Jefferson Karthikeyan Rajah

Conventional Integrated Assessment Models (IAMs), including those used in Shared Socioeconomic Pathway (SSP) projections, often represent climate damages with limited cross-sectoral coupling, potentially missing system–wide and nonlinear risks. Policy approaches that assume gradual, reversible, and predictable change are therefore likely to underestimate both the risks of delayed action and the benefits of early, coordinated intervention. These limitations highlight the need for IAMs to incorporate increasingly comprehensive representations of climate–society feedbacks in order to better characterize climate risks under real-world conditions. To address this gap, this study investigates how explicit climate–society feedbacks alter long-term projections in a coupled human–Earth system. We use the Feedback-based knowledge Repository for Integrated Assessments version 2.1 (FRIDA v2.1), a global IAM designed to capture bidirectional feedbacks between climate and multiple socio-economic modules—Energy, Finance, Demography, Human Behaviour, Land Use, and Resource Infrastructure—via 19 climate impact channels grouped into 9 broader categories. We compare a counterfactual simulation ensemble without climate-society impacts (NoImpacts) to a fully coupled experiment with all of the impact channels (AllImpacts), and to experiments where impact channels are activated individually. Across these experiments, we find that explicit climate feedbacks fundamentally alter socioeconomic trajectories, with the AllImpacts case exhibiting substantially lower economic growth than the NoImpacts case due to cascading feedback loops that propagate through financial, energy, demographic, and resource systems. Indirect economic channels—particularly climate-induced changes in investment and bank assets—emerge as the dominant drivers of system-wide outcomes, while other impacts remain largely sector-specific. These cascading mechanisms imply a growth-damage rather than a level-damage representation of climate impacts relative to canonical IAMs (e.g., DICE), resulting in substantially larger economic losses. The analysis reveals strongly nonlinear climate-society interactions driven by cross-sectoral feedbacks, state-dependent responses, and regime-switching dynamics. Nonlinearities are particularly pronounced in food demand, crop yield, agricultural water use, and surface temperature anomaly, reflecting heterogeneous response mechanisms across coupled biophysical and socio-economic systems. These results demonstrate that tightly coupled human-Earth systems can generate non-linear system-wide changes even in the absence of explicit tipping elements.

The Feedback-based knowledge Repository for IntegrateD Assessments (FRIDA) v3.1 is a fully coupled climate-human systems model designed to explore the global long-term co-evolution of environmental, technological, economic, financial, and social processes under deep uncertainty. While many integrated assessment models represent climate damages primarily through aggregate economic impacts, important feedbacks among human behaviour, energy systems, land use, climate impacts, technological change, and financial stability remain weakly represented or exogenous. FRIDA v3.1 addresses this limitation through a system dynamics framework that embeds these processes within a unified feedback-rich structure, enabling emergent dynamics to arise from interactions across sectors rather than from externally prescribed assumptions. Building on earlier model versions, FRIDA v3.1 introduces representations of climate-driven wildfire dynamics, endogenous fertilizer use, sectorally disaggregated and behaviourally-informed energy demand, climate impacts on wind energy productivity, endogenous life expectancy, carbon capture and storage technologies, and macro-financial policy mechanisms designed to reduce climate-related financial risks. The model integrates these additions with existing representations of the climate system, carbon cycle, hydrology, population, economy, agriculture, land use, and energy supply. Calibration and evaluation against a large historical dataset and extensive uncertainty ensembles show that the updated model reproduces history while generating novel feedbacks that influence future development pathways. FRIDA v3.1 advances the capacity of integrated assessment modelling to investigate how behavioural, biophysical, technological, and financial processes jointly shape climate outcomes, providing a transparent and computationally efficient platform for uncertainty analysis, scenario exploration, and policy assessment in a rapidly changing world.

