Working paper sub-group 3B on metrics and midpoint characterisation factors

(1)

Working paper sub-group 3B on metrics and midpoint characterisation factors

4 July 2019

Produced by CDC Biodiversité as a part of the Aligning Biodiversity

Measures for Business initiative

(2)

workshop held in Brussels on 26-27 March 2019. It was prepared by the chair of SG3B sub-group as a draft for discussion on metrics and midpoint characterisation factors¹. It is circulated as a draft for discussion with all SG3B members to be elaborated into a SG3B position paper as input for the workshop which will be held in Brazil on October 2019 and subsequent discussions.

Its target audience is mainly methodology developers (and the working paper is thus extremely technical). The working paper is a hybrid between a set of proposals for discussions at the first SG3B webinar, and texts fully redacted as the future position paper (from SG3B and the broader project).

Overall, it should be considered as a draft – a starting point - to amend and complete (as some topics may not be covered at all at this stage) during and after the first SG3B webinar. In line with the objectives of the sub-group, the aim of the working paper is to build common ground between measurement approaches.

Objectives of the sub-group

1. Explore the differences between metrics and midpoint² calculations across different measurement approaches and the reasons for the current divergence.

2. Propose bridges between metrics (e.g. conversion factors or translation of characterisation factors in different metrics) and propose common midpoint characterisation factors.

3. Identify how to disaggregate footprinting metrics and aggregate site level metrics, creating complementarity between the two.

The first objective is mainly about creating a good understanding of the reasoning behind each metric (see Figure 1). The most promising area for collaboration between initiatives however lies in the second objective, so the time dedicated to the first objective will be limited and the subgroup will avoid debates about “the best metric” as much as possible.

Figure 1. Potential outcome of the Metrics and midpoint characterisation factors and Corporate data input for impact assessment subgroups: a (partial) harmonisation of inputs and midpoints facilitating conversions between metrics

1 Midpoint is used in this document to mean all points in the cause-effect chain (environmental mechanism).

Characterisation factors are impact coefficients used to calculate impacts.

2See Annex 1: Definitions.

Inputs (activity, pressure-

related data…) Midpoints Impacts

Initiative 1

Initiative 2 Initiative 3

Initiative 1

Initiative 2 Initiative 3 Corporate data

input sub- group #3A

Midpoint sub-group

#3B

(4)

The following outputs are expected of the sub-group:

1. Analysis of differences between metrics and midpoint calculations and reason for divergence

2. Mapping of the language of the LCA community with language used to describe a more direct measurement of biodiversity. This mapping will comprise language used by LCA practitioners, EIA practitioners, biodiversity specialists and natural capital accounting and assessment (Natural Capital Protocol)

3. The different measurement approaches will work together to explore linkages and how they can positively reinforce each other, identifying bridges between metrics and common mid-point characterisation factors and determining how site based and portfolio approaches can link and complement each other.

Context

Perimeter of the sub-group

In the cause-effect chains leading from data inputs to impacts on biodiversity, some intermediary steps (“midpoints”) are not biodiversity-specific (e.g. the assessment of the global mean temperature increase due to greenhouse gas emissions). Converging on these intermediary calculations is the main focus of sub-group 3B. The sub-group also focuses on the rationale behind the use of different metrics.

Input data are not within the perimeter of sub-group 3B as they are dealt with by sub-group 3A.

Figure 2 summarizes the perimeter of sub-group 3B and its connection with sub-group 3A.

Figure 2. Perimeter of sub-group 3B: metrics and midpoint characterisation factors of the corporate biodiversity measurement approaches³

Current issues regarding metrics and midpoint characterisation factors

Several initiatives seek to assess impacts on biodiversity. Different metrics are however used. Can bridges be found between these metrics in a similar way to conversion factors between meter and feet?

These initiatives include more or less the same main pressures on biodiversity, identified by the IPBES as Land / sea use change, Direct exploitation, Invasive alien species, Pollution, and Climate change.

Several steps are followed to go from inputs (e.g. greenhouse gas or GHG emissions in kg CO2 eq.) to biodiversity impacts (which could be called “endpoints” in life cycle assessment, or LCA, parlance).

