Social Simulation Projects

Projects using agent-based models [1] at The James Hutton Institute comprise a portfolio of work funded by the Scottish Government [2]'s Strategic Research Programme [3] (2011-2016 and 2016-2021), the European Commission [4] (Framework Programme 7 and Horizon 2020), the Macaulay Development Trust [5], Research Councils [6] in the UK, and the Research Council of Norway [7].

Resilience of Rural Industry Supply Chains

In the 2016-2021 Strategic Research Programme, we have been working on simulating the potential effects of systemic disturbances such as changes in trade agreements on Scotland's beef and dairy value chains. Initial work has focused on the effect of Brexit [8] on the phenomenon of the 'disappearing middle [9]' in Scottish farms, in which there has been a long-term trend of a decline in the number medium-sized farms together with a growth in the numbers of small and large farms. A screenshot from the model is shown below. It features farm business agents specializing in various areas of cattle production in the beef and dairy value chains. The model is being used to develop methods for empirical validation using Approximate Bayesian Computation.

SMARTEES

Funded by Horizon 2020, the SMARTEES [10] project has been exploring how social innovations can work to bring about the transition to the energy union in ways that are sensitive to local conditions and foster citizen participation. The Hutton's modelling work in SMARTEES centres on case studies in Aberdeen and Timişoara, Romania on the adopting of district heating to address fuel poverty; and in Stockholm and Malmö on engaging citizens in district regeneration. The screenshot below shows a hypothetical heat-network adoption scenario from a model simulating the Torry area of Aberdeen using an artificial population of agents.

Global Food Security

The James Hutton Institute's participation in the DEVIL [11] project on delivering food security on limited land has been funded by the 2016-2021 Strategic Research Programme. Our work in this project has focused on the development of a model of global trade in food and nutrition, capable of exploring scenarios relating to climate change, waste reduction, dietary change and sustainable intensification. The model [12] simulates agents that represent countries, using UN FAO [13] data. The countries try to achieve the goal of obtaining enough food to feed their population through production and trade, affected by each of the scenarios. The model, a screenshot from which is shown below, is currently being used in a project [14] exploring the resilience of the UK's food industry in the light of the COVID crisis.

FEARLUS-SPOMM

In the 2011-2016 Strategic Research Programme, we coupled a legacy model of agricultural land use change (FEARLUS) with a biodiversity model (SPOMM). This allowed us to explore scenarios of incentivizing biodiversity [15] in agricultural landscapes, including payment-by-results (as opposed to payment for specific activities) and 'clustered' payments where groups of land owners owning neighbouring land are given extra money if they all have target species, or undertake activities creating habitat for them. Our main result is that, above a certain threshold in incentivization amount, payment-by-results is most robust to other influences on decision-making by the agents in the model. Analysing the outputs of agent-based models is a significant challenge, as we typically run them several thousands (sometimes even millions) of times. In this work, we used machine learning algorithms to process the outputs. The screenshot below shows FEARLUS-SPOMM running.

[16]

GLAMURS

The GLAMURS [17] project, funded by the European Commission's 7th Framework Programme, studied options for greening lifestyles. Hutton's agent-based modelling work in GLAMURS concentrated on case studies in Northeast Scotland. Using a traffic model, we compared flexible working arrangements with new infrastructure [18] for their effects on tailpipe emissions on traffic and mean commute times (see screenshot below), and also explored hypothetical workplace sharing schemes [19]. With a larger-scale lifestyle transitions model, we examined options for a 'green recovery [20]' in Aberdeen following the post-2014 decline in the oil industry.