ESPA Models to Scenarios Workshop 22–24 October, 2012 Charles Darwin House, London Both modeling and scenario development are central to a number of Ecosystem Services for Poverty Alleviation (ESPA) projects, yet the terms are used loosely and in different ways The goal of the workshop was to develop a shared understanding of how models and scenarios can best be used for ESPA research, and to develop links among approaches that would support individual projects and the ESPA programme more generally One central question concerned how a research-based understanding of ecosystem services, and their role in poverty alleviation, could be used to develop approaches to sustainable poverty alleviation over a 20–40 year timescale This timescale is challenging because it is longer than the simple projections of current trends (usually 5–10 years), but shorter than the timescales that many scenarios and visioning processes tend to use (often 50 years) The workshop examined: 1) the full range of tools and models available; 2) how to link such models with scenario-based approaches for examining ecosystem servicebased pathways for the sustainable alleviation of poverty; 3) how scenario development over the 20–40 year timescale can be advanced within the ESPA context Workshop Objectives To agree on a typology and on the concepts for models and for scenariobased tools and methods, appropriate for ESPA This is necessary to achieve the goals of the workshop To develop a general approach to using scenarios and models for ESPA To identify gaps in knowledge and understanding, or obstacles to implementing the agreed approach Define next steps for the work in ESPA Models and Scenarios Typology Models and scenarios should challenge each other and are a means of communicating the science, data and knowledge around ecosystem services for poverty alleviation Models There are many models used in the ESPA context, ranging from simple statistical models to highly complex integrated-assessment or agent-based models An important distinction was drawn between static and dynamic models, and between phenomenological and mechanistic models Static, phenomenological models (e.g species distribution models) may provide robust means to understanding current environment-species linkages, and to make predictions about the future, but they lack processes that represent how one state transforms to another Models contain various levels of uncertainty including structural uncertainty, parameter uncertainty, driver uncertainty and context dependency Each of these needs to be considered separately and the uncertainties made clear If necessary, the significance of each kind of uncertainty for model outcomes can be tested Therefore, the assumptions and uncertainties in models must be considered when making them fit for purpose Models should allow relevant simplifications of complex processes They can then be used to test assumptions Creating abstractions based on models can be helpful for clarifying what matters most, where uncertainties are influential and what the most influential parameters or variables in a system might be Scenarios Scenarios also take a range of forms Scenarios may be used in a range of ways including for designing interventions, exploring plausible alternative futures, for developing desirable versus undesirable visions for the future, or for projecting into the future in different ways from a specified starting point Each of these links to modeling in different ways, with some being model-based (interventions, projections), while others are more based on story lines only Often there is a combination and productive interplay with models being used to challenge scenarios, while scenarios are used to design and develop models Models and scenarios are therefore intimately linked The interaction between them is important and at times they merge into each other Discussion and Conclusions A range of different but complex systems is involved in ecosystem services for poverty alleviation research The use of models and their implementation in scenarios can support understanding and better decisions There are many processes, states, feedbacks and emergent properties that are important, but that can be defined only once the context is clearly prescribed A range of concerns about the definition of terms, sources and extents of uncertainties need to be clarified at the outset Conceptual frameworks are useful for this purpose and could play a very important role in many ESPA projects in ensuring that there is a good understanding of how and where models and scenarios are being developed During the workshop, participants were asked to explore three different scenarios to project the future (A, B & C in Figure 1), based on their own experience, exploring