3 Fall back data are data used when company’s data are not available, see below.

Input data

Sub-group 3A

Biodiversity impact assessment Tools or approach

Impact factors Midpoints Endpoints

Sub-group 3B

(midpoint characterisation factors)

① Company’sdata

②Fall back data sets

Sub-group 3B (rationale of the different metrics)

(5)

Figure 3 illustrates these steps and Table 1 lists some databases and models which allows to move from one step to the next.

Figure 3. Example of path between inputs (in blue), midpoints (in green) and endpoints (in red) (ASN Bank, 2016)

Table 1. Models and databases providing biodiversity-related midpoint and endpoint characterisation factors (non- exhaustive)

From data inputs to midpoints (midpoint CF) From midpoints to endpoints (endpoint CF)

Life cycle assessment databases such as ecoinvent

Environmentally extended input-output models such as EXIOBASE

UN Environment lifecycle initiative

CML (outdated) GLOBIO ReCiPe LC Impact

IUCN mean % decline over 10 years USEtox

Some of the intermediaries (“midpoints”) are common to most methodologies, for instance the global mean temperature increase (GMTI) is usually used to go from GHG emissions to the impact of climate on biodiversity. These midpoints are not always consistent between different biodiversity measurement approaches. Agreeing on them would ensure consistency between methodologies and increase the credibility of all the initiatives as a whole. Table 2 shows a beginning of convergence on one midpoint characterisation factor (IAGTP, integrated average global temperature potential⁴, i.e. the impact of the emission of a gas on the global temperature over time).

4 IAGTP is sometimes considered an "endpoint characterisation factor" but is called here a midpoint characterisation factor to keep things simple.

(6)

Table 2. Summary of existing approaches to assess the biodiversity impact of climate change (CDC Biodiversité, 2019)

Similarly, other considerations related to midpoint, like how to deal with impacts which last over many years (like the warming induced by GHG emissions, see Figure 4), or the issue of time integration (see below), could benefit from convergence between methodologies.

Figure 4. How to deal with impacts lasting beyond the period assessed like climate change impacts. Figures from Joos et al. (2013).

Rationale for convergence on midpoint characterisation factors

One potential significant interest of focusing on midpoints is that it could allow to build bridges between metrics, allowing translations between them in a way that is complicated or even impossible through direct conversion⁵.Figure 5. Midpoints can be used to create bridges between metrics (non exhaustive example of MSA and PDF and unit of extinction risk): instead of finding a direct conversion factor between MSA and PDF, intermediary midpoint results (in grey) can be used to calculate impacts (in green) in both of the metrics.

Two main options exist to build bridges between metrics using different units, as illustrated by Figure 5:

• Define conversion factors, allowing the direct translation between results

• Rely on common midpoints and apply the characterisation factors associated to each metric

5 Ideally, it would however be best to find direct conversion factors.

(7)

Figure 5. Midpoints can be used to create bridges between metrics (non exhaustive example of MSA and PDF and unit of extinction risk): instead of finding a direct conversion factor between MSA and PDF, intermediary midpoint results (in grey) can be used to calculate impacts (in green) in both of the metrics.

Common framework on metrics and midpoint characterisation factors

The following provides proposals of common definitions and concepts to be agreed by SG3B members.

Common definitions

A number of definitions are agreed upon and listed in Annex 1: Definitions. They include: midpoint.

Other definitions are being agreed upon by SG3A, including indicator, measure, metric, unit.

Time integration

A significant difference between approaches lies in how they deal with long-lasting impacts such as the impact of GHG emissions on climate change or chemical pollution which remains harmful in soil, air or water over several years.

GHG continue to warm the climate dozens (or even hundreds) of years after their emissions, well beyond the usual timeframe of most assessments. Tools using the PDF unit [BFFI, PBF] deal with this question by integrating impacts over time. Other tools like the GBS do not integrate over time but nonetheless takes into account the persistence of impacts through a static footprint⁶. These two approaches are illustrated in Figure 6.

This issue has consequences in terms of the units used (PDF.m².yr or PDF.m² for instance) and the values reported (much higher if time-integrated assessments are used). It thus impacts the capacity of non-expert stakeholders to understand the results but also how targets should be set (time-integrated targets need to be set if the methodology integrates its results over time).