Conventional Integrated Assessment Models (IAMs), including those used in Shared Socioeconomic Pathway (SSP) projections, often represent climate damages with limited cross-sectoral coupling, potentially missing system–wide and nonlinear risks. Policy approaches that assume gradual, reversible, and predictable change are therefore likely to underestimate both the risks of delayed action and the benefits of early, coordinated intervention. These limitations highlight the need for IAMs to incorporate increasingly comprehensive representations of climate–society feedbacks in order to better characterize climate risks under real-world conditions. To address this gap, this study investigates how explicit climate–society feedbacks alter long-term projections in a coupled human–Earth system. We use the Feedback-based knowledge Repository for Integrated Assessments version 2.1 (FRIDA v2.1), a global IAM designed to capture bidirectional feedbacks between climate and multiple socio-economic modules—Energy, Finance, Demography, Human Behaviour, Land Use, and Resource Infrastructure—via 19 climate impact channels grouped into 9 broader categories. We compare a counterfactual simulation ensemble without climate-society impacts (NoImpacts) to a fully coupled experiment with all of the impact channels (AllImpacts), and to experiments where impact channels are activated individually. Across these experiments, we find that explicit climate feedbacks fundamentally alter socioeconomic trajectories, with the AllImpacts case exhibiting substantially lower economic growth than the NoImpacts case due to cascading feedback loops that propagate through financial, energy, demographic, and resource systems. Indirect economic channels—particularly climate-induced changes in investment and bank assets—emerge as the dominant drivers of system-wide outcomes, while other impacts remain largely sector-specific. These cascading mechanisms imply a growth-damage rather than a level-damage representation of climate impacts relative to canonical IAMs (e.g., DICE), resulting in substantially larger economic losses. The analysis reveals strongly nonlinear climate-society interactions driven by cross-sectoral feedbacks, state-dependent responses, and regime-switching dynamics. Nonlinearities are particularly pronounced in food demand, crop yield, agricultural water use, and surface temperature anomaly, reflecting heterogeneous response mechanisms across coupled biophysical and socio-economic systems. These results demonstrate that tightly coupled human-Earth systems can generate non-linear system-wide changes even in the absence of explicit tipping elements.

Conventional Integrated Assessment Models (IAMs), including those used in Shared Socioeconomic Pathway (SSP) projections, often represent climate damages with limited cross-sectoral coupling, potentially missing system–wide and nonlinear risks. Policy approaches that assume gradual, reversible, and predictable change are therefore likely to underestimate both the risks of delayed action and the benefits of early, coordinated intervention. These limitations highlight the need for IAMs to incorporate increasingly comprehensive representations of climate–society feedbacks in order to better characterize climate risks under real-world conditions. To address this gap, this study investigates how explicit climate–society feedbacks alter long-term projections in a coupled human–Earth system. We use the Feedback-based knowledge Repository for Integrated Assessments version 2.1 (FRIDA v2.1), a global IAM designed to capture bidirectional feedbacks between climate and multiple socio-economic modules—Energy, Finance, Demography, Human Behaviour, Land Use, and Resource Infrastructure—via 19 climate impact channels grouped into 9 broader categories. We compare a counterfactual simulation ensemble without climate-society impacts (NoImpacts) to a fully coupled experiment with all of the impact channels (AllImpacts), and to experiments where impact channels are activated individually. Across these experiments, we find that explicit climate feedbacks fundamentally alter socioeconomic trajectories, with the AllImpacts case exhibiting substantially lower economic growth than the NoImpacts case due to cascading feedback loops that propagate through financial, energy, demographic, and resource systems. Indirect economic channels—particularly climate-induced changes in investment and bank assets—emerge as the dominant drivers of system-wide outcomes, while other impacts remain largely sector-specific. These cascading mechanisms imply a growth-damage rather than a level-damage representation of climate impacts relative to canonical IAMs (e.g., DICE), resulting in substantially larger economic losses. The analysis reveals strongly nonlinear climate-society interactions driven by cross-sectoral feedbacks, state-dependent responses, and regime-switching dynamics. Nonlinearities are particularly pronounced in food demand, crop yield, agricultural water use, and surface temperature anomaly, reflecting heterogeneous response mechanisms across coupled biophysical and socio-economic systems. These results demonstrate that tightly coupled human-Earth systems can generate non-linear system-wide changes even in the absence of explicit tipping elements.

