possibilities for modelling sustainability

45
Possibilities for modelling sustainability scenarios Technical Report of Work Package 3 DRAFT VERSION 16 Aug 2008 Mats G E Svensson, Stefan Anderberg Roy Haines-Young, Allison Rollett Stefan Bringezu, Mathieu Saurat Lund University, Centre for Sustainability Studies University of Nottingham, Centre for Environmental Management Wuppertal Institute for Climate, Environment and Energy

Upload: liu-se

Post on 04-Feb-2023

0 views

Category:

Documents


0 download

TRANSCRIPT

Possibilities for modelling sustainability scenarios

Technical Report of Work Package 3

DRAFT VERSION 16 Aug 2008

Mats G E Svensson, Stefan Anderberg Roy Haines-Young, Allison Rollett Stefan Bringezu, Mathieu Saurat Lund University, Centre for Sustainability Studies University of Nottingham, Centre for Environmental Management Wuppertal Institute for Climate, Environment and Energy

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

1

Table of Contents

EXECUTIVE SUMMARY...............................................................................................3

1. INTRODUCTION ....................................................................................................4

2. SCENARIO ELEMENT SYSTEMATICS IN FORESCENE .....................................5

2.1. Cross-cutting driving forces: economic activities and underlying factors................. 9

2.2. Desirable sustainability goal references ......................................................................... 11

2.3. Key sustainability strategies ............................................................................................. 15

3. PRELIMINARY NARRATIVES IN FORESCENE..................................................18

4. REVIEW OF RELEVANT SCENARIO STUDIES AND SIMULATION MODELS IN COMPARISON WITH FORESCENE ...............................................23

4.1. Global and regional scenario studies of particular relevance to FORESCENE........ 23 4.1.1. Millennium Ecosystem assessment scenarios............................................................. 24 4.1.2. The Global Scenario Group ........................................................................................... 25 4.1.3. ATEAM and EURURALIS projects................................................................................ 26

4.2. Studies related to the FORESCENE topic areas ............................................................ 27 4.2.1. Material and energy flow models .................................................................................. 27 4.2.2. Land use and biodiversity models................................................................................. 30 4.2.3. Water use models .......................................................................................................... 32

4.3. Summing-up ......................................................................................................................... 33

5. SCENARIO MODELLING IN FORESCENE .........................................................34

5.1. Need for a framework or ‘meta-model’............................................................................. 34

5.2. Short presentation of Bayesian Networks....................................................................... 34

5.3. Preliminary model structure for FORESCENE................................................................ 36 5.3.1. Problem fields/environmental pressures ...................................................................... 36 5.3.2. Activities .......................................................................................................................... 37 5.3.3. Cross-cutting driving forces ........................................................................................... 37 5.3.4. Goals ............................................................................................................................... 37 5.3.5. Key strategies ................................................................................................................. 37 5.3.6. Submodeliing systems ................................................................................................... 37

5.4. Uncertainties......................................................................................................................... 39

6. CONCLUSIONS ...................................................................................................40

7. REFERENCES .....................................................................................................41

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

2

List of Figures

Figure 1: Scenario elements systematics in FORESCENE’s scenario construction ... 5

Figure 2: Overview of the approach in WP1 and WP2 ............................................... 6

Figure 3: Mapping of the cross-sectoral and multi-beneficial sustainability strategies ...................................................................................................17

List of Tables

Table 1: Overview of the three levels of underlying factors .........................................8

Table 2: Analysis of underlying drivers for the three environmental topics (resource use and waste, water and water use, and landscape,

biodiversity and soils) .................................................................................10

Table 3: Sustainable goal references for the FORESCENE project ..........................13

Table 4: Sustainable goal references for the three topic areas at different scale levels .................................................................................................14

Table 5: Summary of scenario models and their relation to FORESCENE................33

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

3

E X E C U T I V E S U M M A R Y

This technical report of Work package (WP) 3 in the FORESCENE project describes

the basis for the scenario construction and modelling in the project. FORESCENE aims

to develop an analytical framework for consistent environmental sustainability scenario

building. Preceding WP 1 and 2 had identified cross-cutting key driving forces for the

environmental problem areas “biodiversity, soil and landscape”, “water” and “resources

and waste”, and have described goals and sustainable strategies for relevant activity

fields (agriculture, land use/infrastructure, industry/economy). The aim of WP 3 is to

analyse the options for parameterization and simulation/modelling. It takes up the key

sustainability elements from earlier WPs, and combines them to preliminary narratives

which can be further integrated to comprehensive scenarios. A literature review is per-

formed to describe earlier scenario and modelling work relevant for the FORESCENE

scenario framework. As existing models use to have a more narrow scope, it is con-

cluded that there is a need for a meta-model which allows to model the broader per-

spective of FORESCENE while allowing to include knowledge and data derived from

specific models. This technical report presents the development of the FORESCENE

model framework describing the main elements and outlining the structure of the major

submodules.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

4

1 . I n t r o d u c t i o n

This technical report of Work Package 3 in the FORESCENE project describes the basis for the modelling and scenario construction in the project. FORESCENE aims to develop an analytical framework for consistent environmental sustainability scenario

building. The project seeks to identify elements of a desirable future and it uses back-casting techniques in order to contribute to the identification of strategies needed to reach different sustainability goals.

In Work Package 1, FORESCENE reviewed past and ongoing research projects and

policy frameworks with regard to the three particularly focused areas ‘water’, ‘biodiver-sity, soil and landscape’, and ‘resource use and waste’. Key drivers behind environ-mental problems in these areas were identified, with a particular focus on cross-cutting drivers that influence several problem areas. In Work Package 2, the focus was put on

the concerned activity fields (economic sectors, policy fields): agriculture, infrastruc-ture/land use, and industry/economy. The aim was to define essential elements of sus-tainable development for these activity fields through answering what the desired future should look like. In a backcasting fashion, FORESCENE then aimed at finding which

cross-sectoral measures could be expected to exert a multi-beneficial impact over the environmental fields considered, in the perspective of reaching the previously defined sustainability goals.

The objective of Work Package 3 is to consistently combine and expand selected sce-

nario elements (i.e. driving forces and environmental pressures from WP1, and sustainability strategies and goals from WP2) into preliminary scenario narratives. Those preliminary narratives are also scenario elements, but of a higher level of inte-gration than the single 'puzzle pieces' of specified key elements delineated in WP1 and

WP2. The next step, which will occur in WP4 and WP5, aims to further develop, com-bine and quantitatively parameterize the preliminary narratives in order to develop and test business-as-usual and alternative scenarios, using existing models or additional modelling frameworks developed in FORESCENE. It is expected, in the end, to give

insight into effective integrated approaches towards sustainability.

The present report has therefore two goals. First, in chapter 2, the key scenario ele-ments from WP1 and WP2 are filtered out in order to select and thoroughly define con-sistent clusters of drivers, pressures, strategies and goals. Based on this selection,

more integrated scenario elements are developed in the form of preliminary narratives in chapter 3.

Second, a review of existing models and scenarios relevant to FORESCENE's topics is undertaken in chapter 4. The criteria for the review shall allow a comparison between

existing modelling and scenario frameworks and FORESCENE's broad coverage of environmental issues. The basis for the comparison, in terms of the driving forces, en-vironmental issues, and sustainability strategies and goals to be modelled, is described in chapter 2 and 3. Chapter 5 draws the consequences from the review in the preced-

ing chapter for the development of FORESCENE's modelling framework. That is to which extent existing models can be used for FORESCENE's purposes or there is a need to develop additional modelling frameworks in FORESCENE.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

5

2 . S c e n a r i o E l e m e n t S y s t e m a t i c s i n F O R E S C E N E

In WP1 and WP2, a participative process of stakeholder involvement (1st and 2nd inte-gration workshops) and expert assessment (three targeted workshops) has resulted in the delineation of the cross-cutting drivers influencing the three environmental topics at stake, and the elicitation of sustainability goals and possible strategies to reach them.

Figure 1 shows the scenario element systematics adopted in FORESCENE for sce-nario construction. The results from WP1 and WP2 represent single key scenario ele-ments derived from a systemic perspective. In this chapter some of these elements will be selected for their potential in scenario construction and further modelling. They will

be refined with the objective to combine them in the next chapter into preliminary narra-tives, which represent certain combinations of those scenario elements.

Figure 1: Scenario elements systematics in FORESCENE’s scenario construction

Before that, the research process undertaken in WP1 and WP2 shall be briefly re-minded. Figure 2 depicts in a nutshell the path followed in WP1 and WP2. A combina-tion of the concept of socio-industrial metabolism and the EEA’s DPSIR framework

Single key combined scencario elements scenario elements

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

6

provided a framework for the analysis of the three topic areas: ‘resource use and waste

generation’, ‘water and water use’, and ‘landscape, biodiversity and soils’. The combi-nation of these approaches resulted in a comprehensive systems analysis tool, which allowed quantification of the interaction between the anthroposphere and the environ-ment (socio-industrial metabolism), as well as a qualitative evaluation of the results of

this interaction (DPSIR).

Figure 2: Overview of the approach in WP1 and WP2

In WP1, the relevance of the influence of a number of underlying factors on each of the three environmental themes has been assessed in the context of eleven economic activities. The relevance analysis is based on stakeholders' and experts’ views, pub-lished data and literature. The results of the relevance study was compiled in a matrix

form, with each couple 'underlying factor'—‘activity’ classified as very relevant, relevant or not relevant, depending whether, respectively, a direct, indirect or no link between one underlying factor and the pressures related to a given activity was assessed. A systematic ‘scoring’ method was then conducted in order to assess the relative impor-

tance of the activities and underlying factors. A score was assigned to each couple ‘underlying factor’— ‘activity’ according to its relevance classification and its cross-

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

7

cutting character over the three environmental topics. Finally, adding the scores of the

‘underlying factor’— ‘activity’ couples row and column wise gave insight into the overall relevance of a given underlying factor regarding all activities and environmental themes, and into the overall importance of a given activity regarding all underlying fac-tors and environmental themes, respectively. The scoring results were given colour

codes to ease the interpretation of the results of the relevance and cross-cutting analy-sis.