6 ‘Dynamic footprint’ is the footprint caused by changes, consumptions or restorations. However, existing pressures can limit the ability of biodiversity to thrive even without any change in pressures. For instance, the very existence of a palm oil plantation prevents the area it occupies from growing back into a natural tropical forest and thus prevents biodiversity from reaching its full potential. This is the ‘static footprint’ or ‘ecological opportunity cost’ and it includes all the ‘persistent effects’

which remains over time.

Impacts 9 MSA.m²

? PDF.m².yr Difficulties

to translate directly impacts?

9 MSA.m²

8.9 PDF.m².yr (local effect) Use

midpoints instead

10 m²of natural forest converted to intensive agriculture

? unit of extinction risk (STAR) Midpoints

(8)

Figure 6. Illustration of the question of time-integration (CDC Biodiversité, 2019)

SG3B recognizes the importance to take into account the persistence of impacts over time and the need for each measurement approach to clarify how it currently deals with the issue. Table 3 provides a preliminary analysis.

Table 3. Overview of current practices regarding time integration among measurement approaches

Time integration approach Measurement approaches Time integration embedded

in the unit used (e.g.

PDF.m².yr)

BFFI, PBF

Distinction of dynamic (integrated over the assessment period) and static impacts

GBS

No time integration AI, BF, BIE, BIM, EP&L, LIFE Index, STAR

Differences between metrics

Four main synthetic "metrics" are currently used to aggregate quantitatively impacts or dependencies:

mean species abundance (MSA), potentially disappeared fraction (PDF), risk of extinction units and monetary (e.g. euros). A fifth option, followed by some initiatives, does not use synthetic metrics and instead relies on qualitative aggregations of assessments. LIFE Institute also has a specific approach with the use of an index which combines measures in MSA with other indicators to calculate an index used to assess whether companies meet a threshold and qualify or not for the LIFE certification.

(9)

Table 4, Table 5, Table 6 and Table 7 provide preliminary analysis of the different approaches regarding metrics. The work of the subgroup would improve this analysis and briefly describe the limitations associated to each metric.

(10)

Table 4. Aggregation method used by each metric

Metric [initiatives using the metric]

Aggregation method Reasoning behind the aggregation

Mean species abundance (MSA) [GBS, BIM, BF]

Arithmetic mean of abundances (same weight for all species)

Equal weights are a good default and explicit weighting is also possible.

Another aspect is that all species contribute to ecological functions.

Potentially disappeared fraction (PDF) [BFFI, PBF]

Number of species (same weight for all species)

Similar to MSA.

Risk of extinction unit [STAR]

Sum of the risks of extinction of species weighted by their threat status

Threat status of species has been evaluated in a scientifically consistent, multi- stakeholder, global process and the presence of threatened species in a site or habitat is an indication that the ecosystem is under pressure.

Natural capital monetary value (, e.g. EUR) [EP&L]

Sum of the economic value of ecosystem services (i.e. more weight to more valuable services)

Economic valuation gives the expression of the worth of the benefits people gain from the environment. Using this assessment allows to better understand and address impacts and prioritize actions.

[BIE, …] No single quantitative metric.

Aggregation approach is still to be determined

State / pressure / response indicators are required to meet sites’ needs and such indicators are difficult to aggregate quantitatively, so a qualitative aggregation is used.

Table 5. State of biodiversity covered by each metric

Metric [initiatives using the metric]

State of biodiversity covered

Reasons why some state of biodiversity are not covered

Capacity to assess biodiversity state based on ecological surveys (direct measurements)

Mean species abundance (MSA) [GBS, BIM, BF]

Terrestrial and aquatic (freshwater)

No endpoint characterisation factors for marine biodiversity

Possible in theory

(11)

Potentially disappeared fraction (PDF) [BFFI, PBF]

Terrestrial, aquatic (freshwater) and marine

?

Risk of extinction unit [STAR]

Terrestrial, aquatic (freshwater) and marine?

Possible

Natural capital monetary value (e.g. EUR) [EP&L]

Terrestrial only? ? Likely to be challenging given that values of

biodiversity are known not to be well represented currently into natural capital assessments

[BIE, …] Terrestrial, aquatic (freshwater) and marine?

