FELiX Regional Model: From Global Visions to Nuanced Pathways
Introduction: A Zoomed-In Future
FeliX is a globally aggregated, feedback-rich simulation model of climate, economy, environment and society interactions, developed at International Institute for Applied Systems Analysis (IIASA) originally for the purpose of supporting Earth observations. It represents global systems including population, education, economy, dietary changes, land use, energy, water, carbon cycle, climate, and biodiversity, as well as the feedback between them and provides future trajectories of these systems under different scenarios.
A regional version of FeliX (FeliX-R5) has been recently developed to address the limitations of the original global-scale FeliX in capturing socioeconomic and cultural differences across world regions. Given high heterogeneity of social, environmental and economic factors across regions and nations, such as local food systems, policies and behaviors varying widely, a regionalised version was highly needed for finer-scale analysis of socioeconomic-environmental dynamics and solutions tailored to specific geographic contexts. Still, maintaining a computationally efficient and traceable feedback-rich model required limiting the resolution level. This is why a five regional disaggregation is chosen. FeliX-R5 breaks the world down into five regions according to the United Nations’ regional classifications: Africa, Asia-Pacific, Eastern Europe, Latin America and Caribbean, Western Europe and other (developed) states. Ten key human-nature systems, such as population (including labor forces), economy, dietary changes, land use, are adapted to reflect regional variability, including the interregional trade of food and energy sources between the regions in a system dynamic setting.
What the Regional Model Does
To regionalise the FeliX model, we kept the model structure the same, but calibrated each region’s model to its observed patterns using historical data. For this purpose, we first created a large dataset covering regional and national historical data of key variables, such as population, energy production or land use change. This involved collecting, cleaning, and processing data from different sources such as FAOSTAT (Food and Agriculture Organization statistics), IEA or IRENA.
Trade becomes critically important in a regionalized model setting to capture the complex interactions and dependencies among regions. We developed an import-based trade mechanism to model the trade of fossil fuels and food products among the five regions. This also allows modelers to trace environmental responsibilities along the supply chain and quantify how one region’s policy may influence economic outcomes, production adjustments, and emissions in others.
The main challenge in this regionalization process was the complexity of the model with many connections and feedbacks between variables from different modules. This made the calibration process multi-dimensional calibration process very time consuming, especially because we strived for logical consistency besides numerical accuracy in the regional model as well.
The resulting regional model simulates food demand resulting from heterogenous population and diet shift dynamics, food security, land use change, CO2 emissions (not only from AFOLU but also energy) and biodiversity pathways. It captures the main socioeconomic and cultural characteristics of CHOICE pilot countries (Austria, Greece, South Africa and Colombia) in respective regions (Western Europe, Africa, Latin America), and hence enables exploring the implications of CHOICE behavior change scenarios.
Figure 1: Overview of the FeliX-R5 model, adapted from Moallemi et al. (2022)
Use in CHOICE
In the CHOICE project, FeliX is used for two main purposes:
- Representing food-related lifestyle changes emerging from the interaction of cognitive, social, environmental and economic factors, hence exploring behavior–driven food demand scenarios;
- Developing an online interactive simulation environment where the users can explore various behavior change scenarios and their outcomes.
FeliX-R5 will particularly serve for the first purpose and create behavior-driven food demand scenarios based on scenario narratives developed in CHOICE either by the modeling teams or in pilot countries using FABLE calculator. FeliX-R5 will particularly shed light on the feedback-based (e.g. climate impacts on economy, population, or dietary shifts behaviors) implications of these scenarios beyond the AFOLU sector.
Key Takeaways
The regionalised FeliX (FeliX-R5) provides a more flexible framework for scenario analysis targeting region-specific policy assessment and decision-making. FeliX-R5 supports quantification of the mitigation potential of food system interventions analyzed in CHOICE pilots, such as dietary behavior shifts, reduction of food loss and waste, and increased livestock productivity. It provides an assessment of these interventions in different geographic regions from an economy-wide and system dynamics perspective including the energy system and non-land-based emissions.
Open Dataset accessibility
- Online documentation: https://iiasa.github.io/felix_docs/
- Online repository for open-access model: https://github.com/iiasa/Felix-Model
Relevant publications:
- Eker, S., et al. (2025). Full of Economic-Environment Linkages and Integration dX/dt (FeliX): Technical Model Documentation. Available at https://github.com/iiasa/Felix-Model/tree/master/Documentation
- Ye, Q., et al. (2024). FeliX 2.0: An integrated model of climate, economy, environment, and society interactions. Environmental Modelling & Software 179, 106121. https://doi.org/10.1016/j.envsoft.2024.106121
References
Moallemi, E.A., Eker, S., Gao, L., Hadjikakou, M., Liu, Q., Kwakkel, J., Reed, P.M., Obersteiner, M., Guo, Z., Bryan, B.A., 2022. Early systems change necessary for catalyzing long-term sustainability in a post-2030 agenda. One Earth, 5, 792–811. DOI: https://doi.org/10.1016/j.oneear.2022.06.003
