Realizing adaptive governance of coupled climate–social systems: A deep reinforcement learning framework

Abstract

The escalating pressure of global warming necessitates practical and effective climate governance policies to meet the urgent goals of theParis Agreement. However, for conventional policy simulation models like integrated assessment models (IAMs), their reliance on predetermined scenarios makes it challenging to adequately address the substantial uncertainties inherent in climate-social system evolution. This study proposes an adaptive decision-making framework that integrates deep reinforcement learning (DRL) with a climate–social system model to identify governance strategies that can reduce the risk of transgressing planetary boundaries. In this framework, we operationalize Social Tipping Elements (STEs) as targeted actions. A reward function provides feedback to optimize these interventions, ensuring the system remains within critical planetary boundaries. Our results demonstrate that, compared to conventional static models, the framework discovers adaptive policies through dynamic learning, enabling real-time responses to evolving climate-social conditions. These adaptive policies exhibit an “early-stage intensive intervention, mid-term moderation, and late-stage reinforcement” pattern that reduces planetary boundary overshoot time by 55 years while stabilizing global warming below 1.5 °C by 2100. By varying governance objectives, we further reveal the trade-off mechanisms between competing climate and socioeconomic goals. Notably, the designed multi-objective reward function enables a synergistic balance across competing objectives, resulting in the shortest planetary boundary overshoot duration (15 years) among all evaluated scenarios. This framework overcomes the limitations of static simulations, offering a robust and interpretable tool to design adaptive strategies crucial for navigating the competing objectives of time-sensitive climate governance.

Keywords

Deep reinforcement learning / Complex adaptive system / Earth governance / Planetary boundary / Social tipping elements / Decision-making

References

[1] An, L., Turner II, B.L., Liu, J.G., Grimm, V., Zhang, Q., Wang, Z.Y., Huang, R.H., 2025. Complex adaptive systems science in the era of global sustainability crisis. Geogr. Sustain. 6 (1), 100250. doi: 10.1016/j.geosus.2024.09.011.

[2] Allen, C., Metternicht, G., Wiedmann, T., Pedercini, M., 2021. Modelling national transformations to achieve the SDGs within planetary boundaries in small island developing states. Glob. Sustain. 4, e15. doi: 10.1017/sus.2021.13.

[3] Bouckaert, S., Pales, A.F., McGlade, C., Remme, U., Wanner, B., Varro, L., D’Ambrosio, D., Spencer, T., 2021. Net Zero by 2050: A Roadmap for the Global Energy Sector. International Energy Agency (IEA), Paris.

[4] Castelletti, A., Galelli, S., Restelli, M., Soncini-Sessa, R., 2010. Tree-based reinforcement learning for optimal water reservoir operation. Water Resour. Res. 46 (9), 2009WR008898. doi: 10.1029/2009WR008898.

[5] Chen, J.W., Cui, H.J., Xu, Y.Y., Ge, Q.S., 2020. An investigation of parameter sensitivity of minimum complexity earth simulator. Atmosphere 11 (1), 95. doi: 10.3390/atmos11010095.

[6] Chen, M.T., Cui, Y.L., Wang, X.N., Xie, H.W., Liu, F.P., Luo, T.Y., Zheng, S.Z., Luo, Y.F., 2021. A reinforcement learning approach to irrigation decision-making for rice using weather forecasts. Agric. Water Manage. 250, 106838. doi: 10.1016/j.agwat.2021.106838.

[7] Cheng, G.D., Li, X., 2015. Integrated research methods in watershed science. Sci. China Earth Sci. 58 (7), 1159-1168. doi: 10.1007/s11430-015-5074-x.

[8] Conijn, J.G., Bindraban, P.S., Schröder, J.J., Jongschaap, R.E.E., 2018. Can our global food system meet food demand within planetary boundaries? Agric. Ecosyst. Environ. 251, 244-256. doi: 10.1016/j.agee.2017.06.001.

[9] Dariane, A.B., Moradi, A.M., 2016. Comparative analysis of evolving artificial neural network and reinforcement learning in stochastic optimization of multireservoir systems. Hydrol. Sci. J. 61 (6), 1141-1156. doi: 10.1080/02626667.2014.986485.

[10] Ding, L., Bai, Y.L., Zheng, D.H., Pan, X.D., Fan, M.H., Li, X., 2025. Chaotic climate system forecasting using an improved echo state network with sparse observations. Sci. China Earth Sci. 68 (7), 2346-2360. doi: 10.1007/s11430-024-1593-9.

[11] Edenhofer, O., 2015. Climate Change 2014: Mitigation of Climate Change. Cambridge University Press, Cambridge.

[12] Eker, S., Wilson, C., Höhne, N., McCaffrey, M.S., Monasterolo, I., Niamir, L., Zimm, C., 2024. Harnessing social tipping dynamics: a systems approach for accelerating decarbonization. One Earth 7 (6), 976-988. doi: 10.1016/j.oneear.2024.05.012.

[13] ENERGY STAR, 2017. The Light Bulb Revolution. U.S. Environmental Protection Agency, Washington, D.C.

