Conference Agenda

Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

Session Overview
307R: Large-scale behavioural models of land use change
Thursday, 25/Apr/2019:
1:30pm - 3:00pm

Session Chair: Mark Rounsevell
Session Chair: Peter Verburg
Session Chair: Calum Brown
Location: MB-205
Main Building, room 205, second floor, east wing, 90 seats
Session Topics:
How do we support transformation?

Session Abstract

The session will explore the development of the next generation of large-scale (global to continental/national scales), land-use models that are based on human behaviour, agency and decision-making processes. The purpose of these models is to explore a wide range of key research (and policy) questions at the nexus of food, ecosystems, water, climate and energy. This will support understanding of adaptation and mitigation processes within the land system as an exemplar of other socio-ecological systems. The session will provide alternatives to the current range of ‘top-down’ global models.

Since there are many different ways of modelling land use change processes, especially with respect to theories of land-use decision-making, we will explore alternative model realisations of decision processes. This includes new representations of institutional processes and their relationships with local land users and taking account of telecoupling across a globally (inter-)connected world. The session will also evaluate the coupling of large-scale, land-use models with other models types, such as Dynamic Global Vegetation Models (DGVMs), biodiversity models and/or climate emulators to explore a wide range of environmental change drivers and to evaluate the consequences of these for ecosystem services.

This session is directly relevant to the GLP topic of modelling land system change (section 4.1.4 of the GLP Science Plan), specifically the use of multi-agent models as learning-tools to test alternative conceptualisations of land system dynamics and scenario analysis. The session will also explore the GLP thematic area of telecoupling of land use systems (section 4.2.1 of the GLP Science Plan) and land governance (section 4.2.4 of the GLP Science Plan) through the development and testing of models of institutions (public policy organisations).

Session Organisers: Mark Rounsevell, Peter Verburg, and Calum Brown

Show help for 'Increase or decrease the abstract text size'
Full talk
ID: 284 / 307R: 1
307R Large-scale behavioural models of land use change
Keywords: global models, land use decision making, mapping, meta-analysis

From case-studies to global land use decision mapping for modelling

Peter Verburg, Ziga Malek

VU University Amsterdam, Netherlands, The

Global integrated assessments are strongly determined by implicit assumptions about land use decision making. Often economic optimal behaviour is assumed while an evidence base for such important assumptions is lacking. Based on a meta-analysis of land use case studies worldwide we derived a typology of decision-making strategies worldwide. While clearly distinct archetypical decision-making strategies can be distinguished this meta-analysis also revealed a general lack of in-depth analysis of decision making mechanisms in land use case-studies.

The spatial occurrence of the decision-making tyhpes was related to location and contextual factors. Overall levels of explanation of the occurrence of 6 types of decision making strategy were high, allowing for extrapolation of the relations to the global land area. This way we were able to map the likelihood of finding a specific land use decision strategy at a location. The resulting map reveals that strategies based on economic profit optimization are dominant worldwide, but in considerable areas of the world other strategies are found, often mixing economic, survival and lifestyle objectives as part of the decision making strategy.

This analysis allows to differentiate global models regionally (and temporally) based on differences in decision making strategies, leading to a more realistic inclusion of agency in global integrated assessments.

Full talk
ID: 597 / 307R: 2
307R Large-scale behavioural models of land use change
Keywords: Agent based model; climate change mitigation; adaptation; behavioural model

National to continental-scale agent-based modelling of land use change

Calum Brown, Bumsuk Seo, Mark Rounsevell

Karlsruhe Institute of Technology, Germany

Models that attempt to represent human behaviour within the land system are usually applied over relatively small geographical extents, with larger-scale applications hampered by shortages of input data and computational power. As a result, national to global scale models have to make broad, unsatisfactory assumptions about human behaviour, usually following a paradigmatic reductionist approach that emphasises the role of macro-economic drivers of land use change. However, the importance of understanding behavioural processes involved in climate change mitigation and adaptation make the scaling-up of behavioural models a clear priority.

We present national to continental (European) scale applications of an agent-based modelling framework in order to explore the feasibility and value of such an approach. We investigate the effects of different forms of behaviour at individual and institutional levels, and in a range of scenario contexts. We find that non-economic behaviours have greater effects in more extreme scenarios, and can substantially affect the provision of different ecosystem services. We also find an increasing probability of land system breakdown and food insecurity as the assumption of economic equilibrium is relaxed, with social behaviours controlling the rate of adaptation in such circumstances. Finally we explore the prospects for coupling behavioural models to biological or ecological models to describe integrated socio-ecological systems. We conclude that behavioural models can make an important contribution to the evaluation of potential developments in the land system.

Full talk
ID: 890 / 307R: 3
307R Large-scale behavioural models of land use change
Keywords: agent-based model, spatial, macro economy, decision-making, upscaling

Upscaling human behavior representation in models for large geographical regions

Tatiana Filatova1, Leila Niamir1, Olga Ivanova2

1University of Twente, Netherlands, The; 2PBL Netherlands Environmental Assessment Agency, Netherlands, The

Agent-based modeling is a prominent way to represent human behavior in coupled models of human-environment systems. This method allows for a detailed representation of various behavioral strategies of agents grounded in different theories and data, explicit modeling of their learning and adaptation behavior and of social interactions, which are essential for information and innovation diffusion. However, agent-based models (ABMs) are originally designed to be small scale. As computer science literature highlights, ABMs – especially spatial ones – are difficult to scale up. While the technical part of the challenge could be resolved by more (distributed) computational power, the architectural solutions regarding heterogeneity, interactions, coordination and synchronization of actions of a much larger population are also in demand. Importantly, as spatial ABMs scale up, other socio-economic processes than decisions of farmers or households start to play a role making it insufficient to just multiply those agents in numbers. It implies that new institutional decisions and processes relevant at larger scales have to be modelled endogenously.