In WP2, the expert workshops followed by the 2nd Integration Workshop delivered a number of items for the core elements of sustainability scenarios ‘sustainable goal ref-

erences’ and ‘sustainability strategies’. In the backcasting approach adopted at this stage of FORESCENE, the latter shall enable our society to proceed on the way to-ward the former. The goals, which remained at this stage quite broadly formulated, were divided into environmental and socio-economic objectives. Twenty-five sustain-

ability strategies were defined and grouped into seven sets. With regard to their poten-tial role for a sustainable future, the strategies are expected to have cross-sectoral and multi-beneficial characteristics. The former refers to elements which can be related to the three activity fields delineated for WP2 (agriculture, industry/economy, infrastruc-

ture/land use). The latter corresponds to strategies which foster improvement towards several of the sustainable goal references. Based on the descriptions provided by the workshops, the strategies are further classified with regard to the level of specification (low, medium, high) which determines the probability for direct operationalization, as

the inverse of the effort expected for it (a low level of specification will require more detailed information before the strategy can be translated into effective measures). The second dimension in the classification reflects the state of the consensus with regard to the proposed strategies, from more controversial to more or less undisputed ap-

proaches. It should be noted that there was an agreement in the workshops that all strategies listed are important, but for some a common understanding was hampered through lacking degree of concreteness, which indicates that level of specification and status of consensus are not independent.

The workshops not only contributed to identification of cross-sectoral sustainability elements but also sought to determine measurable indicators for each of the sustainabil-ity elements identified. The indicators were viewed as Level 3 in the identification of underlying factors. This identification of indicators is in FORESECENE viewed as the first step in developing scenarios; the indicators can be used as parameters in the models. Table 1 presents an overview of the results of the analysis of underlying with various indicators for level 2 connected to the problem areas. The indicators are listed for the problem areas landscape, biodiversity and soils, resource use and waste, and water and water use. The indicators have then been further classified according to the DPSIR framework since they may play different roles in constructing and operationalis-ing a potential scenario.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

8

Table 1: Overview of the three levels of underlying factors

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

9

2.1. Cross-cutting driving forces: economic activities and underly-

ing factors

A major aim of WP1 was to determine cross-cutting drivers that together influence the three problem areas) for the use of the subsequent work packages. Table 2 presents e.g. the results of the relevance analysis. ‘XX’ means that a Level 2 underlying factor has been classified as very relevant, and has a direct effect on the pressures and the resulting impacts due to a given activity on a given environmental topic. According to

these results from the evaluation of relevance of the participative stakeholder work-shop, energy supply, agriculture, water supply and construction appeared to be the activities most susceptible to cause pressures and impacts on the three environmental themes. Transport, forestry, chemicals, basic metals, and food products were also ac-

tivities or product groups potentially important to consider.

The underlying factors were sorted into five categories: economic development, pro-duction patterns, consumption patterns, demography and natural system. The catego-ries ‘production patterns’ and ‘economic development’ were the groupings of underly-

ing factors that achieved the highest scores in the evaluation. The factors under pro-duction patterns (‘material intensity’, ‘composition of material input’, ‘innovation’ and ‘recycling’) are all among the most powerful underlying factors. They all have a strong, direct and cross-cutting influence on the most important activities in relation to the

three topic areas. ‘Globalisation’, ‘economic growth’ and ‘investment patterns’ have also considerable influence, but only cross-cutting environmental effects within a more limited number of activities.

The underlying factor categories ‘natural system’ and ‘consumption patterns’ follow

‘production patterns’ and ‘economic development’ in the ranking. For 'natural system', ‘depletion of resources’ and ‘climate change’ were the most relevant underlying factors, and ‘food and drink’ and ‘transport and communication’ were the most important under 'consumption patterns'. ‘Natural system’ and ‘consumption patterns’ are more indirect

drivers in nature but according to the relevance analysis they are important and should not be neglected, particularly in connection with agriculture, construction, energy and water supply, and transport. In relation to these activities, these underlying factors have a considerable cross-cutting influence.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

10

Table 2: Analysis of underlying drivers for the three environmental topics (resource use and waste, water and water use, and landscape, biodiversity and soils)

The preliminary narratives which will be presented in this report are mostly based upon drivers connected to “production patterns” and “consumption patterns”, but they are

also strongly based on economic development factors that set frames, have to be mo-bilized and may be strongly effected by described developments. Economic growth is central for setting frames for all kinds of societal development.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

11

Three narratives are based in changes of production patterns toward higher resource

productivitiy, changed agricultural production patterns and radically increased use of renewable energy, respectively. Higher resource productivity is connected to all groups of factors under “production factors”, including material intensity, composition of mate-rials, recycling and innovation. This narrative is also closely related to economic devel-

opment factors such as investment patterns and have implications for globalization. The other two have mostly connection with land use, agriculture and forestry.

The remaining two narratives are based upon changes in consumption patterns toward increased service demand and vegetable diet, respectively. They have connections

with production patterns, particularly related to food and agriculture and the service sector, and economic developments such as investment patterns and international trade of goods and services.

One narrative is based on climate change mitigation and development of renewable

energy, which has strong implications for landscape and nature. Otherwise the narra-tives do not take a direct starting point in demography or natural system and these are only addressed indirectly in the narratives. Demographic factors such as age and population patterns may be of great relevance for consumption developments. Differ-

ent responses to ageing and population development problems would have a possible and perhaps interesting option for a narrative, but it was considered difficult to connect to the resource related themes of FORESCENE. Climate change may have far-reaching cross-cutting effects toward the mid-21st century, but since effects are very

uncertain and the aim of narratives and scenarios are to analyze various policy options, it seemed most promising to concentrate on the issue of biofuels.

2.2. Desirable sustainability goal references

In the European context, EU’s key objectives for sustainable development, which are

summarized below constitute the main guidance and these objectives also provide the major goals for FORESCENE. The fundamental objectives constitute:

1. Environmental protection Safeguard the earth's capacity to support life in all its diversity, respect the limits of

the planet's natural resources and ensure a high level of protection and improve-ment of the quality of the environment. Prevent and reduce environmental pollution and promote sustainable consumption and production to break the link between economic growth and environmental degradation.

2. Social equity and cohesion

Promote a democratic, socially inclusive, cohesive, healthy, safe and just society with respect for fundamental rights and cultural diversity that creates equal oppor-tunities and combats discrimination in all its forms.

3. Economic prosperity

Promote a prosperous, innovative, knowledge-rich, competitive and eco-efficient economy which provides high living standards and full and high-quality employment throughout the European Union.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

12

4. Meeting EU’s international responsibilities

Encourage the establishment and defend the stability of democratic institutions across the world, based on peace, security and freedom. Actively promote sustain-able development worldwide and ensure that the European Union’s internal and ex-ternal policies are consistent with global sustainable development and its interna-

tional commitments.

EU’s specific strategic sustainability objectives constitute:

1. Climate change and clean technology

To limit climate change and its costs and negative effects to society and the envi-ronment. Kyoto Protocol commitments of the EU-15 and most EU-25 to targets for reducing greenhouse gas emissions by 2008 – 2012, whereby the EU-15 target is for an 8% reduction in emissions compared to 1990 levels. Aiming for a global sur-

face average temperature not to rise by more than 2ºC compared to the pre-industrial level.

2. Sustainable transport To ensure that our transport systems meet society’s economic, social and environ-

mental needs whilst minimising their undesirable impacts on the economy, society and the environment. Thus, decoupling economic growth and the demand for transport with the aim of reducing environmental impacts.

3. Sustainable consumption and production

To promote sustainable consumption and production patterns. Promoting sustain-able consumption and production by addressing social and economic development within the carrying capacity of ecosystems and decoupling economic growth from environmental degradation.

4. Conservation and management of natural resources To improve management and avoid overexploitation of natural resources, recogniz-ing the value of ecosystem services. Improving resource efficiency to reduce the overall use of non renewable natural resources and the related environmental im-

pacts of raw materials use, thereby using renewable natural resources at a rate that does not exceed their regeneration capacity. Improving management and avoiding overexploitation of renewable natural resources such as fisheries, biodiversity, wa-ter, air, soil and atmosphere, restoring degraded marine ecosystems by 2015 in line with the Johannesburg Plan (2002) including achievement of the Maximum Yield in

Fisheries by 2015. Halting the loss of biodiversity and contributing to a significant reduction in the worldwide rate of biodiversity loss by 2010.

5. Public Health To promote good public health on equal conditions and improve protection against

health threats.

6. Global poverty and sustainable development challenges To actively promote sustainable development worldwide and ensure that the Euro-pean Union’s internal and external policies are consistent with global sustainable

development and its international commitments.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

13

7. Social inclusion, demography and migration

To create a socially inclusive society by taking into account solidarity between and within generations and to secure and increase the quality of life of citizens as a precondition for lasting individual well-being.

These EU objectives formed the basic starting point for the definition of goal references within FORESCENE. The EU objectives are sometimes not very well-defined. Their priority and balance is not always clear and they have to be translated to more con-crete terms, strategies and actions.

Table 3: Sustainable goal references for the FORESCENE project

Table 3 presents an overview of the sustainable goal references condensed from the outcome of the workshops. In this general formulation, the goals listed are relevant for a large span of economic activities and environmental problem fields. These goal refer-ences were further elaborated and separated in three levels. Table 4 summarizes this

concretization of the goal references.

The preliminary narratives are linked to many of the goals above. The production effi-ciency and changed consumption toward services are closely related to many of the resource and material flows, while the other narratives; changed diet, liberalisation of

agriculture and climate change mitigation have closer connection to agricultural and landscape development goals. The Climate change mitigation narrative is based upon the global climate goal of minimizing temperature but through the emphasized mitiga-tion strategies the sketched development may challenge land use and water goals

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

14

Table 4: Sustainable goal references for the three topic areas at different scale levels

Level Resource use and Material flows

Land use and biodiversity Water management

Global - No increase of global tem-

peratures by more than 2°C (mean annual temperature) by the year 2050 compared to today’s level => requiring a reduction of greenhouse gas emissions by industrial-ised countries of up to 80%

- Prevent further loss of biodiver-sity

- Balance of demand and supply of water

- Maintain and enhance regulation services; water quality and quantity

EU - Absolute decoupling of

resource use and GDP - Maximal individual happi-

ness/well-being - Resource extensive life-

styles, production and con-sumption patterns

- Resource self-sufficiency of Europe

- Low impact per consump-tion unit

- Transport culture based on proximity, slowness and suf-ficiency

- Reduce use of non-renewable resources

- Welfare more in the sense of happiness, reduction of social deprivation, and edu-cation

- Competitiveness

- Shift from non-renewable resources to renewable ones while not increasing use of biomass

- Zero emissions of hazard-ous substances - reduce pollution below critical loads

- Prevent further loss of biodiver-sity

- Maintain or restore eco-system services

- Large and growing biodiversity within agro- and urban ecosys-tems

- Preserved natural and semi-natural ecosystem enclaves

- Optimal multifunctionality of land use

- Overall low volume of transport - Minimise urban sprawl by using

brownfield resource - Multifunctionality - Inter- and intraregional Diversity - Minimise nutrient losses - Food production of good quality

- Balance of demand and supply of water

- Maintain and enhance regulation services; water quality and quantity

Regional - Overall low volume of trans-

port - Transport culture based on

proximity, slowness and suf-ficiency

- Minimize urban sprawl by using brownfield resource

- Prevent further loss of biodiver-sity

- Maintain or restore eco-system services

- Large and growing biodiversity within agro- and urban ecosys-tems

- Preserved natural and semi-natural ecosystem enclaves

- Optimal multifunctionality of land use

- Minimize urban sprawl by using brownfield resource

- Balance of demand and supply of water

- Maintain and enhance regulation services; water quality and quantity

It will not be possible to model all of the goals listed above in the context of FORES-CENE. For example, the goal of competitiveness for the EU can of course be used in

scenario narratives but it will not be modelled quantitatively. It would require the use of econometric models which exceed the scope of FORESCENE. The modelling of hap-piness and well-being will similarly remain at the qualitative level allowed by scenario narratives.