[14] Feng, K.R., Lin, N., Kopp, R.E., Xian, S.Y., Oppenheimer, M., 2025. Reinforcement learning-based adaptive strategies for climate change adaptation: an application for coastal flood risk management. Proc. Natl. Acad. Sci. U.S.A. 122 (12), e2402826122. doi: 10.1073/pnas.2402826122.

[15] Flatow, I., 2012. Fifty Years Ago, a Bright Idea. NPR.

[16] François-Lavet, V., Henderson, P., Islam, R., Bellemare, M.G., Joelle, P., 2018. An introduction to deep reinforcement learning. Found. Trends Mach. Learn. 11 (3-4), 219-354. doi: 10.1561/2200000071.

[17] Hung, F., Yang, Y.C.E., 2021. Assessing adaptive irrigation impacts on water scarcity in nonstationary environments-a multi-agent reinforcement learning approach. Water Resour. Res. 57 (9), e2020WR029262. doi: 10.1029/2020WR029262.

[18] IPCC, 2014. Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK, and New York, NY, USA doi: 10.1017/CBO9781107415416.

[19] Johansson, T.B., Patwardhan, A., Nakićenović, N., Gomez-Echeverri, L., 2012. Global Energy Assessment: Toward a Sustainable Future. Cambridge University Press, Cambridge.

[20] Jacobsson, S., Johnson, A., 2000. The diffusion of renewable energy technology: an analytical framework and key issues for research. Energy Policy 28 (9), 625-640. doi: 10.1016/S0301-4215(00)00041-0.

[21] Koster, R., Pîslar, M., Tacchetti, A., Balaguer, J., Liu, L.Q., Elie, R., Hauser, O.P., Tuyls, K., Botvinick, M., Summerfield, C., 2025. Deep reinforcement learning can promote sustainable human behaviour in a common-pool resource problem. Nat. Commun. 16 (1), 2824. doi: 10.1038/s41467-025-58043-7.

[22] Lamperti, F., Mandel, A., Napoletano, M., Sapio, A., Roventini, A., Balint, T., Khorenzhenko, I., 2019. Towards agent-based integrated assessment models: examples, challenges, and future developments. Reg. Environ. Change 19 (3), 747-762. doi: 10.1007/s10113-018-1287-9.

[23] Lee, S.Y., Lei, B.W., Mallick, B., 2020. Estimation of COVID-19 spread curves integrating global data and borrowing information. PLoS One 15 (7), e0236860. doi: 10.1371/journal.pone.0236860.

[24] Lelieveld, J., Klingmüller, K., Pozzer, A., Burnett, R.T., Haines, A., Ramanathan, V., 2019. Effects of fossil fuel and total anthropogenic emission removal on public health and climate. Proc. Natl. Acad. Sci. U.S.A. 116 (15), 7192-7197. doi: 10.1073/pnas.1819989116.

[25] Li, X., Cheng, G.D., Lin, H., Cai, X.M., Fang, M., Ge, Y.C., Hu, X.L., Chen, M., Li, W.Y., 2018. Watershed system model: the essentials to model complex human-nature system at the river basin scale. J. Geophys. Res. Atmos. 123 (6), 3019-3034. doi: 10.1002/2017JD028154.

[26] Li, X., Feng, M., Ran, Y.H., Su, Y., Liu, F., Huang, C.L., Shen, H.F., Xiao, Q., Su, J.B., Yuan, S.W., Guo, H.D., 2023. Big data in Earth system science and progress towards a digital twin. Nat. Rev. Earth Environ. 4 (5), 319-332. doi: 10.1038/s43017-023-00409-w.

[27] Li, X., Su, J.B., Yuan, S.W., 2026. Navigating Earth governance: a cybernetic and pragmatic pathway to digital twin Earth. Innov. Geosci. 4, 100203. doi: 10.59717/j.xinn-geo.2026.100203.

[28] Littman, M.L., 1994. Markov games as a framework for multi-agent reinforcement learning. In: Machine Learning Proceedings 1994. Elsevier, Amsterdam, pp. 157-163. doi: 10.1016/B978-1-55860-335-6.50027-1.

[29] Lowe, R., Wu, Y., Tamar, A., Harb, J., Abbeel, P., Mordatch, I., 2017. Multi-agent actor-critic for mixed cooperative-competitive environments. In: Proceedings of the 31st International Conference on Neural Information Processing Systems. ACM, pp. 6382-6393.

[30] Luderer, G., Pietzcker, R.C., Bertram, C., Kriegler, E., Meinshausen, M., Edenhofer, O., 2013. Economic mitigation challenges: how further delay closes the door for achieving climate targets. Environ. Res. Lett. 8 (3), 034033. doi: 10.1088/1748-9326/8/3/034033.

[31] Milkoreit, M ., 2023. Social tipping points everywhere?-Patterns and risks of overuse. Wires Clim. Change 14 (2), e813. doi: 10.1002/wcc.813.