There are several options to upscale a spatial ABM, including (i) expanding it to national or global scales, (ii) integrating it with another type of model designed to capture macro processes, or (iii) taking a modular approach with nested model elements, and parameterizing certain processes across scales. This talk will zoom into the second approach and will illustrate it with an example of linking a spatial ABM with a macroeconomic Computable General Equilibrium model (CGE). The spatial ABM models behavioral changes among households with respect to energy use and climate mitigation actions, which are grounded in theories from psychology and survey data. CGE estimates economy-wide impacts of such behavioral changes and in a step-wise aggregation procedure goes from regional to national and EU scales. I will discuss cons and pros of the approach and the ways forward.

Full talk
ID: 625 / 307R: 4
307R Large-scale behavioural models of land use change
Keywords: food security, scale, agency, simulation models, transformation, uncertainty

Modelling food security: bridging the gap between the micro and the macro scale

Birgit Müller1, Falk Hoffmann1, Thomas Heckelei2, Christoph Müller3, Thomas W. Hertel4, J. Gary Polhill5, Mark van Wijk6, Thom Achterbosch7, Peter Alexander8, Calum Brown9, Jiaqi Ge5, James Millington10, Ralf Seppelt1, Peter H. Verburg11, Heidi Webber2,12

1Helmholtz Centre for Environmental Research - UFZ, Leipzig, Germany; 2University of Bonn, Germany; 3Potsdam Institute for Climate Impact Research (PIK), Member of the Leibniz Association, Germany; 4Purdue University, US; 5The James Hutton Institute, Aberdeen, Scotland, UK; 6International Livestock Research Institute (ILRI), Kenya; 7Wageningen University & Research, The Netherlands; 8University of Edinburgh, UK; 9Karlsruhe Institute of Technology (KIT), Garmisch-Partenkirchen, Germany; 10King's College London, UK; 11Vrije Universiteit Amsterdam, The Netherlands; 12Leibniz Centre for Agricultural Landscape Research (ZALF), Müncheberg, Germany

Achieving food security in a changing and globalized world remains a critical challenge, especially for the most vulnerable parts of the world’s population. This endeavour can be supported by science-based modelling, transcending traditional disciplinary boundaries, of the complex interactions involved across global to household scales. Models have typically addressed selected aspects of food security (such as production, trade, or diets). Some operate at global scale (food trade equilibrium models, dynamic global vegetation models) while others focus on local scales (farm-level crop models, bio-economic models, agent-based approaches).

Macro- and micro-modelling studies can be seen as complementary. However, due to the existing gaps between scales and disciplines in current assessment frameworks, feedbacks across different levels are not properly accounted for. The inclusion of such feedbacks poses a fundamental challenge, as the modelling approaches developed to tackle each scale individually have been elaborated in different disciplines, and hence rely on different conceptual and theoretical bases.

In this contribution, we first synthesize key pieces of a fragmented landscape of food security modelling. By providing a comprehensive scheme, we portray achievements and gaps in three contextual domains of food security (production, trade, and consumption) at different spatial scales. Second, we dig deeper along three archetypical research questions which embody distinct core issues of food security: (i) uncertainty in supplies and prices, (ii) technological innovation and diffusion, and (iii) transformation of food systems. Third, as a conclusion from our reflection, we identify a set of methodological issues of relevance for current research topics in food-security modelling. These include the need for an advancement of the currently limited representation of agency by exploring the roles of new agent types in food systems, such as transnational companies, large land owners, and agribusinesses; and by enabling a more explicit representation of agency in models.

Full talk
ID: 564 / 307R: 5
307R Large-scale behavioural models of land use change
Keywords: food security, agent-based model, global trade

Exploring scenarios affecting international food and nutrition security with an agent-based model of global trade

Gary Polhill1, Jiaqi Ge1, Nuala Fitton2, Jennie Macdiarmid3, Robin Matthews1, Terry Dawson5, Heather Clark3, Stephen Whybrow3, Mukta Aphale4, Pete Smith2

1The James Hutton Institute, United Kingdom; 2University of Aberdeen, United Kingdom; 3Rowett Institute of Nutrition and Health, University of Aberdeen, United Kingdom; 4Robert Gordon University, United Kingdom; 5Kings College London, United Kingdom

The FeedUs model is an agent-based model of international food trade simulating nation states as agents. It has been developed as part of the DEVIL (Delivering Food Security on Limited Land) project to explore various scenarios influencing our ability to meet food and nutrition security demands of the projected global population in 2050. The country agents in the model aim to meet their nutritional needs through a combination of trade, domestic production of food, and depleting stocks, and produces a simulated FAO Food Balance Sheet for each country each time step. This output can be used off-line to see what the demands are for agricultural land in each country, and hence to gain some idea of the degree to which a scenario of the model has feasible food production assumptions underpinning it.

There are four scenarios to examine in DEVIL. A waste reduction scenario looks at reducing waste in both food supply and food consumption, to see how this affects international nutrition security. A dietary change scenario explores various options for food demand based on (i) transition to a Western diet; (ii) achieving nutrition security with minimum dietary change; (iii) transition to widespread adoption of plant-based diets. A sustainable intensification scenario examines the consequences of increases in maximum yield. A climate change scenario is more closely linked to land through adjusting grazing land available for livestock. Since none of the scenarios will occur in the absence of the others, the FeedUs model has been designed such that combinations of these scenarios can be brought together. The presentation will show preliminary results from the simulation of these scenarios, and the implications for food security and demand for land.