On the other hand, some targets mentioned above are already quantitatively set. In connection with climate change mitigation, the emissions of GHG which according IPCC studies should be cut by 80% in 2050 in the industrial countries. Some other tar-get values have also been suggested by other bodies (e.g. 50% reduction target from

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

15

the G8) which could alternatively be used for assessing the progress of the EU towards

sustainability.

Some goals related to resource use and waste generation can be translated in con-crete and quantitative terms using existing material flow indicators, such as total mate-rial requirement (TMR). An ambitious target for TMR could be a 80% reduction by

2050. Because TMR is a very good indicator for the physical basis of a society, such a change would certify that the EU has adopted a resource extensive way-of-life. Setting targets to the different components of the TMR would also cover a number of the goals listed in the first column of table 4. For example, the ratio TMR non renewables to bio-

mass should not increase. The ratio TMR foreign to domestic should not increase ei-ther if the issue of problem shifting is to be efficiently addressed.

2.3. Key sustainability strategies

The FORESCENE workshops also resulted in sets of overall sustainability strategies

expected to have cross-sectoral and multi-beneficial characteristics. As a result of the discussions at the 2nd Integration Workshop, the defined twenty-five sustainability strategies were grouped into seven sets as derived from:

• Improving orientation and target setting

• Improving information and decision processes

• Improved planning

• Changing use of capital

• Changing environmental performance of production and consumption

• Improvement of product management and procurement

• Improving state finance and social security systems.

The environmental goals in Tables 3 and 4 are connected to the strategy set. Figure 3 presents a mapping of the strategies with reference to level of consensus and specifi-cation. Most of the strategies from the set "Changing environmental performance of production and consumption", which is closely related to the goals in Tables 3 and 4 are found in the top right corner, which translates to high levels of both consensus and specification. The general strategies must however be operationalized through more specific ones. Energy/resource efficiency through innovation can be, for example, be promoted via an ecological tax reform shifting labour taxes to energy and natural re-sources. This can be further targeted to discourage consumption related to greenhouse emissions or encourage use of renewable energy. Strategies in the lower parts of the figure are characterized by important disagreements and often address sensitive as-pects of the socio-economic sustainability. Food production within the EU may reduce transport and contribute to the security of food supply, but may also deprive less de-veloped countries of an opportunity to develop through international trade. A similar case is arising in connection with bioenergy.

The preliminary narratives started with strategies that are characterized by high levels of consensus. Today, for instance, there is growing support for increasing resource productivity, and the use of renewable energy. However, there may be important dis-

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

16

agreements concerning the appropriate measures under these strategies, for instance

about the relation of intra vs. intersectoral change, and the potential contribution of biomass to the renewables. This leads to the consideration of narratives which are more controversial, involve higher risks of trade-offs and problem shifting, and could be used to model worse case development compared to the baseline and more sustain-

able alternatives. Such a preliminary narrative concerns the opening up the European food market for foreign producers with decreasing production subsidies which con-fronts the strategy of localising markets and could have side effects in the EU and be-yond.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

17

Figure 3: Mapping of the cross-sectoral and multi-beneficial sustainability strategies

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

18

3 . P r e l i m i n a r y n a r r a t i v e s i n F O R E S C E N E

The identified scenario elements form the basis for the scenario construction and mod-eling activities in FORESCENE. Scenarios are stories or `snapshots' of what might be. Decision makers use them to evaluate what to do now, and what the options are, based on different possible futures. The options for the future reflect either an extrapo-lation of current trends or introduced changes, such as policies and management plans. Modelling is the activity where a model is formulated, which is a way of creating a sim-plified and workable representation of a real-world problem. A model is designed for a specific purpose. A model can never be totally ‘right’ – it is an indicative instrument only. However it should be plausible and rigorous enough to give decision support. Scenarios generally have a qualitative component - the narrative - and a quantitative component - the numbers that illustrate and support the story. Different scenario analyses will re-quire different balances between narrative and number.

In the FORESCENE project the aim is to develop a framework for scenario modelling purposes as a decision support tool. The constructed scenarios must be scientifically sound and quantitative and meet the practical needs of policy makers. All scenarios will have a defined temporal scale. The major time frames consist of middle-term (2015-2030) and long-term (2050). The spatial scale for the FORESCENE scenarios are in-

tended to mainly focus at the EU level and EU’s exchange with the surrounding world, but will be broken down to subregions of the EU as well.

In the previous chapter, the key scenario elements from WP1 and WP2 have been picked up and further described in the perspective of using them for scenario construc-

tion and modelling in FORESCENE. In the present chapter, these 'puzzle pieces' will be assembled in several preliminary scenario narratives. Basically, the procedure is as follows: one selects one key sustainability strategy deemed relevant in the previous chapter and describes, in qualitative terms, its expected influence on the cross-cutting

drivers, which in turn will impact the pressures on the three environmental topics in a way that may or may not be satisfactory with regard to the sustainable goals. The qualitative narratives should also give hints regarding the parameters and indicators that may be used in later work packages to translate them into quantitatively observ-

able terms.

Though these narratives are of a higher level of integration than the elements pre-sented in the previous chapter (see figure 1), they are not yet fully fletched scenarios. But they are built around a “what if?” kind of question which gives room to further inte-

grate the narratives with one another in order to develop consistent scenarios. This, and the quantitative modelling part, is the task of WP4 and WP5.

Preliminary narrative 1: Increased resource productivity

A both high consensus and high specification key strategy (see section 2.3) is that of increasing material, energy and water efficiency. The production system would be the target of this strategy that may be fostered through economic instruments such as an economic tax reform, increased R&D investment, material efficiency programmes for

manufacturing etc. Such instruments have been identified as important underlying fac-tors in WP1 (see section 2.1). They are deemed necessary to consider for their poten-

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

19

tial influence on economic activities and their associated environmental pressures. The

quantitative modelling, however, would require the adequate use of econometric mod-els. Unless such models are accessible and can be integrated in the FORESCENE framework, the unfolding of the economic instrumentarium will have to remain at the narrative level.

The outcome of these strategically applied measures would be an overall increase in material productivity in the production of goods. The amount of nonrenewable re-sources, such as metals, industrial minerals and construction minerals, mobilised per monetary unit of final demand (whether domestic or for exports) would decrease. This

would possibly also have a positive effect on the energy and water productivity. The more efficient use of materials in goods production can be expected to lead to an in-crease in energy and water productivity as well.

At the end of the causal chain, the consequences of a resource productivity increase

would have a positive influence for the sustainable goals related to the use of non re-newables (minerals and fossil fuels), which would in turn reduce waste generation and greenhouse gases emissions. A water productivity increase would as well be expected to improve the water balance. FORESCENE should be capable to mode those effects,

and it should also be possible to estimate the changes in the total material requirement of Europe, and whether its domestic and foreign components remain balanced.

Preliminary narrative 2: Changed consumption pattern towards service economy

The habits of European consumers could be modified in the direction of more service demand, while overall consumption volume in terms of available income spent grows unchanged with the trend of GDP. The overall European economy could actually move towards a service economy, which would imply that the EU would export a higher share of services. While there is a high consensus concerning the goal of strengthen-ing the development of the service economy, there may be important ideological con-flicts on many of possible means for bringing about consumption changes.

The hypothesis is that services show a higher resource productivity than goods. A higher share of services in the final demand would therefore lead to a – at least rela-tively – lower resource demand. The expected consequences for the environmental pressures would be similar in qualitative terms as those mentioned in the first prelimi-nary narrative. The use of non renewables, waste generation, TMR, greenhouse gases emissions, water balance etc are likely to evolve in quantitative terms towards the sus-tainable goals. Comparing the modeling results of the preliminary narrative 2 to those of the preliminary narrative 1 would give insight into the respective potentials of the strate-gies around which they are respectively built. By preserving the level of consumption while decreasing the environmental impacts associated with resource use, one could also assume that within the wealthy EU the population would experience higher levels of happiness. It is however most probably not possible to model this aspect quantita-tively in the ongoing project. Instead, the FORESCENE modelling may concentrate on modelling the effects of higher share of service consumption on the environmental pressures quantified.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

20

Preliminary narrative 3: Changed diet

Changing the average European diet toward a more vegetable based food consump-tion is also about changing consumption patterns. To change food consumption via direct state intervention is probably very controversial, even if such a change may seem very desirable from a resource consumption point-of-view since animal produc-

tion demands much higher inputs of energy, nutrients and land resources. Many stud-ies on food and nutrient flows in industrial societies argue for a changed diet as the perhaps most obvious option for an important decrease of resource consumption and losses of nutrients in all steps of the food flows. This would reduce the pressure to

natural land conversion for feed production, and the need of long-distance imports of animal feed, and could mitigate the GHG emissions of cattle breeding, thus considera-bly decreasing the ecological foot-print of European food consumption. Such a change could be achieved by a combination of economic instruments bringing a changed price

relationship between vegetable and animal products, and changed attitude toward put-ting higher value on vegetable diets. The latter can be supported by health arguments and may be inspired by low animal diets in e.g. Asia or Latin America. The modelling in FORESCENE would have to reflect relevant changes with regard to changes of envi-

ronmental pressure to the intra EU environment due to changes of the structures of agricultural production in Europe as well as the impacts of changed agricultural imports on extra EU land use and resulting pressures.