Conventional Integrated Assessment Models (IAMs), including those used in Shared Socioeconomic Pathway (SSP) projections, often represent climate damages with limited cross-sectoral coupling, potentially missing system–wide and nonlinear risks. Policy approaches that assume gradual, reversible, and predictable change are therefore likely to underestimate both the risks of delayed action and the benefits of early, coordinated intervention. These limitations highlight the need for IAMs to incorporate increasingly comprehensive representations of climate–society feedbacks in order to better characterize climate risks under real-world conditions. To address this gap, this study investigates how explicit climate–society feedbacks alter long-term projections in a coupled human–Earth system. We use the Feedback-based knowledge Repository for Integrated Assessments version 2.1 (FRIDA v2.1), a global IAM designed to capture bidirectional feedbacks between climate and multiple socio-economic modules—Energy, Finance, Demography, Human Behaviour, Land Use, and Resource Infrastructure—via 19 climate impact channels grouped into 9 broader categories. We compare a counterfactual simulation ensemble without climate-society impacts (NoImpacts) to a fully coupled experiment with all of the impact channels (AllImpacts), and to experiments where impact channels are activated individually. Across these experiments, we find that explicit climate feedbacks fundamentally alter socioeconomic trajectories, with the AllImpacts case exhibiting substantially lower economic growth than the NoImpacts case due to cascading feedback loops that propagate through financial, energy, demographic, and resource systems. Indirect economic channels—particularly climate-induced changes in investment and bank assets—emerge as the dominant drivers of system-wide outcomes, while other impacts remain largely sector-specific. These cascading mechanisms imply a growth-damage rather than a level-damage representation of climate impacts relative to canonical IAMs (e.g., DICE), resulting in substantially larger economic losses. The analysis reveals strongly nonlinear climate-society interactions driven by cross-sectoral feedbacks, state-dependent responses, and regime-switching dynamics. Nonlinearities are particularly pronounced in food demand, crop yield, agricultural water use, and surface temperature anomaly, reflecting heterogeneous response mechanisms across coupled biophysical and socio-economic systems. These results demonstrate that tightly coupled human-Earth systems can generate non-linear system-wide changes even in the absence of explicit tipping elements.

Conventional Integrated Assessment Models (IAMs), including those used in Shared Socioeconomic Pathway (SSP) projections, often represent climate damages with limited cross-sectoral coupling, potentially missing system–wide and nonlinear risks. Policy approaches that assume gradual, reversible, and predictable change are therefore likely to underestimate both the risks of delayed action and the benefits of early, coordinated intervention. These limitations highlight the need for IAMs to incorporate increasingly comprehensive representations of climate–society feedbacks in order to better characterize climate risks under real-world conditions. To address this gap, this study investigates how explicit climate–society feedbacks alter long-term projections in a coupled human–Earth system. We use the Feedback-based knowledge Repository for Integrated Assessments version 2.1 (FRIDA v2.1), a global IAM designed to capture bidirectional feedbacks between climate and multiple socio-economic modules—Energy, Finance, Demography, Human Behaviour, Land Use, and Resource Infrastructure—via 19 climate impact channels grouped into 9 broader categories. We compare a counterfactual simulation ensemble without climate-society impacts (NoImpacts) to a fully coupled experiment with all of the impact channels (AllImpacts), and to experiments where impact channels are activated individually. Across these experiments, we find that explicit climate feedbacks fundamentally alter socioeconomic trajectories, with the AllImpacts case exhibiting substantially lower economic growth than the NoImpacts case due to cascading feedback loops that propagate through financial, energy, demographic, and resource systems. Indirect economic channels—particularly climate-induced changes in investment and bank assets—emerge as the dominant drivers of system-wide outcomes, while other impacts remain largely sector-specific. These cascading mechanisms imply a growth-damage rather than a level-damage representation of climate impacts relative to canonical IAMs (e.g., DICE), resulting in substantially larger economic losses. The analysis reveals strongly nonlinear climate-society interactions driven by cross-sectoral feedbacks, state-dependent responses, and regime-switching dynamics. Nonlinearities are particularly pronounced in food demand, crop yield, agricultural water use, and surface temperature anomaly, reflecting heterogeneous response mechanisms across coupled biophysical and socio-economic systems. These results demonstrate that tightly coupled human-Earth systems can generate non-linear system-wide changes even in the absence of explicit tipping elements.

