S1 -1.2% =(1+M)^(1/(H-A))-1
Q Normalized CAR S2 -9.4% =(1+N)^(1/(H-A))-1
R Normalized CAR S12 =(1+O)^(1/(H-A))-1
Scope 1 benchmark Emissions | CO2 | Energy and Industrial Processes
S Scope 1 TS (°C) 2.35 =2,58-0,19*(-P)*100
Scope 2 benchmark Emissions | CO2 | Energy | Supply
T Scope 2 TS (°C) 1.50 (*) =2,85-0,15*(-Q)*100
Scope 1 + Scope 2 TS =(S*F+T*G)/(F+G)
(*) 1.44°C before application of temperature floor
2.28
Company Gamma
Cement
-1.8%
1,000,000
9,000,000
8,000,000
700,000
60.0%
-27.8%
72
© CDP Worldwide and WWF International 2024
Table 19: Company Delta (Scope 3 target)
Label Variable Company Delta Calculation details
Company activity Auto manufacturer
Target scope(s) Scope 3
Target Scope 3 categories covered
Cat. 1 Purchased goods and services
Cat. 11 Use of sold products
Target type Absolute
A Base year 2019
B Base year total emissions (total Scope 3, tCO2e) 10,000,000
C Base year total emissions (Scope 3, cat. 1 & 11 only, tCO2e) 8,000,000
D Base year value covered (Scope 3, cat. 1 & 11 only, tCO2e) 5,000,000
E Current year total emissions (total Scope 3, tCO2e) 9,250,000
F Current year total emissions (Scope 3, cat. 1 & 11 only, tCO2e) 7,000,000
G Target year 2034
H Target timeframe Mid-term (10-year horizon)
I Targeted reduction from base year (%) 35.0%
J Boundary coverage (Scope 3, cat. 1 & 11 only, tCO2e) 62.5% =D/C
K Boundary coverage (total Scope 3, tCO2e) 50.0% =J*C/B
L Normalized reduction ambition (Scope 3, cat. 1 & 11 only, tCO2e) 21.9% =J*I
M Normalized reduction ambition (total Scope 3, tCO2e) 17.5% =K*I
N Normalized CAR (Scope 3, cat. 1 & 11 only, tCO2e) -1.6% =(1+L)^(1/(G-A))-1
O Normalized CAR (total Scope 3, tCO2e) -1.3% =(1+M)^(1/(G-A))-1
P Scope 3 benchmark Emissions | Kyoto Gases
Q Scope 3 TS (°C) 2.15 =2,46-0,24*(-O)*100
73
© CDP Worldwide and WWF International 2024
Annex 5: Summary of required data for applying the Temperature Scoring method
The following tables present a legend of the data required to apply the Temperature Scoring
method for a portfolio using data tools.
Portfolio data
Data field
Optional/Required
Format
Explanation
company_name
Optional
Text
Name of the company in your portfolio.
company_id
Required
Text
Identifier for the company in your portfolio, used to map target and
fundamental data to the company.
company_isin
Optional
Text
ISIN and/or LEI are used to map companies to the SBTi database.
company_lei
Optional
Text
ISIN and/or LEI are used to map companies to the SBTi database.
investment_value
Required
Number
The monetary value invested in the company. Used for aggregation.
engagement_target
Optional
TRUE or FALSE
Used for engagement analysis.
Table 20: Data legend for portfolio data
74
© CDP Worldwide and WWF International 2024
Data field
Optional/Required
Format
Explanation
company_name
Optional
Text
Name of the company in your portfolio.
company_id
Required
Text
Identifier for the company in your portfolio, used to map target and
fundamental data to the company.
target_type
Required
Absolute, Intensity, T_score or other
Type of target. Can be absolute or intensity based GHG emission
reduction target. From v1.5 can also be a temperature score (eg for and
FI).
intensity_metric
Required for intensity
targets
Text
The metric the intensity based GHG emission reduction target is based
on.