Preliminary narrative 4: Climate change mitigation: increased use of biofuels

Responding to the threat of global climate change seems to become one of greatest challenges for Europe in coming decades. In the few last years consensus has devel-oped that the EU needs to take the lead for making it possible for the world to stabilize

and later reduce emissions of greenhouse gases. An ambitious European response will focus on both reducing fossil energy consumption through efficiency increase (prelimi-nary narrative 1) and increased use of alternative energy fuels. Two key targets have already been set by the European Council for the year 2020: 20% reduction of green-

house gas emissions and a 20% share of renewable energy (from 8,5% in 2007). These targets are likely to be a first step toward at 50% reduction and perhaps 50% renewable energy by 2050, which can be considered a minimum of necessary Euro-pean efforts for the realization of the temperature goal of not exceeding 2o warming by 2050.

Biofuels may contribute a certain share to GHG mitigation. However, additional land requirements, particularly for the provision of growing imports, may result in counter-productive effects, i.e. increased GHG emissions and pressures on biodiversity through

the conversion of tropical forests and savannahs. The global land use requirements of the EU to supply the demand for agricultural goods are determined by the the intensity and productivity of cultivation within and outside the EU. The FORESCENE modelling should reflect the implementation of the currently invisaged biofuel quota for the EU

(10% in 2020), and the resulting implications on domestic environmental pressures and impacts on the foreign environment.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

21

Preliminary narrative 5: Liberalisation of commercial agriculture

The Common Agricultural Policy (CAP) was traditionally based on production support, which resulted in overproduction of many agricultural in Europe, which was delivered to the world with support of export subsidies. An abolition of production subsidies to the agricultural sector has been discussed intensively for the past two decades and CAP

has been gradually transformed towards an emphasis on other types of support, particularly toward landscape conservation and rural development. However, production support is still important, although it has changed in many ways, e.g. from production to areal support. With increasing global pressures on EU, e.g. in connection

with negotiations for a new international trade treaty, to abolish agricultural subsidies and allow products from other regions to compete on equal terms on the European market, one potential development route may foresee that EU will continue to reduce production support. However, since many large member countries are opposed to this

development, it can hardly be described as a consensus strategy, also before the background of social development of rural regions. A profound liberalisation of commercial agriculture with maintained support for land-

scape conservation in marginal areas would have important many-sided structural ef-fects on European agriculture and landscape development. This would lead to in-creased concentration of agricultural production to regions that are able to compete on a globalizing market. In such competitive regions, farming will be more specialized,

industrialized and intensified putting extra pressure on scarce water resources and biodiversity. In major parts of Europe, however, agriculture is likely to become extensi-fied and depend on support on landscape conservation. The future of the European agriculture would increasingly depend on the development of the world market. How

agriculture will develop in various individual regions may depend on both traditional specialization, organization and competitiveness of local agriculture and food industry. Depending on the input from agro-economic models, an advanced modelling effort may be able to identify production areas that will be most affected by abolition of subsidies

and show some of the effects on land use and water by intensification or extensifica-tion. However, because of the complex uncertainties surrounding the development of the global market for agricultural market, the development of individual products and agricultural regions quantitative modelling will be difficult..

A possible way for a structured move from the preliminary narratives developed above

towards the final alternative scenarios (to come in WP4 and WP5) is to introduce a similar procedure that was used in the model “Limits to growths - A 30 year update” In that model, the scenario starts with one main element and new elements are succes-sively added,. In the case of FORESCENE, the elements to be combined are the pre-

liminary narratives. For example :

Scenario 1.1: Increase resource productivity

Scenario 1.2: Increase resource productivity and changed consumption pattern

Scenario 1.3 Increase resource productivity and changed consumption pattern and diet

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

22

Scenario 1.4 Increase resource productivity, changed consumption pattern and diet,

and increase of biofuels

Scenario 1.5 Increase resource productivity, changed consumption pattern and diet, and increase of biofuels and liberalisation of commercial agriculture

The last scenario would be the closest to a final integrated alternative scenario. It would have to specify the conditions under which certain elements such as the in-crease of bioenergy and liberalization of agriculture markets would be beneficial, thus probably describing the features of a “balanced” increase of biofuels and the degree of

liberalization which seems appropriate.

The final scenarios might in the end contain less elements than the sum of the prelimi-nary narratives because of the impossibility to model them all. But it should also allow a more detailed description of the interactions between the scenario elements.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

23

4 . R e v i e w o f r e l e v a n t s c e n a r i o s t u d i e s a n d s i m u l a -

t i o n m o d e l s i n c o m p a r i s o n w i t h F O R E S C E N E

The following section is a review of selected scenario modelling studies with relevance for the FORESCENE scenario construction. It mostly includes scenario-based studies with similar thematic aims and scope as FORESCENE. The review primarily assesses

how selected scenario studies and models deal with the three environmental topics of FORESCENE, but also what problem area, cross-cutting drivers, scope and time hori-zon have been used in these studies. Most existing models have been developed to deal with rather specific aspects of sustainability, whereas FORESCENE adopts a much broader perspective, where the aim is to integrate environmental problems and

activities which have been treated separately so far, and with an extended spatial and temporal perspective.

4.1. Global and regional scenario studies of particular relevance to FORESCENE

A number of large scenario projects have during the last decade been conducted at the global level aimed at unravelling the impacts of human activities on natural systems. The most well-known examples include the IPCC Assessment (IPCC, 2000), the Global Environment Outlook (UNEP, 2002) and the Millennium Ecosystem Assessment

(MEA, 2005). These global assessments provide in general too little detail at the re-gional level for making concrete contributions to a regional analysis. Global land use change assessment are largely inadequate since most processes influencing global change are the result of decisions and changes at the local scale that most often are

poorly represented in global scale assessments (Houghton, 2003; Ellis, 2004). Thus, global studies have therefore often limited direct relevance to support national or Euro-pean policy and planning. Regional scenario assessments at the EU level has a more direct relevance for FORESCENE. However, global thematic scenarios may still be

valuable for providing an international context of future development within the EU, and are thus of some relevance for the scope of FORESCENE. For FORESCENE, the scenario construction by Millennium Ecosystem Assessment at the global level and Global Scenario Group with both global and regional scenarios have provided valuable

inspiration. Among recent European studies, the ATEAM and EURURALIS projects have been particularly valuable, and the MOSUS and Prelude projects in terms of land use approaches.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

24

Box 1. Examples of recent important global and regional scenario studies

Box 2. Scenario links

4.1.1. Millennium Ecosystem assessment scenarios

This project developed its first global scenarios in 1996-97, which have later been up-

dated. The project has the developing world as the primary focus and the scenarios include both exploratory and normative approaches. Exploratory stands for adding quantitative rigor to scenarios through the use of global models and other systematic numerical approaches, and thus possible future pathways. The Millennium Ecosystem

Assessment Scenarios include the following set of four scenarios (http://www.millenniumassessment.org/en/Scenarios.aspx):

• The World Futures Studies Federation • The Global Business Network (GBN), • The Millennium Project at the American Council for the United Na-

tions University. • The scenario page at Royal Dutch/Shell. • Le Laboratoire d’Investigation en Prospective, Strategie et

Organisation (LIPSOR). • The International Institute for Applied Systems Analysis (IIASA). • The Futures Group, • The EU’s information portal for environmental scenarios and pro-

spective studies. • The Society for International Development (SID)’s Future Searches

programme. • Global Multi-Thematic Scenarios • Barry Hughes’ International Futures (IFs) • The Millennium Institute’s Threshold 21 • UNEP’s Global Environment Outlook • The Global Scenario Group and the PoleStar Project • The Great Transition Initiative • RIVM’s IMAGE model

Global Thematic Scenarios

• The IPCC’s Special Report on Emissions Scenarios • The FAO’s World Agriculture: Toward 2015/2030. An FAO Perspec-

tive • The FAO’s Global Fibre Supply Study and other forestry outlook

studies • Long-Term Scenarios of Livestock-Crop-Land Use Interactions in

Developing Countries, prepared for the FAO by A.F. Bouwman

Regional Multi-Thematic Scenarios

• African Futures - National Long Term Perspectives Studies • NIES’s Asian-Pacific Integrated Model (AIM)

• Five Scenarios for Europe (Europe 2010) by the Forward Studies Unit of the European Commission

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

25

Global Orchestration -This scenario depicts a globally connected society that

focuses on global trade and economic liberalization and takes a reactive ap-proach to ecosystem problems but that also takes strong steps to reduce pov-erty and inequality and to invest in public goods such as infrastructure and edu-cation. Economic growth in this scenario is the highest of the four scenarios,

while it is assumed to have the lowest population in 2050;

Order from Strength – This scenario represents a regionalized and fragmented world, concerned with security and protection, emphasizing primarily regional markets, paying little attention to public goods, and taking a reactive approach

to ecosystem problems. Economic growth rates are the lowest of the scenarios (particularly low in developing countries) and decrease with time, while popula-tion growth is the highest

Adapting Mosaic – In this scenario, regional watershed-scale ecosystems are

the focus of political and economic activity. Local institutions are strengthened and local ecosystem management strategies are common; societies develop a strongly proactive approach to the management of ecosystems. Economic growth rates are somewhat low initially but increase with time, and population in

2050 is nearly as high as in Order from Strength.

TechnoGarden – This scenario depicts a globally connected world relying strongly on environmentally sound technology, using highly managed, often en-gineered, ecosystems to deliver ecosystem services, and taking a proactive ap-

proach to the management of ecosystems in an effort to avoid problems. Eco-nomic growth is relatively high and accelerates, while population in 2050 is in the mid-range of the scenarios.

All these scenarios point at the importance of maintaining ecosystem services.

FORESCENE is emphasizing biodiversity, which may be used as a proxy for ecosys-tem services, but differs in terms of trade-offs between gains in provisioning services and the potential losses of other services or functions. The MEA also points at 2050 as a point of reference, which will be the same in the FORESCENE project. From the

MEA project one may also conclude that it is important to relate the EU perspective to the surrounding world.

4.1.2. The Global Scenario Group

The Global Scenario Group developed during the 1990s a set of global and regional scenarios. It consisted of six global scenarios; three main scenario types in two vari-

ants:

• Conventional Worlds (Policy Reform and Market Forces),

• Barbarization (Fortress world and Breakdown)

• Great Transitions (Eco-communalism and New sustainability paradigm).

These scenarios were compared to a reference scenario for the period 1995-2050, the same end year that has been chosen for the FORESCENE scenarios. Global popula-tion, food requirements, agricultural output, agricultural, forest, cropland and pasture-land areas, requirements for energy, water and resource requirements including re-

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

26

source use efficiency were focused by the scenarios. Global carbon dioxide emissions

were used as a proxy for global climate change. The issue of water sufficiency and quality were also given special attention, similar to the FORESCENE focus on water.