the steps necessary to obtain plausible but ‘desired’ future scenarios, as opposed to simply modeling business-as-usual ‘A’ represented forecasting from the present; ‘B’ represented mechanistic models of alternative future pathways, and; ‘C’ represented back-casting from future desirable states It was agreed that all three approaches had a role in ESPA, however, tools and techniques for C appear to be less well developed than those for A and B B was generally easier to envisage, while C created a difference in opinion, with some enjoying the opportunity to explore more visionary goals and others worrying that C tends to become too complex and global, introducing too many external drivers It is also important to consider who it is that has defined B or C, and to question whether we have the best B or C scenario Figure 1: Scenario planning for long-term changes in poverty alleviation through the effective management of ecosystem services All stakeholders should be involved in the decision-making process as well as in the development both both models and of scenarios They should all be fully involved in the process of developing models and scenarios which should be a dynamic process to develop and implement the models and scenarios, with feedbacks and checks on inputs and outputs Key issues include the different power relationships between stakeholders, their input to model development and design, and how to strike the right balance between complexity and reality, which may be perceived differently by different stakeholders Next Steps for ESPA Directorate Review ESPA relevant case-studies, methods and application of results with respect to scenarios and modeling Document learning on scenarios and models Facilitate access to available models and best practice Facilitate communication between projects and provide opportunity to apply for funding for future workshops through the Regional Opportunities Fund Annex 1: AGENDA ESPA Models to Scenarios Workshop 22–24 October, 2012 Charles Darwin House, London ESPA is an interdisciplinary research programme that aims to deliver high-quality, cutting-edge research that will improve our understanding of the way ecosystems function, the services they provide and their relationship with the political economy and sustainable growth The workshop will focus on forging the links between modelling and scenarios in ecosystem services and poverty alleviation research The aim of the workshop is to share experiences and knowledge and to develop a shared approach to tools and techniques for examining options for ‘sustainability’ in the 20– 40 year timescale We will examine: The full range of tools and models available (e.g agent-based, statistical, processbased and global change models); How to link such models into scenario-based approaches for examining ecosystem service based pathways for the sustainable alleviation of poverty; and How scenario development over the 20–40 year timescale can be advanced within the ESPA context Workshop objectives: To agree on a typology and on the concepts for models and for scenariobased tools and methods, appropriate for ESPA This is necessary to achieve the goals of the workshop To develop a general approach to using scenarios and models for ESPA To identify gaps in knowledge and understanding, or obstacles to implementing the agreed approach Define next steps for the work in ESPA OUTLINE AGENDA Day 1: Monday 22 October 12:00–1:00 Registration and lunch 1:00–1.15 Introduction 1:15–2:15 Introduction (by participants) 2:15–2:30 External speaker 2:30–2:45 External speaker: Application of the LPJmL generic ecosystem model for large-scale ecosystem service assessments 2:45–3:00 Coffee Georgina Mace Drew Purves (Microsoft Research Cambridge) Kirsten Thonicke (Potsdam Institute for Climate Impact Research) 3:00–3:15 3:15–3:30 3:30–4:00 External speaker: Developing scenarios to assess ecosystem service trade-offs – tools and case studies External speaker 4:00–4:10 4:10–5:30 Discussion (of external speakers’ presentations) Introduction to ESPA Group work 5:30–7:00 Drinks & nibbles Emily McKenzie (Natural Capital Project) Andrew Ainslie (Oxford Brookes/University of Reading and Joost Vervoort ECI Oxford University) Paul van Gardingen Introduction by Helen Suich Groups A–F Group work = This session will explore the range of approaches to scenario development and the typology of models and modelling techniques in order to establish a common framework of understanding What types of questions can you answer with these approaches? What might be the main difficulties with the definitions and concepts you are using, especially for ESPA? What new information or understanding would help to resolve these difficulties? Day 2: Tuesday 23 October 9:00–10:30 