input_temp_score
Requred for T_score
targets
Number in decimals
For targets set using the CDP/WWF temperature scoring approach.
scope
Required
S1, S2, S1+S2, S1+S2+S3, S3
The scope(s) covered by the target.
s3_category
Required for S3 targets
Integer between 1 and 15
The scope 3 category of the target.
coverage_s1
Required for S1 targets
Number in decimals, between 0 and 1
The part of emissions covered in scope 1 for the target.
coverage_s2
Required for S2 targets
Number in decimals, between 0 and 1
The part of emissions covered in scope 2 for the target.
coverage_s3
Required for S3 targets
Number in decimals, between 0 and 1
The part of emissions covered in scope 3 for the target.
reduction_ambition
Required
Number in decimals, between 0 and 1
The emission reduction that is set as ambition in the target.
base_year
Required
Year (4-digit integer)
Base year of the target. Defines time frame together with end year.
end_year
Required
Year (4-digit integer)
End year of the target. Defines time frame together with base year.
start_year
Optional
Year (4-digit integer)
Year the target was announced.
statement_date
Required
Number in decimals, between 0 and 1
The date the target was confirmed or updated. If not specified,
the start year will be assumed.
base_year_ghg_s1
Required
tCO2e
Total reported GHG emissions for scope 1 for the company at the
base year of the target.
base_year_ghg_s2
Required
tCO2e
Total reported GHG emissions for scope 2 for the company at the
base year of the target.
base_year_ghg_s3
Required
tCO2e
Total reported GHG emissions for scope 3 for the company at the
base year of the target.
achieved_reduction
Optional
Number in decimals, between 0 and 1
Part of the reduction ambition of the target that is already achieved by
the company.
target_ids
Optional
Text
Identifier of individual targets.
Table 21: Data legend for target data
75
© CDP Worldwide and WWF International 2024
*Note: GHG data needed for aggregation of Scope 1+2 temperature scores as of version 1.5 of methodology.
**Note: Required if needed for the selected portfolio aggregation method.
Data field
Optional/Required
Format
Explanation
company_name
Optional
Text
Name of the company in your portfolio.
company_id
Required
Text
Identifier for the company in your portfolio, used to map target and fundamental data to the
company.
Isic
Required
Text
Sector classification code for the company based on the International Standard Industrial.
Used to map targets to the correct regression model. Should include at least the first two
levels: Section and Division.
country
Optional
Text
Country where the company has its headquarter. Used for analysis purposes only.
region
Optional
Text
Region where the company has its headquarter. Used for analysis purposes only. Can be
continental or more granular.
industry_level_1
Optional
Text
Level 1 through 4 of the industry classification of the company. Used for analysis purposes
only. Can be based on any industry classification system.
industry_level_2
Optional
Text
Level 1 through 4 of the industry classification of the company. Used for analysis purposes
only. Can be based on any industry classification system.
industry_level_3
Optional
Text
Level 1 through 4 of the industry classification of the company. Used for analysis purposes
only. Can be based on any industry classification system.
industry_level_4
Optional
Text
Level 1 through 4 of the industry classification of the company. Used for analysis purposes
only. Can be based on any industry classification system.
sector
Optional
Text
Sector of the company. Used for analysis purposes only. Can be based on any
classification system.
ghg_s1
Required*
tCO2e
tCO2e
Total GHG emissions for scope 1 for the company. Used for aggregation of temperature
scores on company level.
ghg_s2
Required*
tCO2e
Total GHG emissions for scope 2 for the company. Used for aggregation of temperature
scores on company level.
ghg_s3
Required*
tCO2e
Total GHG emissions for scope 3 for the company. Used for aggregation of temperature
scores on company level.
company_revenue
Required**
Number
In single currency (can be any currency you choose). Revenue of the company in the most
recent year.
company_market_cap
Required**
Number
Market capitalization of the company in single currency.
company_enterprise_value
Required**
Number
Enterprise value of the company in single currency.
company_total_assets
Required**
Number
Total assets of the company in single currency.
company_cash_equivalents
Required**
Number
Cash equivalents of the company in single currency.