The GSG project also developed some alternative scenarios:

• Market forces – a scenario, which is mainly supply-demand-based, which

may be analogous with a “business-as-usual” scenario. The “free market” corrects for inefficiency and thus mitigates the environmental crisis;

• Policy reform – a scenario that maintains the essential assumptions of the reference scenario paradigm. Remaining is the steady march of economic

globalization, the gradual convergence of all regions toward the evolving model of development in industrial regions, and progressive homogenization of global culture around the values of materialism and individualism. This puts emphasis on the role of the global market and the world outside EU,

which is should be considered in the FORESCENE scenarios.

4.1.3. ATEAM and EURURALIS projects

Global scenarios can give valuable inputs and inspiration for alternative development pathways in the world and its dynamics. Regional scenarios may have more direct con-

tributions to the analysis of the particular European development, but can also provide methodological inspiration. Two recent European scenario projects - ATEAM and EU-RURALIS - focussed on ecosystems and rural areas, respectively. These two projects followed a similar conceptual approach, but used different models and tools. The first

step of both studies involved the definition of the scenarios and the elaboration of the narrative storylines.

The ATEAM project (Advanced Ecosystem Analysis and Modelling) was conducted 2001–2004 by a consortium of European universities and research institutes with fund-

ing from the EU commission (Ewert et al., 2005; Rounsevell et al., 2006). Ecosystem development was in focus, and the project used the land use/land cover classes urban, cropland, grassland and forest land as well as introducing new land use classes such as bioenergy crops. This gives an indication of which land use classes that the

FORESCENE landscape submodel may use.

EURURALIS was undertaken in 2004-2005 by a consortium of Dutch universities and research institutes funded by the Dutch Ministry of Agriculture, Nature and Food Qual-ity (Klijn et al., 2005), and recently (December 2007) Eururalis 2.0 was released. This project focus on processes of change in Europe’s rural areas with the aim of providing

an overview of threats and opportunities to inform and encourage policy discussion. This objective is implemented through a scenario-based modelling approach to land use change and impacts on the environmental, social and economic domains. Strate-gic scenarios were elaborated, built on a 2x2 matrix, where the axes represent the

most critical uncertainties. The four scenarios were structured along two axes: (1) rang-ing from increasing globalisation to a world of regional economic and cultural blocks and (2) ranging from a future of lean governments to a future with ambitious govern-ment regulation. This way of constructing and combining strategic scenario elements is

a useful way for the scenario construction and has been used by FORESCENE as well

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

27

(WP2). The storylines were inspired by earlier work such as the emission scenarios of

the Intergovernmental Panel on Climate Change (SRES), the GEO-3 scenarios and scenarios of the Netherlands Bureau for Economic Policy Analysis (CPB).

A recent scenario study of EURURALIS (van Meijl et al, 2006; Eickhout et al, 2007) has the aim of making an assessment of land-use intensity and the related biodiversity

in agricultural landscapes in EU-25 for the year 2000 (reflecting the current situation), and explore future trends, based on the four EURURALIS scenarios up to 2030, which corresponds with suggested mid-term time perspective for the FORESCENE scenar-ios. The scenarios are quantified with a chain of models, ranging from global models to

a spatially explicit model, which simulated land use on a 1kmx1km grid for the whole EU. Data from the Farm Accountancy Data Network (FADN) were used to classify farm types in 100 regions of the EU15, according to agricultural intensity. For the ten New Member States (EU10), which are not yet considered by the FADN, country level data

were used to obtain similar farm types, which also limit the resolution for the FORES-CENE modelling efforts with an EU-25 scale. Three processes were considered for the assessment of future trends in agricultural land-use intensity: (1) land-use change, (2) conversion into organic farming, and (3) changes in productivity of crop and grassland

production. Scenario results show that for the Global economy scenario, the highest loss in ecosystem quality will take place in all regions in croplands and grasslands. The Regional communities scenario provides the best opportunities to improve ecosystem quality of agricultural landscapes. In most scenarios, agricultural land is decreasing,

while the remaining agricultural areas tend to be used more intensively. The negative impact of intensification on biodiversity is partly set off by (active or spontaneous) na-ture development on abandoned agricultural areas, but the overall trend seems to be generally negative (Reidsma et al. 2006). The very ambitious data set of the EU-

RURALIS project may be out of reach for the FORESCENE modelling efforts, but the strategic scenario constructing is of high relevance for the FORESCENE scenarios.

4.2. Studies related to the FORESCENE topic areas

Here follows a short overview of some inspiring studies for the scenario making in rela-

tion to the three topic areas of FORESCENE.

4.2.1. Material and energy flow models

There are numerous material and energy flow models from several scientific fields (for a review covering ecology and economy, see Suh, 2005). Below are two selected pro-jects with particular relevance described which have followed a more comprehensive

approach; and three recent papers are mentioned, which focus on selected aspects of potential relevance for the FORECENE meta-modelling work.

The MOSUS project was funded by the 5th framework programme (sub-programme energy, environment and sustainable development) of the European Union. The MO-

SUS project aimed at integrating the three policy themes; Sustainable development; Competitiveness and social cohesion in the knowledge-based society and; Globalisa-tion and international trade, within a macroeconomic, multi-sectoral framework. The

framework was based on an existing macro-economic model. The first application of

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

28

IO analysis of material flows using an international IO model system was being re-

alised in the project MOSUS. The four major key objectives and targets of this project were: 1. Assessing and quantifying the European use of resources (scale), including “ecological rucksacks” induced by international trade; 2. Formulating and evaluating sustainability scenarios, linking economic performance with resource use and environ-mental deterioration; 3. Refining environmental indicators to assess resource produc-

tivities, material and energy intensities and labour intensities of resource use for the EU; 4. Elaborating policy strategies and actions that reconcile long-term economic de-velopment, international trade and environmental protection.

Scenarios of the economic and social/distributional impacts of key environmental policy

measures wer made with a time horizon to 2020. The baseline scenario projects fur-ther trends observed between 1980 and 2003. The weak sustainability scenario re-flects sustainability policy goals and measures derived from strategic documents of the European Community. The strong sustainability scenario defines policy goals and in-struments, which are more ambitious from the point of view of sustainable development compared to those, included in the EU documents. The major cross-cutting drivers were economic developments, energy consumption as the two interrelated major drivers. The policy recommendations derived from the project were the following;

Environmental policy measures primarily geared towards decoupling economic activity from material and energy throughput can be positive for economic growth, contrary to the popular assumption that such policies will mainly raise costs for enterprises, de-crease competitiveness and thus have an opportunity cost in terms of reduced eco-

nomic performance.

The TransSustScan project (Scanning Policy Scenarios for the Transition to Sustain-able. Economic Structures, http://www.transust.org/transust.scan.htm) is mainly focus-

ing at economic modelling, targeted to support policies that aim at the transition to sus-tainable economic structures. The models used within the project will therefore be able to deal with; Competitiveness; Economic development; Security; The preparations for Beyond-Kyoto policies; The interaction between technological change and the use of

natural resources on the EU scale. TranSuctScan scenario building work will involve; forecasting of future states given the implementation of currently know policies; simula-tion of deviations from business-as-usual strategies; backcasting the policy patterns needed for achieving certain policy targets, with a time horizon: 2015-2030 The major

drivers derived by the preceding project TransScan are of three major kinds: techno-logical change, economic development and natural resources use. This project’s aims are in common with the FORESCENE’s project, but differ in the modelling approach and will mainly make use of already existing economic models, but may be contribu-tory.

Holz et al (2008) is presenting a model for the future supply of gas in Europe. The natural gas market in the European Union is undergoing considerable changes. Three main challenges for the next decades can be identified: the liberalization of the industry

initiated by the European Union, the increasing demand for natural gas and, simulta-neously, an increasing import dependency on gas supplied from outside the European

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

29

Union. Analyzing oligopolistic energy markets with large-scale simulation models, in

terms of data input, regional disaggregation, etc. is challenging and these models work with the underlying assumption of perfect competition as well, which makes them less appropriate for the studying the European market. This paper (and the following two below) stresses out increasing global demand for energy as an important driver for the

FORESCENE models.

Hirsch (2006) presents in his article Mitigation of maximum world oil production: Short-age scenarios, future scenarios that depicts development paths with a shortage of en-ergy resources. A framework is developed for planning mitigation of the oil shortages

that will be caused by world oil production reaching a maximum and going into decline. To estimate potential economic impacts, a reasonable relationship between percent decline in world oil supply and percent decline in world GDP was determined to be roughly 1:1. As a limiting case for decline rates, giant fields were examined. Actual oil

production from Europe and North America indicated significant periods of relatively flat oil production (plateaus). However, before entering its plateau period, North American oil production went through a sharp peak and steep decline. Examination of a number of future world oil production forecasts showed multi-year rollover/roll-down periods,

which represent pseudoplateaus. Consideration of resource nationalism posits an Oil Exporter Withholding Scenario, which could potentially overwhelm all other considera-tions. Three scenarios for mitigation planning resulted from this analysis: (1) A Best Case, where maximum world oil production is followed by a multi-year plateau before

the onset of a monatomic decline rate of 2-5% per year; (2) A Middling Case, where world oil production reaches a maximum, after which it drops into a long-term, 2-5% monotonic annual decline; and finally (3) A Worst Case, where the sharp peak of the Middling Case is degraded by oil exporter withholding, leading to world oil shortages

growing potentially more rapidly than 2-5% per year, creating the most dire world eco-nomic impacts. As energy demand, and the shift from non-renewables to renewables is considered as one of the more important drivers in the resource use, this study is em-phasising this, and also points out the importance of plateus. The implications for the

FORESCENE scenarios is the importance to identify major cross-cutting drivers, whereof the energy demand situation is one, which is further illustrated in the two stud-ies below.

Smeets et al. (2007) used a model for estimating bioenergy production potentials in 2050 on a global level, called the Quickscan model. The Quickscan model uses a bot-

tom-up approach and its development is based on an evaluation of data and studies on relevant factors such as population growth, per capita food consumption and the effi-ciency of food production. Three types of biomass energy sources are included: dedi-cated bioenergy crops, agricultural and forestry residues and waste, and forest growth.

The bioenergy potential in a region is limited by various factors, such as the demand for food, industrial roundwood, traditional fuelwood, and the need to maintain existing forests for the protection of biodiversity. Only the surplus area of agricultural land is included as a source for bioenergy crop production. The model results indicate that the

application of very efficient agricultural systems combined with the geographic optimi-zation of land use patterns could reduce the area of land needed to cover the global food demand in 2050 by as much as 72% of the present area. A key factor was the area of land suitable for crop production, but that is presently used for permanent graz-

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

30

ing. Another key factor is the efficiency of the production of animal products. The global

potential of bioenergy production from agricultural and forestry residues and wastes was calculated to be 76-96EJyr-1 in the year 2050. The paper is indicating the impor-tance and the potential of the shift from non-renewables to renewables in the FORES-CENE scenarios and what biomass sources that may be considered and included.

