Between integration and modularity: reviewing global Integrated Assessment Models for water–energy–land–economy nexus analysis

Introduction Integrated Assessment Models (IAMs) link human activities, emissions and the climate system, underpinning global mitigation pathways and national climate strategies. As policy questions increasingly encompass water, land and materials, the depth of integration across these domains is becoming as important as their representation. This study maps the current landscape of global IAM configurations and examines the trade-offs between integrated models and modular multi-model frameworks. Methods We conducted a documentation-based review of 21 globally scoped IAM configurations, including individual models and documented model couplings. Using explicit coding rules, we assessed 17 criteria and the integration depth of eight domains. Configurations were classified along an integration spectrum ranging from single code bases to institutionally maintained model stacks and bespoke multi-model pipelines. We also examined how these configurations represent welfare and evaluated the advantages and limitations of bespoke pipelines, using the modelling suite of the UN Sustainable Development Solutions Network (SDSN) Global Climate Hub (GCH) as an illustrative case. Results Energy is represented endogenously in 16 of the 21 configurations, whereas water is represented endogenously or through two-way links in seven. Bidirectional water–energy feedback is resolved in four configurations, based on three independent model systems, while fully endogenous three-way energy–land–water modelling is present in three. Materials remain an emerging domain: two configurations endogenise bulk-material production and two account for critical-mineral demand, but none represents mineral supply, depletion or price feedback. Welfare is assessed using output indicators, cost- or surplus-based objectives, or consumption utility; beyond-GDP indicators appear only as reported outputs. Discussion The comparison highlights trade-offs between integration depth and modularity. Bespoke multi-model pipelines offer greater sectoral depth, country-level detail and flexibility in model selection, but require substantial harmonisation and may achieve looser integration when model linkages are not iterated. The GCH modelling suite illustrates this approach. These findings are not intended to rank IAM configurations, but to provide practical guidance for model selection and identify a shared agenda for strengthening both IAMs and modular frameworks.