Possible

Table 6. Impacts on biodiversity, and associated pressures, covered due to the endpoint characterisation factors available for each metric

Endpoint characterisation factors and associated capacity to assess the impact of pressures Metric

[initiatives using the metric]

Available endpoint characterisation factors

Land / sea use change

Direct exploitation Invasive alien species

Pollution Climate change Other

Mean species abundance (MSA) [GBS, BIM, BF]

GLOBIO’s pressure-impact relationships

Land use,

Fragmentation, Encroachment, Hydrological disturbance, Wetland conversion

Not covered directly

Not covered Atmospheric nitrogen deposition, Nutrient

emissions, Land use change in catchment

Climate change

Potentially disappeared

ReCiPe or LC Impact’s

Land occupation, Land

Not covered Not covered Terrestrial ecotoxicity,

Climate change

(12)

fraction (PDF) [BFFI, PBF]

characterisation factors

transformation, (regional) Water scarcity

Terrestrial acidification, Marine ecotoxicity, Marine

eutrophication, Freshwater eutrophication, Freshwater ecotoxicity

Risk of extinction unit [STAR]

No

characterisation factor but assessment of the level of pressures through the IUCN Red List

Residential &

Commercial Development, Agriculture &

Aquaculture, Energy Production

& Mining, Transportation &

Service Corridors, Human Intrusions

& Disturbance, Natural System Modifications

Biological Resource Use

Invasive &

Problematic Species,

Pathogens &

Genes

Pollution Climate Change Geological Events

Natural capital value (monetary, e.g. EUR) [EP&L]

No

characterisation factor

[BIE, …] No

characterisation factor⁷

7 The level of pressure is nonetheless assessed based on site documentation.

(13)

Table 7. Limitations of each metric (to be completed after the 11 June webinar)

Metric [initiatives using the metric]

Type of biodiversity covered⁸

Other limitations (on top of those listed in the previous tables)

Mean species abundance (MSA) [GBS, BIM, BF]

Ecological integrity

To be completed

Potentially disappeared fraction (PDF) [BFFI, PBF]

Ecological integrity

To be completed

Risk of extinction unit [STAR]

Extinction risk

To be completed

Natural capital value (monetary, e.g. EUR) [EP&L]

Ecosystem services

To be completed

[BIE, …] Ecological integrity &

extinction risk⁹

To be completed

8 Biodiversity types include Extinction risk, Ecosystem integrity and Ecosystem services.

9 BIE does not actually provide figures on the risk of extinction but rather on the risk of impact on important biodiversity features. “Extinction risk” is the current name of the category of biodiversity including this, but the name might change in the future to better reflect all the situations it covers.

(14)

The different metrics and aggregation approaches meet different needs and answer different questions such as: how best to maintain global functional diversity? What are the conservation actions with the highest potential to prevent species extinction? What are the impacts on other people’s dependencies?

How to ensure site-level no net loss?

To assess the consequences for decision making, it is useful to consider examples.

In example 1, a company considers transforming two patches of natural forest into intensive agriculture.

One patch of natural forest is located in Cambridge, United Kingdom, and the other is located in the Atlantic forest in Brazil. In the example, both are large patch of contiguous intact forest with healthy ecosystems. The forest in Cambridge hosts a few hundred species and only 1 endangered species while the Atlantic forest hosts a couple of thousands of species and many endangered species.

Intactness metrics like MSA and PDF will consider both forests equivalent because they are both undisturbed. So the company might decide to cut down the Atlantic forest.

Species-focused metrics like the risk of extinction will value the Atlantic forest more because of its high number of species and in particular endangered species.

Results from ecosystem service metrics like the natural capital value will depend on the potential beneficiaries of the services provided by both forests.

In example 2, another company is considering developing an undisturbed barren area with a few dozen species and no endangered species, far from any human activity.

Intactness metrics will warn against the destruction of this undisturbed area.

Species-focused metrics will conversely consider the low number of species means losses are limited.

Ecosystem service metrics will similarly consider the lack of beneficiaries mean this ecosystem has a low value.

However, the artificialisation of such an ecosystem would still lead to the complete loss of ecological functions, and potentially put at risk the survival of species whose habitats would be destroyed.