Presentations on Group Work (5 minutes Liz Carlile (chair) each) with feedback from externals and then a general discussion 10:30–10:45 Introduction to Scenarios group work Georgina Mace 10:45–11:15 Coffee 11:15–1:00 Group work Groups 1–6 1:00–2:30 Lunch (and emailing time!) 2:30–4:00 Feedback on Group Work Liz Carlile (chair) 4:00–4:15 Coffee 4:15–5:00 Externals’ comment on Group Work 5:00–5:30 Plenary discussion Liz Carlile (chair) 5:30–5:45 Wrap up – where are we now, and what are Liz Carlile we doing tomorrow Group work = What kind of processes (environmental, social and political), over what temporal and spatial scales are necessary for the issues of ecosystem services and poverty alleviation to be dealt with jointly in scenarios? What are the steps necessary to obtain plausible but ‘desired’ future scenarios, as opposed to simply modelling business as usual and the anticipated outcomes of interventions? Day 3: Wednesday 24 October 9:00–10:15 Group work Introduction by Caroline Howe Groups 7–12 10:15–11.00 11:00–11:30 11:30-12.00 12:00–12:15 12:15-1:00 1:00–1:45 13:45–14:30 14:30–15:00 15:00 Feedback on Group Work (including coffee) Externals’ comment on Group Work Plenary discussion on Group Work Introduction to ESPA understanding of ES and PA Lunch Group work Feedback from group work Wrap up and next steps End & go home! Caroline Howe & Helen Suich Groups 13–18 Georgina Mace Group work = What approaches are available that work/can be adapted to work at the spatial and temporal scales (identified in the scenarios work) appropriate for ES and PA research in order to most successfully integrate the use of models and scenarios? What are the data requirements for these models? What are the obstacles in terms of data, knowledge and techniques and how might these gaps be addressed? Group work = What is different/particular about an ESPA approach to this? What kinds of questions are these scenarios and models most likely to be useful for? How best can they be integrated (i.e scenarios and models) in the context of ES & PA research? Keynote speakers a) Kirsten Thonicke (Potsdam Institute for Climate Impact Research) Application of the LPJmL generic ecosystem model for large-scale ecosystem service assessments b) Drew Purves (Microsoft Research Cambridge) An introduction to models for ecosystem service-based research c) Emily McKenzie (Natural Capital Initiative) Developing scenarios to assess ecosystem service trade-offs – tools and case studies d) Andrew Ainslie (Oxford Brooks/University of Reading) and Joost Vervoort (ECI Oxford University) Annex 2: PARTICIPANTS Name Caroline Howe Georgina Mace Helen Suich Paul van Gardingen Ruth Swanney Liz Carlile Andy Frost Pamela Kempton Wouter Buytaert Aiden Keane Susannah Sallu Affiliation ESPA Directorate ESPA Directorate ESPA Directorate ESPA Directorate ESPA Directorate ESPA Directorate ESPA Directorate ESPA Programme Executive Board Imperial College London UCL and ZSL University of Leeds Sheila Heymans Rob D'hondt Martin Todd Johnstone Omuhaya Mark Mulligan Kerry Waylen Birgit Muller Emmanuel Nzunda Madhu Verma John Forrester Luuk Fleskens Richard Taylor Shrinivas Badiger Ricardo Crespo Brian Voigt Marco Van de Wiel Fai Fung John Gathenya Ferdinando Villa Simon Willcock Gianni Lo Iacano Dave Redding Andrew Riddick Ke Zhang Drew Purves Emily McKenzie Kirsten Thonicke Ronju Ahammad Andrew Ainslie Terry Dawson Dominic Andradi-Brown Scottish Marine Institute University of Ghent University of Sussex Kenya Marine and Fisheries Research Institute Kings College London James Hutton Institute Helmholtz Centre for Environmental Research Sokoine University of Agriculture Indian Institute of Forest Management University of York University of Leeds Stockholm Environment Institute Oxford ATREE ETH, Zurich University of Vermont University of Western Ontario University of Oxford The Rockefeller Foundation, Kenya Basque Centre for Climate Change University of Southampton University of Cambridge UCL & ZSL British Geological Survey University of Southampton Microsoft Research Natural Capital Project PIK UNDP-Bangladesh Oxford Brookes/University of Reading University of Dundee Imperial College London ... Pamela Kempton Wouter Buytaert Aiden Keane Susannah Sallu Affiliation ESPA Directorate ESPA Directorate ESPA Directorate ESPA Directorate ESPA Directorate ESPA Directorate ESPA Directorate ESPA Programme... on story lines only Often there is a combination and productive interplay with models being used to challenge scenarios, while scenarios are used to design and develop models Models and scenarios. .. Review ESPA relevant case-studies, methods and application of results with respect to scenarios and modeling Document learning on scenarios and models Facilitate access to available models and