ABMs are increasingly applied in energy transition research due to their capacity to represent complex socio-technical systems. However, the integration of behavioural insights remains limited: models typically include a narrow set of psycho-logical determinants, rely on ad hoc decision rules, and provide limited transparency on how behavioural assumptions are selected. This is particularly problematic when ABMs are used for policy support, where a strong alignment between behavioural determinants and the simulated behaviour, referred to as the determinant–behaviour fit, is required. This paper introduces the Behavioural Analysis (BA) method: a systematic, ex ante procedure to identify, select, and prioritize behavioural determinants in line with empirical evidence, behavioural theory, and purpose of the ABM. The BA consists of five steps: defining the model goal, specifying the target behaviour and actor group, mapping candidate determinants across psychological, contextual, and social domains, selecting relevant behavioural theory, and developing an empirical strategy for operationalization. The method is demonstrated for household load shifting in the context of grid congestion. Current load shifting models predominantly focus on price responsiveness, while key behavioural determinants of routine behaviour, such as habit strength, remain underrepresented. Applying the BA leads to a theoretically grounded se-lection of behavioural determinants and highlights the interaction between hassle and habit as a critical but underexplored mechanism. The BA approach enhances transparency, theoretical grounding, and empirical traceability in ABM design, improving its decision support power to policy design.

Climate change is posing threats to cities through increasing severity of flood events and heatwaves, as well as through the acceleration of biodiversity loss. Nature-based solutions (NbS) are increasingly recognised as an important strategy for addressing these threats, yet their implementation remains limited. This is especially critical in cities, where space for green NbS is scarce and most of the land is in private ownership. This paper investigates how large-scale NbS adoption by households can be achieved, leading to cumulative city-wide climate adaptation. To this end, we developed a spatially explicit theory- and data-grounded agent-based model (ABM), which simulates the decision-making process of households regarding the installation of green roofs and raingardens on their property. The formalisation is performed in two phases. Phase I results in an ABM grounded in the Theory of Planned Behaviour and the Value Belief Norm theory and calibrated using secondary data. This model is further advanced in Phase II with data from our tailored NbS survey among Dutch households. Using the ABM, we explore the mechanisms shaping outscaling and deepscaling of household level NbS, investigate whether socio-economic tipping points exist, and explore what policy levers could be used to accelerate private NbS implementation in cities.

Increasing urban heat stress places growing pressure on households to adapt, often in contexts of uneven capacity and fragmented institutional support. Collective action is therefore increasingly discussed as a pathway for collective heat adaptation and more just outcomes. This study examines how community-based adaptation (CBA) and collective-led adaptation (CLA) shape household and neighborhood heat responses, with a strong focus on agent-based modeling systems (ABMS) for social simulation.

The research adopts a mixed-method design combining a household survey, interviews with decision-makers, and a rule-based ABMS. Protection Motivation Theory and social capital theory inform survey construction and model rules, capturing how perceived heat risk, adaptive capacity, social ties, trust, reciprocity, and perceptions of justice influence participation in collective action. Empirical data are used to parameterize the ABMS, enabling simulations of interactions between households, neighborhoods, and facilitators across short-term heat events and long-term adaptation dynamics.

The ABMS conceptualizes collective heat adaptation as a complex adaptive system, allowing exploration of emergence and sustainability under different scenarios. By explicitly incorporating perceptions of fairness and justice, the model examines how collective approaches distribute adaptation responsibilities across groups and scales. The case study in Antwerp provides a concrete context for testing how CBA and CLA influence collective heat adaptation.

Keywords: CBA, CLA, ABMS, collective heat adaptation, justice.

Agent-Based Modelling has increasingly been employed in inter- and transdisciplinary research settings. While Agent-Based Models can integrate diverse data, theories, and stakeholder perspectives, the collaborative and iterative nature of their development remains largely under-documented, limiting transparency, reflexivity, and cumulative learning. This paper introduces InTrA, a structured reflection framework designed to address these challenges. Conceptualized at the intersection of three domains (i.e., Agent-Based Modelling understood as a socio-technical system; inter-/transdisciplinary knowledge production; and practice of reflecting on experience) InTrA Reflection Framework provides a principled approach for documenting and reflecting on transdisciplinary modelling processes. It is built around three core principles: (1) Agent-Based Modelling is a multi-phase, iterative process, with each phase, including its steps and sub-steps, open to systematic documentation; (2) social and epistemic integration can occur at any phase; and (3) reflection on experience can target any element of the Agent-Based Modelling socio-technical system (people, goals, culture etc.). By guiding modellers to make explicit decisions, assumptions, and collaborative interactions, InTrA fosters transparency, individual as well as group learning, and knowledge integration. Through the framework’s personal and subjective nature encouraging deep, context-sensitive reflection, we hope that InTrA can be used as a tool toward making Agent-Based Modelling processes more transparent and satisfactory for the involved parties, supporting methodological rigor, the integration of knowledge in transdisciplinary contexts, and long-term skill and community development.