We propose a wish list of features that would greatly enhance population synthesis methods from the perspective of agent-based modelling. The challenge of synthesising appropriate populations is heightened in agent-based modelling by the emphasis on complexity, which requires accounting for a wide array of features. These often include, but are not limited to: attributes of agents, their location in space, the ways they make decisions and their behavioural dynamics. In the real-world, these aspects of everyday human life can be deeply interconnected, with these associations being highly consequential in shaping outcomes. Initialising synthetic populations in ways that fail to respect these covariances can therefore compromise model efficacy, potentially leading to biased and inaccurate simulation outcomes.

Agent-based models (ABMs) are increasingly utilized in ecology and related fields, yet concerns persist regarding the lack of consideration for lessons learned from previous models. This study explores the potential of systematically conducted ABM reviews to contribute to cumulative science and theory development by synthesizing individual ABM findings more effectively. We are conducting a meta-review of ABM reviews to assess current practices, compare them to systematic literature review (SLR) literature recommendations, and evaluate their engagement with theory and theory development. Our analysis of the ecology and social science sample reveals that many reviews are not conducted systematically and lack transparency. The analysis step of SLRs holds significant potential to advance theory development. Reviews primarily focus on model design, while other avenues of theory development receive less attention. Our findings suggest ways to improve current practices and may guide future ABM reviews via benchmarks for methodological decisions and dimensions for advancing theory development.

Agent-based models are particularly suitable to reflect the dynamics of humans, nature, and their interactions, making them a crucial approach for understanding social-ecological systems. The formalisations of human decision-making are central to resulting model behaviours. Despite awareness of the complexity of human behaviour in social-ecological systems research, scholars tend to represent human decision-makers as simplified, perfectly informed rational optimisers, without explicitly considering the fit with decision context. Key reasons are a lacking uptake of social theories and insights. To advance, we need a practice of reflecting, sharing, and inquiring on the justification of the decision model fit with its context. This paper stimulates this practice by 1) supporting the justification of decision model (DM) fit by describing the DM landscape and providing guiding questions; and 2) by supporting researchers in considering alternative DMs through a survey-based impression of modeller practices, and through highlighting DM frontiers as inspiration for future research.

There has so far been no shared understanding of validity in agent-based simulation. We here conceptualise validation as systematically substantiating the premises on which conclusions from simulation analysis for a particular modelling context are built. Given such a systematic perspective, validity of agent-based models cannot be ensured if validation is merely understood as an isolated step in the modelling process. Rather, valid conclusions from simulation analysis require context-adequate method choices at all steps of the simulation analysis including model construction, model and parameter inference, uncertainty analysis and simulation. We present a twelve-step protocol to highlight the (often hidden) premises for methodological choices and their link to the modelling context. It is designed to aid modelers in understanding their context and in choosing and documenting context-adequate and mutually consistent methods throughout the modelling process. Its purpose is to assist reviewers and the community as a whole in assessing and discussing context-adequacy.

While agent-based modeling (ABM) has become one of the most powerful tools in quantitative social sciences, it remains difficult to explain their structure and performance. We propose to use artificial intelligence both to build the models from data, and to improve the way we communicate models to stakeholders. Although machine learning is actively employed for pre-processing data, here for the first time, we used it to facilitate model development of a simulation model directly from data. Our suggested framework, ML-ABM accounts for causality and feedback loops in a complex nonlinear system and at the same time keeps it transparent for stakeholders. As a result, beside the development of a behavioral ABM, we open the ‘blackbox’ of purely empirical models. With our approach, artificial intelligence in the simulation field can open a new stream in modeling practices and provide insights for future applications.

Abrupt systemic changes in ecological and socio-economic systems are a regular occurrence. While there has been much attention to studying systemic changes primarily in ecology as well as in economics, the attempts to do so for coupled socio-environmental systems are rarer. This paper bridges the gap by reviewing how models can be instrumental in exploring significant, fundamental changes in such systems. The history of modelling systemic change in various disciplines contains a range of definitions and approaches. Even so, most of these efforts share some common challenges within the modelling context. We propose a framework drawing these challenges together, and use it to discuss the articles in this thematic issue on modelling systemic change in coupled social and environmental systems. The differing approaches used highlight that modelling systemic change is an area of endeavour that would benefit from greater synergies between the various disciplines concerned with systemic change.

Increasing attention to regime shifts, critical transitions, non-marginal changes, and systemic shocks calls for the development of models that are able to reproduce or grow structural changes that occur over time periods perceived as abrupt. This paper highlights specific modelling challenges to consider when exploring coupled socio-environmental systems experiencing regime shifts. We explore these challenges in the context of four modelling approaches that have been applied to the study of regime shifts in coupled socio-environmental systems: statistical, system dynamics, equilibrium and agent-based modelling. When reviewing these modelling approaches we reflect on a set of criteria including the ability of an approach (1) to capture feedbacks between social and environmental system, (2) to represent the sources of regime shifts, (3) to incorporate complexity aspects, and (4) to deal with regime shift identification. Many of the modelling examples considered do not provide information on all these criteria, which receive a lot of attention in empirical studies of registered regime shifts. This suggests a need to develop a common modelling terminology in the domain of modelling for resilience and regime shifts. When discussing strengths and weaknesses of various modelling paradigms we conclude that a hybrid approach is likely to provide most insights into the processes and consequences of regime shifts. Challenges and frontier directions of research for designing models to study regime shifts in coupled socio-environmental systems are outlined.

The proliferation of agent-based models (ABMs) in recent decades has motivated model practitioners to improve the transparency, replicability, and trust in results derived from ABMs. The complexity of ABMs has risen in stride with advances in computing power and resources, resulting in larger models with complex interactions and learning and whose outputs are often high-dimensional and require sophisticated analytical approaches. Similarly, the increasing use of data and dynamics in ABMs has further enhanced the complexity of their outputs. In this article, we offer an overview of the state-of-the-art approaches in analyzing and reporting ABM outputs highlighting challenges and outstanding issues. In particular, we examine issues surrounding variance stability (in connection with determination of appropriate number of runs and hypothesis testing), sensitivity analysis, spatio-temporal analysis, visualization, and effective communication of all these to non-technical audiences, such as various stakeholders.

Coupled socio-ecological systems (SES) are complex systems characterized by self-organization, non-linearities, interactions among heterogeneous elements within each subsystem, and feedbacks across scales and among subsystems. When such a system experiences a shock or a crisis, the consequences are difficult to predict. In this paper we first define what a shock or a crisis means for SES. Depending on where the system boundary is drawn, shocks can be seen as exogenous or endogenous. For example, human intervention in environmental systems could be seen as exogenous, but endogenous in a socio-environmental system. This difference in the origin and nature of shocks has certain consequences for coupled SES and for policies to ameliorate negative consequences of shocks. Having defined shocks, the paper then focuses on modelling challenges when studying shocks in coupled SES. If we are to explore, study and predict the responses of coupled SES to shocks, the models used need to be able to accommodate (exogenous) or produce (endogenous) a shock event. Various modelling choices need to be made. Specifically, the 'sudden' aspect of a shock suggests the time period over which an event claimed to be a shock occurred might be 'quick'. What does that mean for a discrete event model? Turning to magnitude, what degree of change (in a variable or set of variables) is required for the event to be considered a shock? The 'surprising' nature of a shock means that none of the agents in the model should expect the shock to happen, but may need rules enabling them to generate behaviour in exceptional circumstances. This requires a certain design of the agents' decision-making algorithms, their perception of a shock, memory of past events and formation of expectations, and the information available to them during the time the shock occurred.