These two examples show the importance of complementary qualitative analyses and environmental safeguards, including the strict need to implement approaches compatible with the mitigation hierarchy at the site level. They also highlight the need to have multiple measures to ensure an appropriate decision is made.

Bridges between metrics

Objective #2 of the sub-group seeks to “Propose bridges between metrics (e.g. conversion factors or translation of characterisation factors in different metrics)”. The following explores two options to propose such bridges between MSA and PDF, which are the two metrics closest to each other and for which bridges seem the easiest to build. The process could be repeated for other metrics. These two options are direct conversion and the use of common midpoints, as illustrated above in Figure 3.

Preliminary information

Before explaining both options, it is useful to explain what characterisation factors are and explain options to deal with time integration and spatial integration.

Background information on characterisation factors

Characterization factors (CF) are used in Life Cycle Impact Assessment and express how much a single unit of mass of the substance of interest contributes to the midpoint or endpoint impact category. CF include a fate factor (FF), an exposure factor (XF) and an effect factor (EF) (Huijbregts et al., 2010)

𝐶𝐹 = 𝐹𝐹 × 𝑋𝐹 × 𝐸𝐹.

For instance, in the case of chemicals ecotoxicological characterisation factors, the fate factor represents the persistence of a chemical in the environment. The exposure factor represents the

(15)

bioavailability of a chemical, i.e. the fraction of chemical dissolved. The fate and exposure factors of chemicals can be retrieved from models such as USEtox¹⁰. The ecotoxicological effect factor reflects the change in the Potentially Affected Fraction (PAF) of species due to change in concentration (PAF.m³.kg^-1). The PAF is a measure of toxic stress, it indicates the fraction of species for which the concentration in the environment exceeds the no observable effect concentration (NOEC). The PAF thus expresses the potential for adverse effect, it gives no indication on the sort of effect nor on its extent. A high PAF may result in a small observable effect.

MSA and PDF

MSA is an index of biodiversity intactness varying between 0 and 1 - or 0% and 100% - but MSA-based metrics usually integrate MSA over surface areas, and are expressed in MSA.km² or MSA.ha. PDF is used for endpoint CF that integrates a time and a spatial dimension, often expressed in PDF.m².yr. PAF is used for EF integrating a spatial dimension (PAF.m³.kg^-1), and for midpoint CF integrating both time and spatial dimensions (PAF.m³.day.kg^-1).

Ignoring the time and spatial dimensions, we could consider rough PDF-MSA and PAF-MSA relationships stating that

• MSA = 1 – PDF. Indeed, on a given ecosystem (say 100 km² of forest) if the potentially disappeared fraction of species is 20%, considering that the remaining MSA is 80% does not seem too inaccurate.

• MSA = 1 – x.PAF. Where x is the share of potentially affected species which actually disappeared. The x parameter is to be determined based on literature.

Such relationship cannot be used unless the spatial and time integrations are handled though.

Handling the spatial integration

Spatial integration means that the biodiversity measure is integrated over space, basically multiplying the biodiversity measure by the area over which it is measured. Thus, PDF.m² means PDF multiplied by m² and not PDF by m².

Over an ecosystem of area S with an homogeneous biodiversity value, this translates into:

∫^𝑥=𝑆[𝑏𝑖𝑜𝑑𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 𝑣𝑎𝑙𝑢𝑒 (𝑀𝑆𝐴 𝑜𝑟 𝑃𝐷𝐹)] = 𝑆 × [𝑏𝑖𝑜𝑑𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 𝑣𝑎𝑙𝑢𝑒]

𝑥=0

For instance 10% MSA over an area S of 10 km2 = 10% x 10 km² = 1 MSA.km².

In order to be able to use the PDF-MSA and PAF-MSA relationships, time integration needs to be handled. The best way to derive PDF or MSA from spatially integrated PDF.m² or MSA.m² would be to divide the latter by the area over which they have been integrated. However, this area is usually unknown so it is not possible.

An alternative might be to (i) assume that impacts are uniformly distributed over the area considered and (ii) use an average biodiversity density of the area to quantify how much biodiversity is affected.