4.2.2. Land use and biodiversity models

Scenario development has become a popular tool for the assessment of land use change and a large number of studies using scenario approaches have been published during recent years (Rabbinge & van Diepen., 2000; Rotmans et al., 2000; de Nijs et

al., 2004; Ewert et al., 2005), whereof many are or relevance for the FORESCENE Land use modelling. From ecosystem functioning and biodiversity to water resources and greenhouse gas emissions, land use is central. In Europe, the most important land uses are agriculture and forestry, which cover about 45% and 36% of the total land

area, respectively (FAO, 2003). A range of models has been developed to better un-derstand, assess and project changes in land use and land cover (Veldkamp and Ver-burg, 2004, Rounsewell, 2006).

In the paper by Rounsewell et al. 2006, a coherent set of future land use change sce-

narios for Europe was made. The paper presents a range of future, spatially explicit, land use change scenarios for the EU15, Norway and Switzerland based on an inter-pretation of the global storylines of the Intergovernmental Panel on Climate Change (IPCC) that are presented in the special report on emissions scenarios (SRES). The

methodology is based on a qualitative interpretation of the SRES storylines for the European region, an estimation of the aggregate totals of land use change using vari-ous land use change models and the allocation of these aggregate quantities in space using spatially explicit rules. The scenarios include the major land use/land cover

classes urban, cropland, grassland and forest land as well as introducing new land use classes such as bio-energy crops and abandoned land and set-asides, which is indicat-ing the appropriate land use classes for FORESCENE as aforementioned The ap-proach to estimate new protected areas is based in part on the use of models of spe-

cies distribution and richness. All scenarios assume some increases in the area of bio-energy crops with some scenarios assuming a major development of this new land use, which is also indicating for the FORESCENE modelling how sub-models for re-source use and land use may be interconnected.

The project ALARM - Assessing Large-scale Risks for biodiversity with tested Meth-ods, a project within EU sixth framework programme, has a research focus on as-sessment and forecast of changes in biodiversity and in structure, function, and dy-namics of ecosystems to 2050. This relates to ecosystem services and includes the

relationship between society, economy and biodiversity. In particular, risks arising from climate change, environmental chemicals, biological invasions and pollinator loss in the context of current and future European land-use patterns are assessed. The objectives of the project are the following: To develop an integrated large scale risk assessment

for biodiversity; To focus on risks consequent on climate change, environmental chemicals, rates and extent of loss of pollinators and biological invasions; To establish

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

31

socio-economic risk indicators related to the drivers of biodiversity; and finally, to pro-

vide a contribution to objective based politics, to policy integration and to derive out-come-oriented policy measures in the field of biodiversity. Socio-economics as a cross-cutting theme will contribute to the integration of driver-specific risk assessment tools and methods and will develop instruments to communicate risks to biodiversity to end

users, and indicate policy options to mitigate such risks. The project is not scenario-focused, instead is there a strong emphasis on risk assessment. The two major cross-cutting drivers used within the project is climate change and land use.

PRELUDE (PRospective Environmental analysis of Land Use Development in Europe), an European Environmental Agency project, initiated in 2006, explores what European landscapes may look like 30 years from now and beyond. Instead of making predic-tions, it tackles the vast uncertainties of the distant future by analysing a range of plau-

sible developments. Prelude has used an innovative Story-and-simulation approach to scenario development, integrating qualitative and quantitative aspects in scenarios, including uncertainties and underlying driving forces, and how these might influence land use.

Five contrasting futures are depicted in a set of coherent scenarios. The five PREL-UDE scenarios are: 1. Great Escape - Europe of contrast; 2. Evolved Society - Europe of harmony; 3. Clustered Networks - Europe of structure; 4. Lettuce Surprise U - Europe of innovation; 5. Big Crisis - Europe of cohesion

Great Escape: This scenario is driven by globalisation, decreasing solidarity and pas-sive government. Societal tension builds up as relatively poor immigrants move to ur-ban city centres. Climate change affects the growing conditions for agriculture. The agricultural market is liberalised and only large-scale farms with intensive manage-

ment survive the pressure from the world market.

Evolved Society: Main ingredients in this scenario are an energy crisis, growing envi-ronmental awareness and active rural development. Serious flooding occurs and peo-ple leave the most vulnerable areas. They rediscover the countryside where small-

scale organic farming, supported by strong policy measures, increases.

Clustered Networks: This scenario is all about optimization of land use and strong spa-tial planning in response to an ageing of society and a declining agricultural sector. Climate change is a less prominent driver in this scenario.

Lettuce Surprise U: The essential drivers here are growing environmental awareness,

technological innovation and decentralization. Agriculture revolutionizes, facilitated by open source mentality and propagation of knowledge. Production becomes small scale and less intensive.

Big Crisis: In this scenario climate change related disasters and increasing solidarity

are all-important. Floods and droughts affect many people and trigger strong European policy interventions, aimed at a balanced regional development.

The scenarios cover a wide spectrum of possible developments of drivers of change in society, economy, governance, environment and technological invention. A full descrip-

tion can be found in EEA 2007. All scenarios except one assumes a net migration from

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

32

current urban centres towards the periphery, Another dominant feature is the loss of

agricultural land,

The PRELUDE project has an extensive list of driving forces behind the changes in land use patterns in Europe, whereof subsidiarity, human population development and settlement, immigration, economic and technological growth, climate change, energy

supply self-sufficiency, agriculture, health, social equity and quality of life, are the most prominent ones. The identified 20 driving forces were further aggregated into 5 major categories; Environmental concern, Solidarity and equity, Governance and interven-tion, Agricultural optimisation, technology and innovation, with various weighing of

these. This list is of particular importance for the FORESCENE scenarios, as are the five scenarios of PRELUDE.

4.2.3. Water use models

Scenario models for water use has a long history within hydrological dynamic model-

ling at various levels, from global down to local. With the Water Framework Directive (WFD, Directive/2000/60/EC), EU established an environmental policy which aims at achieving a good status of surface water and groundwater, which called for new tools and approaches. The Water framework directive (WFD), adopted in October 2000, is a

major document that will guide the management of aquatic environments in Europe during the coming decades. The document shifts from a water-management policy based on usage, e.g. fishing, crop irrigation or electrical production, to a policy focusing on aquatic life to avoid damage to ecosystems. Hydrological modelling has a long his-

tory within the engineering sciences, with models such as

TopModel (http://www.es.lancs.ac.uk/hfdg/topmodel.html), and EU-projects such as Rebecca (http://www.rbm-toolbox.net/rebecca/), BMW (http://www.environment.fi/default.asp?contentid=116046&lan=EN ),

DaNubs (http://www.icpdr.org/icpdr-pages/danubs.htm ) and Harmoni-CA (http://www.harmoni-ca.info/Catchment_Modelling_projects/The_EC_ CatchMod_Cluster.php).

However, this rather technical approach may not always be suitable for policy testing

and cross-disciplinary approaches.

Bayesian Belief Network models are emerging as a valid approach for modelling and policy testing and supporting decision-making in the field of water resource manage-ment (Casteletti & Soncini-Sessa, 2007), which may also be a tentative modelling pathway for the FORESCENE modelling efforts. In the water resource context they

have been used by Batchelor and Cain (1998) in irrigated and rainfed farming system modelling, by Varis and Kuikka (1997) to investigate the effect of climate change on surface waters and by Borsuk et al. (2001, 2004) in studying the eutrophication of river estuaries. In the paper by Casteletti and Soncini-Sessa (2007) a comparison of Baye-

sian networks, mechanistic models, empirical models, and Markov chains is made. They use a list of criteria, such as; Ease of identification, integration potential, dynam-ics and parsimoniousness. They conclude that Bayesian models do not always meet all these criteria. A Bayesian network is a type of model that provides a simplified seman-

tics that is useful when knowledge about the system to be modelled is poor or unstruc-tured, and mainly empirical in nature, but that Bayesian networks lose some of their

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

33

potential when the system being considered is dynamic, and includes recursive deci-

sions. For the FORESCENE Bayesian network model this may be considered, but the main objective for FORESCENE is to combine and cross over domains, but with somewhat restricted capability of capturing dynamics of systems.

4.3. Summing-up

Global scenario studies may give important inspiration for general alternative trends. But they have seldom sufficient detail for more concrete contributions for an analysis at the regional European level. Regional scenarios as well as those related to the particu-lar focus topics of FORESCENE provide more direct and concrete contribution.

Most existing models have been developed to deal with rather specific aspects of sustainability, either within one discipline or a geographical area, whereas FORES-CENE adopts a much broader perspective, trying to integrate sustainability problems and activities, which mostly have been treated separately so far. FORESCENE also

has the ambition to introduce a more extended spatial and temporal perspective. There are many studies that can provide inspiration both on scenario narratives and concrete analysis of various topics, but the most concrete and direct contributions are given by studies such as EURURALIS, which have similarities with the focus and ambitions of

FORESCENE. It also becomes clear that existing models cannot cover FORESCENE’s scope and that another modelling framework is needed. Such a framework should al-low to combine the various aspects so far often treated separately and to model the basic links between the environmental problems, key drivers, and sustainability strate-

gies from an overview perspective, with an option to supply the meta-model with data from more detailed and specific models.

Table 5: Summary of scenario models and their relation to FORESCENE

Relation to FORESCENE

Temporal scope Most models deal with a -25-30 year time horizon, whereof eco-

nomic models usually have a shorter time horizon.

FORESCENE will ad here to the time horizon to 2050,

which is the same as of IPCC scenarios.

Spatial scope Most models deal with parts of EU FORESCENE aim at EU-25-

level

Problem field(s) Usually mono-disciplinary in their

problem field approach

FORESCENE aim at combin-

ing several problem fields

Cross-cutting drivers Economic growth, competitiveness

and climate change are the domi-nating cross-cutting drivers

FORESCENE will use these

cross-cutting drivers as well as various others.

Key strategies Decoupling between economic growth and environmental impact;

Curbing climate change by limiting carbon emissions; adapt to water

shortages by improved technology;

These key strategies will be used by the FORESCENE

meta-model, but also com-bined with others.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

34

5 . S c e n a r i o m o d e l l i n g i n F O R E S C E N E

5.1. Need for a framework or ‘meta-model’

When the scope of FORESCENE project is considered, in the light of the Sustainability Scenario Elements that was identified as a result from the FORESCENE Work pack-

ages 1 & 2 and the short review conducted above, it was decided to develop a specific meta-model. The term meta-model stands for a model at an aggregate level and in a simplified way of the highly complex interactions between the socio-industrial system (especially the activity / policy fields agriculture, infrastructure / built environment, and

industry / economy) and the three environmental fields ‘resource use / waste’, ‘land-scape / biodiversity / soils’ and ‘water / water use’. The meta-model should help testing the preliminary narratives (i.e. combinations of SSEs, see section 3) in comparison with Business-As-Ususal scenarios (developed in WP4) and refine them towards integrated

alternative scenarios (in WP5).