Conventional Integrated Assessment Models (IAMs), including those used in Shared Socioeconomic Pathway (SSP) projections, often represent climate damages with limited cross-sectoral coupling, potentially missing system–wide and nonlinear risks. Policy approaches that assume gradual, reversible, and predictable change are therefore likely to underestimate both the risks of delayed action and the benefits of early, coordinated intervention. These limitations highlight the need for IAMs to incorporate increasingly comprehensive representations of climate–society feedbacks in order to better characterize climate risks under real-world conditions. To address this gap, this study investigates how explicit climate–society feedbacks alter long-term projections in a coupled human–Earth system. We use the Feedback-based knowledge Repository for Integrated Assessments version 2.1 (FRIDA v2.1), a global IAM designed to capture bidirectional feedbacks between climate and multiple socio-economic modules—Energy, Finance, Demography, Human Behaviour, Land Use, and Resource Infrastructure—via 19 climate impact channels grouped into 9 broader categories. We compare a counterfactual simulation ensemble without climate-society impacts (NoImpacts) to a fully coupled experiment with all of the impact channels (AllImpacts), and to experiments where impact channels are activated individually. Across these experiments, we find that explicit climate feedbacks fundamentally alter socioeconomic trajectories, with the AllImpacts case exhibiting substantially lower economic growth than the NoImpacts case due to cascading feedback loops that propagate through financial, energy, demographic, and resource systems. Indirect economic channels—particularly climate-induced changes in investment and bank assets—emerge as the dominant drivers of system-wide outcomes, while other impacts remain largely sector-specific. These cascading mechanisms imply a growth-damage rather than a level-damage representation of climate impacts relative to canonical IAMs (e.g., DICE), resulting in substantially larger economic losses. The analysis reveals strongly nonlinear climate-society interactions driven by cross-sectoral feedbacks, state-dependent responses, and regime-switching dynamics. Nonlinearities are particularly pronounced in food demand, crop yield, agricultural water use, and surface temperature anomaly, reflecting heterogeneous response mechanisms across coupled biophysical and socio-economic systems. These results demonstrate that tightly coupled human-Earth systems can generate non-linear system-wide changes even in the absence of explicit tipping elements.

Demand-side strategies, including shifts toward lower-impact diets, are increasingly recognized as critical levers for climate mitigation. Yet most model-based assessments represent future dietary demand through exogenous trajectories or narrative-consistent assumptions, limiting their ability to capture feedback-driven behavioural change. Here, we use the fully coupled FRIDAv2.1 integrated assessment model, which incorporates endogenous feedback-based representations of dietary behaviours, to examine the conditions under which earlier and more sustained pro-environmental dietary change can emerge. Through variance-based sensitivity analysis, we identify the behavioural-change parameters that most strongly shape global diet size and composition. The results show that influential parameters cluster around three mechanisms: socioeconomic accessibility, sociocultural valuation, and risk-linked personal norms. Scenario experiments that perturb these mechanisms show that earlier and more durable reductions in animal-product and total food demand emerge most clearly when multiple levers are activated together, whereas isolated interventions remain vulnerable to attenuation and backsliding. Although diet shifts yield meaningful reductions in agricultural emissions, their effects on long-term global temperature and sea-level rise remain limited without concurrent reductions in energy-related emissions. Overall, the analysis shows that endogenizing behavioural change reveals non-linear, path-dependent, and reversible demand-side mitigation dynamics that are obscured in exogenously specified demand pathways. These findings indicate that demand-side climate mitigation pathways cannot be adequately evaluated as smooth exogenous demand reductions; they must be understood as feedback-driven processes that can spread, stall, and reverse over time.