Point (i) has significant drawbacks as impacts are not uniformly distributed for several pressures. As the average biodiversity density of an ecosystem is usually unknown, point (ii) implies the use of a less precise biodiversity density, such as a global biodiversity density. This in turns assumes that biodiversity is uniformly distributed over space, which is also debatable.

Since PDF or MSA global average density are not available, the global average species densities of 1.48.10^-8 species.m^-² ¹¹ for terrestrial biodiversity and 7.89.10^-10 species.m^-3 for aquatic biodiversity

10 USEtox is a scientific consensus model endorsed by the UNEP/SETAC Life Cycle Initiative for characterizing human and ecotoxicological impact of chemicals. https://usetox.org/

11 The species density (1.48.10-8 species.m^-²) applied to the total terrestrial area (140 million km²) gives a total number of 2 072 000 terrestrial species.

(16)

(Huijbregts et al. 2017) is used instead. PDF.m² or PAF.m² is thus translated into disappeared species or affected species.

Handling the time integration

Similarly to spatial integration, time integration means that the biodiversity measure is integrated over time (which is a bit trickier to conceptualize), basically multiplying the biodiversity impact by the time over which it will occur. Thus, PDF.yr means PDF multiplied by year and not PDF by year.

Over a period of time of T, if the impact on biodiversity is constant over time, this translates into:

∫^𝑡=𝑇[𝑏𝑖𝑜𝑑𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 𝑖𝑚𝑝𝑎𝑐𝑡 (𝑀𝑆𝐴 𝑜𝑟 𝑃𝐷𝐹)] = 𝑇 × [𝑏𝑖𝑜𝑑𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 𝑖𝑚𝑝𝑎𝑐𝑡]

𝑡=0

Dealing with the time integration requires to know the shape of the impact curve that was assumed to compute the impact and the considered time frame. Knowing these, the footprint could be broken down into its annual components (impacts on year 1, impacts on year 2, etc. until impacts on year N). In Figure 7, the impact on year 1 is called the dynamic footprint, and the persistent impact (which does not vary compared to the initial dynamic footprint) is called the static footprint. These are terms used by CDC Biodiversité to describe the impacts but it is not suggested to use them in this sub-group at this stage.

Figure 7. Illustration of the impact assessed through time-integration and the approximation of the impact through a “rectangular shape” assumption. MSA is used as an example but the principle is the same with any metric.

Most often we know which time frame was considered in the computation of endpoint CF in PDF.m².yr – ReCiPe for instance provides 3 categories of CF based on three different time horizons.

The issue of the shape of the impact curve is less easilly solved. For climate change we know that the considered time-horizon is 100 years and that the impact curve basically follows a rectangular shape so dividing the impact by 100 could be appropriate. This may however not be the case when the impulse-response function is not rectangular.

Gathering information on the shape of the impulse-response function could be done using the fate factor. FF indeed expresses the fate of the considered substance in the environment (concentration in the various compartments following a point emission) and encompasses a time dimension. As such, they somehow correspond to the residence time of the substance in the various

(17)

compartments. For instance the FF of toluene emitted in freshwater is 3.95 kg.kg^-1.day^{-1 12}, stating that 1 kg of toluene emitted directly in freshwater will remain there for 3.95 days [to be confirmed]. The yearly impact could be computed from there based on the ratio between FF (in days) and the duration of the assessment period (usually one year, i.e. 365 days).

Direct (endpoint-based) conversion option

Once spatial integration and time integration are dealt with, translating one metric to the next could be based on a comparison of endpoint characterisation factors related to common pressures, such as climate change. For the climate change pressure, we indeed know both the impact in species.yr (unit used in BFFI) and MSA% (used in the GBS). The computation is as follows:

In the BFFI, the climate change global impact is 2.8.10^-9 species.yr/kgCO2. Assuming a rectangular shape and dividing by 100 gives an impact of 2.8.10^-11species/kgCO2, further dividing by the IAGTP (6.5.10^-14 °C/kgCO2) gives a global impact of 4.3.10⁴ species/°C. Comparatively, the climate change global impact expressed in MSA is 5.21 MSA%/°C (Arets, Verwer, & Alkemade, 2014; Schipper, Meijer, Alkemade, & Huijbregts, 2016). Combining the two figures gives 8.3.10³ species/MSA%.