Several pertinent questions were arising from the previous workshops, considering how to choose a modelling approach for the emerging scenarios (Jörgensen, 2008): Can we model a system that has only uncertain observations/ data? The forcing func-

tions and several ecological processes are stochastic. How to account for that? How to develop models if the databases are very heterogeneous, i.e. based on observations and data from many different system types? How to develop models of systems, when our knowledge is mainly based on a number of rules/properties/ propositions?

There were two major alternative approaches in consideration for constructing such a meta-model; a system dynamic approach or Bayesian Network approach. Both have their strengths and weaknesses (Varis, 1997). System dynamic models have generally a larger data need, while Bayesian networks can be used to build a decision support

system, especially when working under uncertain conditions (Castelletti & Soncini-Sessa, 2007). The latter is therefore better adapted to the situation, where the aim is to connect different areas, as a meta-model approach. The term meta-model stands for a model at an aggregate level and in a simplified way of the highly complex interactions

between the socio-industrial system (especially the activity / policy fields agriculture, infrastructure / built environment, and industry / economy) and the three environmental fields ‘resource use / waste’, ‘landscape / biodiversity / soils’ and ‘water / water use’. The meta-model should help testing the preliminary narratives (i.e. combinations of

SSEs, see section 3) in comparison with BAU scenarios (developed in WP4) and refine them towards integrated alternative scenarios (in WP5). Bayesian Belief Networks can be used to build a decision support system, especially when working under uncertain conditions.

5.2. Short presentation of Bayesian Networks

In the area of environmental and resource management, the applications of Bayesian analysis have been largely dominated by classical Bayesian inference (i.e. parameter estimation and uncertainties). In decision theory, the idea of considering the entire

model as a construct subject to uncertainty and subjectivity stem from the game theory

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

35

of the 1930s and '40s (Shafer, 1990) and decision trees were developed. The basic

theory was developed into more applicable level towards the late 1960s (Raiffa, 1968). For a long time, one of the bottlenecks to practical applications of Bayesian ap-proaches has been the high amount of computation required. Powerful numerical tech-niques have not been available until mid-1980s, and today most personal computers

are powerful enough to run such models. This has proliferated an expansion and use of this approach and has resulted into several methodologies - known as belief net-works, causal networks, Bayesian nets, qualitative Markov networks, or constraint net-works - this division is difficult or impossible to distinguish. Characteristic of these tech-

niques is the principle of networking nodes representing conditional, locally updated probabilities. The local-updating principle allows construction of large and densely cou-pled net- works without excessive growth in computation (Varis, 1997).

Bayesian belief networks (BBNs) are useful tools for modelling complex situations and

aiding in management and decision-making (Jensen, 2001; Marcot et al. 2001). BBNs also serve well as part of a risk-management Framework by explicitly displaying the ‘causal web’ of interacting factors and the probabilities of multiple states of predictor and response variables. Development of a BBN model, however, does not follow a

standard process in the making. The first step is often done in discussion with expert and/or stakeholders, and after an initial review of the literature. From a mathematical point of view, the basic property of Bayesian networks is the chain rule: a Bayesian network is a compact representation of the joint probability table over its universe.

From a knowledge engineering point of view, a Bayesian network is a type of graphical model. The structure of the network is formulated in a graphical communication lan-guage for which the language features have very simple semantics, namely causality. The graphical specification also specifies the requirements for the quantitative part of

the model, the conditional probabilities. The graphical representation is for humans to read, and it helps to focus attention, when working in a group jointly developing a model. The graphical model has to be well-defined, to ensure it can be communicated to a computer. The Bayesian network is a sufficiently well-defined language, and be-

hind the graphical specification in the user interface of Bayesian network software, there is an alpha-numeric specification language. Bayesian network models are acyclic, that is, there are not any feedback loops, which separate them from dynamic models. Tractability may be considered, by using algorithms for probability updating of the Bayesian network.

Bayesian networks has a long history in statistics, and in the first half of the 1980s they were introduced to the field of expert systems through work by Pearl (1982) and Spiegelhalte et al. (1993). Characteristic of these techniques is the principle of net-working nodes representing conditional, locally updated probabilities. The local-

updating principle allows construction of large and densely coupled networks without excessive growth in computation. Furthermore, networks can easily be constructed to operate interactively and on-line. As is usual in such techniques, the entire model - the hypothesis space – is subjected to Bayesian analysis, not only the parameter space. In

recent years, they have spread quickly to many application areas, including fault diag-nosis, reliability theory, medicine, pattern recognition, and decision analysis (Varis, 1997). In historical perspective, the roots of decision trees are even older than those of modern Bayesian decision theory. The basic idea is that the events, both controllable

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

36

(decision) and uncontrollable (chance) ones, are set up in procedural order. Each set

of outcomes from a node, be they decision alternatives or possible outcomes of a chance, define a new branch in the tree. Influence diagrams have gained increasing popularity within decision analysis, originally extended from decision trees. An influ-ence diagram is an acyclic network of nodes connected with one-directional links. The

nodes represent probabilistic variables, deterministic variables, or decisions. Like a decision tree, the diagram describes causality or the flow of information and probabilis-tic dependencies in a system. The influence diagram notation provides the analyst with some attractive and useful properties. One of them is the straight way to perform the

value of information analysis. The network has n nodes that can arbitrarily be intercon-nected. The prior probabilities assigned to the outcomes are updated with the informa-tion linked from other parts of the network, yielding the posterior probability distribution. Generalized belief networks can include models from many methodological families

that have conventionally been considered as being distant, e.g. pragmatic, linguistic, mechanistic, and metric models can be used together in a hybrid model or in a meta-model (Varis 1994).

The Bayesian network approach to modelling is not without its shortcomings (Borsuk et

al 2004). Perhaps the most profound is the inability to explicitly represent system feed-backs. Bayesian networks are defined as being directed acyclic graphs, so relation-ships must represent either one-way causal influences at a particular instant in time or net influences on eventual steady-state conditions. An alternative is to construct a dy-

namic Bayesian network (Jensen, 2001) in which a down-arrow variable in one time step can influence an up-arrow variable in the next. Such a model requires significantly more information to quantify the time dynamics. However, insufficiently representing dynamic aspects of system behaviour can lead to unexpected consequences that are

not adequately captured by the probabilistic predictions (Jorgensen, 1999).

5.3. Preliminary model structure for FORESCENE

Here follows a brief description of the structure of the meta-model in relation to chap-ters 2 and 3, giving and overview of the FORESCENE meta model.

5.3.1. Problem fields/environmental pressures

The current EU environmental policy context is determined by the four priorities of the 6th EAP on climate change, nature and biodiversity, environment and health and qual-ity of life, and on natural resources and waste, which is also indicating the most impor-tant problem fields. These priorities have been translated into seven Thematic Strate-

gies that are being developed according to a common approach independently of the specific content requirements relating to their subject matter:

Soil protection; Protection and conservation of the marine environment; Sustainable use of pesticides; Air pollution; Urban environment; Sustainable use and management

of resources; Waste recycling. FORESCENE’s central hypothesis that the production and consumption patterns have essential influence on the interaction between industry and society on the one hand and the environment on the other hand, is also guiding the integration between the problem fields. A key methodological approach will be the

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

37

analysis of the material flows between and within those spheres, and the transforma-

tion links between one material flow to another one.

5.3.2. Activities

The economically defined activities energy supply, agriculture, water supply and con-struction appear to be most relevant with regard to pressure-causing factors and im-

pacts on the environment. Transport, forestry, chemicals, basic metals, and food prod-ucts are also activities or product groups potentially important that will be considered in the FORESCENE meta-model.

5.3.3. Cross-cutting driving forces

Five major cross-cutting factors were identified during the workshops (see description in Chapter 2 & 3 above), in particular Production patterns & Economic development; Consumption patterns; and natural system dynamics (incl. climate change, and deple-tion of natural resources). These factors are also related to each other, but will be used

separately as driving forces.

5.3.4. Goals

Based on WP2, and before the background of maximised well-being as an over-arching goal for the EU sustainable development, the sustainable goal references

shown in Table 3 and 4 will be taken as starting point.

5.3.5. Key strategies

The twenty-five sustainability strategies that were defined and grouped during WP2, and that were related to three major activity fields (agriculture, industry/economy, and

infrastructure/land use), are the basis for the development of narratives as well as for further modelling. This relates especially to those strategies which can be operational-ized by indicators and other quantified parameters (Fig. 2). The meta-model of FORESCENE should therefore be made in such a way so optional key strategies could

be tested.

5.3.6. Submodeliing systems

To be able to study the impact of the identified cross-cutting driving forces and their related activities on the problem fields, the FORESCENE modelling system is sug-

gested to contain a minimum set of three submodelling systems; Materials and waste flow, Land use and biodiversity, and Water use. There are two major issues that differ-entiate between the three submodelling systems; in the Material and waste production submodel, also the extra EU dimension of the system will be considered, as minerals can be imported, while in the Biodiversity and landscape submodel will focus on the

intra EU impacts while trying to cover also impacts of land use change outside the EU due to increased biofuels imports; whereas in the water submodel, commodities out-side of EU will not be considered, but a regional breakdown within the EU should be implemented according to varying water availability in European regions.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

38

Another aspect that will differentiate the submodels are that the Material and waste

production submodel is assumed to be driven amongst other factors by economic growth, while the other two are indirectly connected to economic growth through the demand for mineral and biomass materials and water. In this model it is therefore as-sumed that GDP is a driver, not an output. Moreover, biodiversity (but also some as-

pects of Water use) is a parameter that also includes qualitative aspects (as well as ethical) which will have to be considered when interpreting the uncertainties for this parameter. A common denominator for all three submodels is Population. The various indicators that are identified in the submodels, respectively, are not shown in the sim-

plified views below. The structural model can also be seen as a synthesis of the work-shops performed previously in the FORESCENE project.

Below is a simplified view of the Material and waste production submodel.