Conventional Integrated Assessment Models (IAMs), including those used in Shared Socioeconomic Pathway (SSP) projections, often represent climate damages with limited cross-sectoral coupling, potentially missing system–wide and nonlinear risks. Policy approaches that assume gradual, reversible, and predictable change are therefore likely to underestimate both the risks of delayed action and the benefits of early, coordinated intervention. These limitations highlight the need for IAMs to incorporate increasingly comprehensive representations of climate–society feedbacks in order to better characterize climate risks under real-world conditions. To address this gap, this study investigates how explicit climate–society feedbacks alter long-term projections in a coupled human–Earth system. We use the Feedback-based knowledge Repository for Integrated Assessments version 2.1 (FRIDA v2.1), a global IAM designed to capture bidirectional feedbacks between climate and multiple socio-economic modules—Energy, Finance, Demography, Human Behaviour, Land Use, and Resource Infrastructure—via 19 climate impact channels grouped into 9 broader categories. We compare a counterfactual simulation ensemble without climate-society impacts (NoImpacts) to a fully coupled experiment with all of the impact channels (AllImpacts), and to experiments where impact channels are activated individually. Across these experiments, we find that explicit climate feedbacks fundamentally alter socioeconomic trajectories, with the AllImpacts case exhibiting substantially lower economic growth than the NoImpacts case due to cascading feedback loops that propagate through financial, energy, demographic, and resource systems. Indirect economic channels—particularly climate-induced changes in investment and bank assets—emerge as the dominant drivers of system-wide outcomes, while other impacts remain largely sector-specific. These cascading mechanisms imply a growth-damage rather than a level-damage representation of climate impacts relative to canonical IAMs (e.g., DICE), resulting in substantially larger economic losses. The analysis reveals strongly nonlinear climate-society interactions driven by cross-sectoral feedbacks, state-dependent responses, and regime-switching dynamics. Nonlinearities are particularly pronounced in food demand, crop yield, agricultural water use, and surface temperature anomaly, reflecting heterogeneous response mechanisms across coupled biophysical and socio-economic systems. These results demonstrate that tightly coupled human-Earth systems can generate non-linear system-wide changes even in the absence of explicit tipping elements.

Conventional Integrated Assessment Models (IAMs), including those used in Shared Socioeconomic Pathway (SSP) projections, often represent climate damages with limited cross-sectoral coupling, potentially missing system–wide and nonlinear risks. Policy approaches that assume gradual, reversible, and predictable change are therefore likely to underestimate both the risks of delayed action and the benefits of early, coordinated intervention. These limitations highlight the need for IAMs to incorporate increasingly comprehensive representations of climate–society feedbacks in order to better characterize climate risks under real-world conditions. To address this gap, this study investigates how explicit climate–society feedbacks alter long-term projections in a coupled human–Earth system. We use the Feedback-based knowledge Repository for Integrated Assessments version 2.1 (FRIDA v2.1), a global IAM designed to capture bidirectional feedbacks between climate and multiple socio-economic modules—Energy, Finance, Demography, Human Behaviour, Land Use, and Resource Infrastructure—via 19 climate impact channels grouped into 9 broader categories. We compare a counterfactual simulation ensemble without climate-society impacts (NoImpacts) to a fully coupled experiment with all of the impact channels (AllImpacts), and to experiments where impact channels are activated individually. Across these experiments, we find that explicit climate feedbacks fundamentally alter socioeconomic trajectories, with the AllImpacts case exhibiting substantially lower economic growth than the NoImpacts case due to cascading feedback loops that propagate through financial, energy, demographic, and resource systems. Indirect economic channels—particularly climate-induced changes in investment and bank assets—emerge as the dominant drivers of system-wide outcomes, while other impacts remain largely sector-specific. These cascading mechanisms imply a growth-damage rather than a level-damage representation of climate impacts relative to canonical IAMs (e.g., DICE), resulting in substantially larger economic losses. The analysis reveals strongly nonlinear climate-society interactions driven by cross-sectoral feedbacks, state-dependent responses, and regime-switching dynamics. Nonlinearities are particularly pronounced in food demand, crop yield, agricultural water use, and surface temperature anomaly, reflecting heterogeneous response mechanisms across coupled biophysical and socio-economic systems. These results demonstrate that tightly coupled human-Earth systems can generate non-linear system-wide changes even in the absence of explicit tipping elements.