This conversion has two limitations. First, the figures are for climate change global impacts (on the whole planet), and would be different if the impacts of (global) climate change was assessed on a smaller ecosystem: the number of species lost per degree of temperature increase would be lower.

Therefore, using this conversion implies that it Is used at a global scale, over about 140 millions km² of terrestrial land. Second, it is important to keep in mind that this species-MSA% relationship is estimated based on climate change impacts, so that it somehow encompasses species sensitivity to climate change. Using it for other pressures thus assumes that the sensitivity to these pressures is the same as for climate change.

Figure 8 presents the summary of the steps to deal with spatial integration, time integration and conversion from PDF.m².yr to MSA%. As explained above, this process is fraught with difficulties and challenging assumptions. It is provided mainly as food for thought to spark discussions on how to translate PDF to MSA or vice-versa.

Figure 8. Summary of the direct conversion option

12 USEtox data

PDF.m².yr species.yr species

Conversion factor 1 PDF.m².yr = 1.8.10^-14MSA%

MSA%

× ×

×

Species density

per m²

Time horizon of 100 years

Species per MSA% computed based on CC impact data

(18)

Common midpoint-based conversion

Table 8 lists the data typically necessary to assess IPBES pressures: by converging on the calculations linking midpoints (central column) to pressure data (right column), the different approaches could increase their common credibility, and increase compatibility between approaches, without changing their proprietary calculations.

Table 8. Preliminary list of data and midpoints typically necessary to assess pressures. Items in bold orange are data which could be required from companies as explained in the dataset subgroups pre-read.

Pressure Midpoints Data typically necessary to

assess the pressures Land / sea use change Agricultural yields and quantities

produced

Water withdrawal and consumption

Land use changes (LUC) Hydrological disturbances

Direct exploitation To be discussed To be discussed

Invasive alien species To be discussed To be discussed

Pollution Emissions of pesticides, N & P Pesticide concentrations N & P concentrations

Climate change Greenhouse gas (GHG) emissions Global Warming Potential (GWP)

Global mean temperature increase (GMTI)

If the different measurement approaches all have characterization factors to translate these midpoints into endpoints (available endpoint CF are described in Table 6), then providing data on midpoints would allow indirect conversion (as illustrated by Figure 1).

Common midpoint characterisation factors

Objective #2 of the sub-group also seeks to “propose common midpoint characterisation factors”. The discussion around common midpoint characterisation factors should occur during the second webinar of the sub-groups.

Potential candidates for alignment are:

• IAGTP: see Table 2

• Agricultural area calculated based on yield data

• Water withdrawal vs water consumption

Link between midpoint-based approaches and site-level approaches

Objective #3 and output #3 of the sub-groups plan to explore how site based and portfolio approaches can link and complement each other. Part of the answer lies in sub-group 3A on data as data inputs useful for midpoint-based approaches could be collected by site-level approaches and vice-versa.

Other linkages and complementarity will be explored within this sub-group.

(19)

Common ground principles

The following principles are proposed for discussion, refinement or rejection in the sub-groups:

• Transparency: to include data sources, gaps, limitations.

• Fit for purpose: the data and methods used should match the objective, application and scope.

• Rigor: the information, data and methods used should be technically robust.

Discussion questions

1. What are the reasons for the differences between the metrics and midpoint characterisation factors used?

2. What are the decision implications of these differences?

3. If alignment is not feasible / practical how can this be communicated to stakeholders to avoid confusion or are translations / conversions between metrics possible?

4. What common approaches and midpoint characterisation factor values can be agreed on?

5. Is it feasible to develop ‘translation’ approaches between different metrics, what are the decision implications of introducing more estimation into the approach?

6. What Common Ground principles for corporate biodiversity measurement could promote alignment?

(20)

Annex 1: Definitions

Endpoint

Impact

Midpoint

Strictly speaking, a midpoint is a point in the cause-effect chain (environmental mechanism) of a particular impact category, prior to the endpoint, at which characterisation factors can be calculated to reflect the relative importance of an emission or extraction in a Life Cycle Inventory.

Midpoint is used with a slightly broader sense in this document, to include all points in the cause-effect chain, without reference to the life cycle framework.

Pressure

Working paper sub-group 3B on metrics and midpoint characterisation factors