Below is a simplified view of the Landscape and Biodiversity submodel. It relates to the other two submodelling systems through Agriculture and Forest in terms of water ab-straction, and to the Material and waste submodelling system through the biofuel de-mand and biomass production.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

39

Below is a simplified view of the Water use submodel. This submodel differs also from

the other two as it is more spatially explicit, as the future water situation is assumed to be very different in different parts of Europe, especially between North and South.

5.4. Uncertainties

Bayesian networks offer a pragmatic and scientifically credible approach to modelling

complex ecological systems, where substantial uncertainties exist (Polino et al. 2007). Bayesian networks are being used to model diverse problems of high complexity (Laskey and Mahoney, 2000 and Korb and Nicholson, 2004), including environmental applications (Borsuk et al., 2004, Bromley et al., 2005, Ticehurst et al., 2007 and Varis

and Fraboulet-Jussila, 2002). In many Bayesian networks, variables have been pa-rameterized using either knowledge or data (Borsuk et al., 2004, Bromley et al., 2005, Rieman et al., 2001 and Ticehurst et al., 2005), but rarely have both these information sources be combined in order to parameterize one variable. Straightforward sensitivity

and uncertainty analysis of a belief network is thus highly time-consuming and difficult, and often the knowledge of experts is incomplete (Morgan and Henrion, 1990). Con-versely, often significant data gaps exist for parameterizing variables with data.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

40

6 . C o n c l u s i o n s

In this paper, based on the preceeding work in the project, we have delineated prelimi-nary narratives for the scenarios to be developed in the FORESCENE project. We

have reviewed relevant existing work on sustainabiliy scenarios and modelling, and shown that there is the need for a meta-model which combines the different problems and activities fields covered by FORESCENE. The Bayesian belief networks method-ology seems a promising tool to establish such kind of meta-model for which we have

described the essential elements and outlined the basic structure of the main sub-models..

Further work will have to develop the meta-model. For that purose, it will need to be filled with data and quantitive descriptions of functional relations of the various nodes

including probabilities. The sub-models will need to be checked for their own working capacity and then combined to test the meta-model with a selection of scenarios, for the appropriate temporal and spatial scales.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

41

7 . R e f e r e n c e s

Batchelor, C.H. and Cain, J.D. 1998. Application of belief networks to water manage-ment studies. Agricultural.Water Management. 40, 51-57.

Borsuk, M., Stow, C., Higdon, D., Reckhow, K., 2001. A Bayesian hierarchical model to predict benthic oxygen demand from organic matter loading in estuaries and coastal zones. Ecological Modelling 143, 165-181.

Borsuk, M., Stow, C., Reckhow, K., 2004. A Bayesian network of eutrophication mod-els for synthesis, prediction, and uncertainty analysis. Ecological Modelling 173, 219-239.

Bromley, J., Jackson, N.A., Clymer, O.J., Giacomello, A.M., Jensen, F.V., 2005. The use of Hugin to develop Bayesian networks as an aid to integrated water resource planning. Environmental Modelling and Software 20,: 231-242.

Castelletti, A. & Soncini-Sessa, R. 2007. Bayesian Networks and participatory model-ling in water resource management. Environmental modelling and software 22: 1075-1088.

de Nijs, T.C.M., de Niet, R., Crommentuijn, L., 2004. Constructing land-use maps of the Netherlands in 2030. Journal of Environmental Management. 72: 35–42.

Eickhout, B., H. van Meijl, A. Tabeau and T. van Rheenen, 2007. Economic and eco-logical consequences of four European land-use scenarios. Land Use Policy, 24: 562 – 575.

Ellis, E.C., 2004. Long-term ecological changes in the densely populated rural land-scapes of China. In: DeFries, R.S., Asner, G.P., Houghton, R.A. (Eds.), Ecosystems and Land Use Change. American GeophysicalUnion, Washington, pp. 303–320.

Ewert, F., Rounsevell, M.D.A., Reginster, I., Metzger, M.J., Leemans, R., 2005. Future scenarios of European agricultural land use: I. Estimating changes in crop productivity. Agriculture Ecosystem Environment 107: 101–116.

FAO, 2003. The agricultural statistics of the Food and Agriculture Organisation of the United Nations. www.FAO.org.

Hirsch, R.L. 2008. Mitigation of maximum world oil production: Shortage scenarios. Energy policy 36: 881-889.

Holz, F., von Hirschhausen, C & Kemfert, C. 2008. A strategic model of European gas supply (GASMOD). Energy Economics 30: 766-788.

Houghton, R.A., 2003. Why are estimates of the terrestrial carbon balance so different? Global Change Biol. 9, 5000–5009.impact. Journal of Environmental Management 72:1–3.

IPCC 2001. Watson, R.T. and the Core Writing Team (Eds.). IPCC Third Assessment report. IPCC, Geneva. www. Ipcc.ch

Jensen, F.V. 2001. Bayesian Networks and Decision Graphs. Springer Ver-lag,Heidelberg, D.

Jørgensen, S.V. 2008. Overview of the model types available for development of eco-logical models. Ecological modelling 215: 3-9.

Klijn, J.A., Vullings, L.A.E., van den Berg, M., van Meijl, H., van Lammeren, R., van Rheenen, T., Veldkamp, T., Verburg, P., Westhoek, H. & Eickhout, B. 2005. The EURURALIS study: technical document. Alterra-rapport, 1196. Alterra, Wageningen.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

42

Korb, K.B., Nicholson, A.E., 2004. Bayesian Artificial Intelligence. Chapman and Hall/CRC Press, London.

Laskey, K.B., Mahoney, S.M., 2000. Network engineering for agile belief network mod-els. IEEE Trans. Know. Data Eng. 12, 487e498.

Marcot, B., Holthausen, R.S., Raphael, M.G., Rowland, M., Wisdom, M.J.. 2001. Using Bayesian belief networks to evaluate fish and wildlife population viability under land management alternatives from an environmental impact statement. Forest ecology and management 153:29-42.

Morgan, M.G., Henrion, M., 1990. Uncertainty: A Guide to Dealing with Uncertainty in Quantitative Risk and Policy Analysis. Cambridge University Press, Cambridge, UK.

MEA. 2005. The Millennium Ecosystem Assessment.Pearl, J., (1988), Probabilistic Reasoning In Intelligent Systems: Networks of Plausible Inference. Morgan Kaufmann Publishers, Inc. San Francisco, California.

Pollino, C. A., Woodberry, O., Nicholson, A., Korb, K. & Hart, B.T. 2007. Parameterisa-tion and evaluation of a Bayesian network for use in an ecological risk assessment. Environmental modelling and software 22: 1140-1152.

Rabbinge, R. & van Diepen, C.A. 2000. Changes in agriculture and land use in Europe. European Journal of Agronomy 13: 85-99.

Raiffa, H. 1968. Decision Analysis. Addison-Wesley. Reading, MA, USA.

Raskin, P., Gallopin, G., Gutman, P., Hammond, A. & Swart, R., 1998. Bending the curve: towards global sustainability. PoleStar Series Report, 8. Stockholm Environment Institute-Boston, Boston.

Reidsman, P., Tekelenburg, T., van den Berg, M. & Alkemade, R. 2006. Impacts of land-use change on biodiversity: An assessment of agricultural biodiversity in the European Union. Agriculture, Ecosystems, Environment. 114: 86-102.

Rieman, B.E., Peterson, J.T., Clayton, J., Howell, P., Thurow, R., Thompson, W., Lee, D.C., 2001. Evaluation of potential effects of federal land management alternatives on trends of salmonids and their habitats in the interior Columbia River basin. Forest Ecology and Management 153: 43-62.

Rotmans, J., van Asselt, M., Anastasi, C., Greeuw, S., Mellors, J., Peters, S., Rothman, D., Rijkens, N. 2000. Visions for sustainable Europe. Futures 32: 809-831.

Rounsevell, M.D.A., Reginster, I., Arau´jo, M.B. , Carter, T.R., Dendoncker, N., Ewert, F., House, J.I., Kankaanpää, J.I., Leemans, R., Metzger, M.J., Schmit, C., Smith, P. & Tuck, G. 2006. A coherent set of future land use change scenarios for Europe. Agricul-ture, Ecosystems and Environment 114: 57-68.

Shafer, G. (1990) Decision making. In Readings in Uncertain Reasoning, eds G. Shafer and J. Pearl, pp. 61-67. Morgan-Kaufmann, San Mateo, CA. USA.

Smeets, E. M.W., Faaij, A.P.C., Lewandowski, I. M. & Turkenburg, W. C. 2007. A bot-tom-up assessment and review of global bio-energy potentials to 2050. Progress in Energy and combustion science 33: 56-106.

Spiegelhalter, D.J., Dawid, A.P., Lauritzen, S.L., Cowell, R.G. 1993. Bayesian analysis in expert systems. Statistical Science 8: 219-283.

Suh, S. 2005. Theory of materials and energy flow analysis in ecology and economics. Ecological modelling 189 (3-4): 251-269.

FORESCENE D.3.2 – Technical report Possibilities for modeling sustainability scenarios

43

Ticehurst, J.L., Rissik, D., Letcher, R.A., Newham, L.H.T., Jakeman, A.J., 2005. Devel-opment of decision support tools to assess the sustainability of Coastal Lakes. In: Zerger, A., Argent, R.M. (Eds.), MODSIM 2005 International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand, (De-cember 2005); Melbourne, Australia, pp. 2414-2420.

Ticehurst, J.L., Newham, L.H.T., Rissik, D., Letcher, R.A., Jakeman, A.J., 2007. Baye-sian network approach for assessing the sustainability of Coastal Lakes in New SouthWales, Australia. Environmental Modelling and Software 22: 1129-1139.

UNEP. 2002. The Global Environment Outlook.

Van Meijl, H., T. van Rheenen, A. Tabeau and B. Eickhout, 2006. The impact of differ-ent policy environments on land use in Europe. Agriculture, Ecosystems and Environ-ment, 114: 21-38.

Varis, O. (1994) A belief network approach to modeling of environmental change: the methodology and prospects for application. IIASA, Laxenburg, Austria, WP-94-40.

Varis, O. 1997. Bayesian decision analysis for environmental and resource manage-ment. Environmental modellinga and software 12: 177-185.

Varis, O., Kuikka, S., 1997. Bayesian approach assessment on surface waters to ex-pert judgment elicitation with case studies on climatic change impact. Climatic Change 37: 539- 563.

Varis, O., Fraboulet-Jussila, S., 2002. Water resources development in the Lower Senegal River Basin: conflicting interests, environmental concerns and policy options. Water Research and Development 18,: 245-260.

Veldkamp, A., Verburg, P.H., 2004. Modelling land use change and environmental im-pact. Journal of Environmental Management 72: 1–3.

Water Framework Directive. 2000. EU Directive/2000/60/EC