Table 22: Data legend for fundamental company data
76
© CDP Worldwide and WWF International 2024
Annex 6: Tables, figures, and equations
Tables
Table 1: Linear regression models associated with different time horizons and variables ......... 13
Table 2: Sector variables and associated linear regression models for each target type and scope
category. ................................................................................................................................... 15
Table 3: Summary of linear regression results (note that rounding differences may occur). ...... 20
Table 4: Target class, wording, and scenario variables ............................................................. 22
Table 5: End-of-century warming projections based on a range of future scenarios (UNEP
Emissions Gap Report, 2023) ................................................................................................... 27
Table 6: Target quality criteria ................................................................................................... 31
Table 7: Target waterfall criteria ................................................................................................ 32
Table 8: Default weighting method objectives ........................................................................... 39
Table 9: Default weighting principles ......................................................................................... 39
Table 10: Details of portfolio aggregation methods ................................................................... 41
Table 11: Assessment of options against weighting objectives ................................................. 44
Table 12: Assessment of options against weighting principles .................................................. 45
Table 13: Document version history .......................................................................................... 53
Table 14: Change log ................................................................................................................ 53
Table 15: Details of sector variables ......................................................................................... 62
Table 16: Company Alpha (combined scope 1+2 target) ........................................................... 69
Table 17: Company Beta (Scope 1 intensity target + no Scope 2 target) .................................. 70
Table 18: Company Gamma (single Scope 1 + single Scope 2 target) ...................................... 71
Table 19: Company Delta (Scope 3 target) ............................................................................... 72
Table 20: Data legend for portfolio data .................................................................................... 73
Table 21: Data legend for target data ........................................................................................ 74
Table 22: Data legend for fundamental company data .............................................................. 75
Figures
Figure 1: Steps of the Temperature Scoring methodology ........................................................ 10
Figure 2: Example of the result of the linear regression model for Emissions | Kyoto Gases for a
30-year time horizon ................................................................................................................. 19
Figure 3: Summary and breakdown of 2100 warming projections based on a range of future
scenarios (Source: Climate Action Tracker, December 2023) ................................................... 26
Figure 4: Step 2b – generating temperature scores at a company level, based on valid publicly
disclosed targets or a default approach for companies with no valid targets. ............................ 30
Figure 5: Example of GHG and CO2 emissions pathways for a subset of scenarios, for illustrative
purposes only (Source: Figure 3.10 IPCC AR6 Chapter 3) ....................................................... 48
Figure 6: Result of the linear regression model for Emissions | Kyoto Gases for a 5-year time frame
................................................................................................................................................. 63
77
© CDP Worldwide and WWF International 2024
Figure 7: Result of the linear regression model for Emissions | Kyoto Gases for a 10-year time
frame ........................................................................................................................................ 64
Figure 8: Result of the linear regression model for Emissions | Kyoto Gases for a 15-year time
frame ........................................................................................................................................ 65
Figure 9: Result of the linear regression model for Emissions | Kyoto Gases for a 20-year time
frame ........................................................................................................................................ 66
Figure 10: Result of the linear regression model for Emissions | Kyoto Gases for a 25-year time
frame ........................................................................................................................................ 67
Figure 11: Result of the linear regression model for Emissions | Kyoto Gases for a 30-year time
frame ........................................................................................................................................ 68
Equations
Equation 1: Linear regression formula ....................................................................................... 13
Equation 2: Compound Annual Reduction Rate ........................................................................ 14
Equation 3: Temperature Score ................................................................................................ 20
Equation 4: Boundary coverage ................................................................................................ 24
Equation 5: Scoring of targets using CDP–WWF Temperature rating method v 1.0 and CDP–WWF
Temperature Scoring method v1.5 ............................................................................................ 34
Equation 6: Aggregation of multiple S3 TS ................................................................................ 37
Equation 7: Temperature score aggregation ............................................................................. 37