[32] Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A.A., Veness, J., Bellemare, M.G., Graves, A., Riedmiller, M., Fidjeland, A.K., Ostrovski, G., Petersen, S., Beattie, C., Sadik, A., Antonoglou, I., King, H., Kumaran, D., Wierstra, D., Legg, S., Hassabis, D., 2015. Human-level control through deep reinforcement learning. Nature 518 (7540), 529-533. doi: 10.1038/nature14236.

[33] Moore, F.C., Lacasse, K., Mach, K.J., Shin, Y.A., Gross, L.J., Beckage, B., 2022. Determinants of emissions pathways in the coupled climate-social system. Nature 603 (7899), 103-111. doi: 10.1038/s41586-022-04423-8.

[34] Mullapudi, A., Lewis, M.J., Gruden, C.L., Kerkez, B., 2020. Deep reinforcement learning for the real time control of stormwater systems. Adv. Water Resour. 140, 103600. doi: 10.1016/j.advwatres.2020.103600.

[35] Müller-Hansen, F., Schlüter, M., Mäs, M., Donges, J.F., Kolb, J.J., Thonicke, K., Heitzig, J., 2017. Towards representing human behavior and decision making in Earth system models-an overview of techniques and approaches. Earth Syst. Dynam. 8 (4), 977-1007. doi: 10.5194/esd-8-977-2017.

[36] Nemet, G.F., 2009. Demand-pull, technology-push, and government-led incentives for non-incremental technical change. Res. Policy 38 (5), 700-709. doi: 10.1016/j.respol.2009.01.004.

[37] Nordhaus, W., 2018. Evolution of modeling of the economics of global warming: changes in the DICE model, 1992-2017. Clim. Change 148 (4), 623-640. doi: 10.1007/s10584-018-2218-y.

[38] Nordhaus, W.D., 2017. Revisiting the social cost of carbon. Proc. Natl. Acad. Sci. U.S.A. 114 (7), 1518-1523. doi: 10.1073/pnas.1609244114.

[39] Otto, I.M., Donges, J.F., Cremades, R., Bhowmik, A., Hewitt, R.J., Lucht, W., Rockström, J., Allerberger, F., McCaffrey, M., Doe, S.S.P., Lenferna, A., Morán, N., van Vuuren, D.P., Schellnhuber, H.J., 2020. Social tipping dynamics for stabilizing Earth’s climate by 2050. Proc. Natl. Acad. Sci. U.S.A. 117 (5), 2354-2365. doi: 10.1073/pnas.1900577117.

[40] Pan, X.D., Chen, D.L., Pan, B.X., Huang, X.Z., Yang, K., Piao, S.L., Zhou, T.J., Dai, Y.J., Chen, F.H., Li, X., 2025. Evolution and prospects of Earth system models: challenges and opportunities. Earth-Sci. Rev. 260, 104986. doi: 10.1016/j.earscirev.2024.104986.

[41] Ramanathan, V., Xu, Y.Y., 2010. The copenhagen accord for limiting global warming: criteria, constraints, and available avenues. Proc. Natl. Acad. Sci. U.S.A. 107 (18), 8055-8062. doi: 10.1073/pnas.1002293107.

[42] Ramanathan, V., Xu, Y.Y., Versaci, A., 2022. Modelling human-natural systems interactions with implications for twenty-first-century warming. Nat. Sustain. 5 (3), 263-271. doi: 10.1038/s41893-021-00826-z.

[43] REN21, 2019. Renewables 2019: Global Status Report. REN21 Secretariat, Paris. https://www.ren21.net/wp-content/uploads/2019/05/gsr_2019_full_report_en.pdf.

[44] Ripple, W.J., Wolf, C., Rockström, J., Richardson, K., Wunderling, N., Gregg, J.W., Westerhold, T., Schellnhuber, H.J., 2026. The risk of a hothouse Earth trajectory. One Earth 9 (2), 101565. doi: 10.1016/j.oneear.2025.101565.

[45] Rockström, J., Gupta, J., Qin, D.H., Lade, S.J., Abrams, J.F., Andersen, L.S., Armstrong McKay, D.I., Bai, X.M., Bala, G., Bunn, S.E., Ciobanu, D., DeClerck, F., Ebi, K., Gifford, L., Gordon, C., Hasan, S., Kanie, N., Lenton, T.M., Loriani, S., Liverman, D.M., Mohamed, A., Nakicenovic, N., Obura, D., Ospina, D., Prodani, K., Rammelt, C., Sakschewski, B., Scholtens, J., Stewart-Koster, B., Tharammal, T., van Vuuren, D., Verburg, P.H., Winkelmann, R., Zimm, C., Bennett, E.M., Bringezu, S., Broadgate, W., Green, P.A., Huang, L., Jacobson, L., Ndehedehe, C., Pedde, S., Rocha, J., Scheffer, M., Schulte-Uebbing, L., de Vries, W., Xiao, C.D., Xu, C., Xu, X.W., Zafra-Calvo, N., Zhang, X., 2023. Safe and just Earth system boundaries. Nature 619 (7968), 102-111. doi: 10.1038/s41586-023-06083-8.

[46] Rockström, J., Steffen, W., Noone, K., Persson, Å., Chapin III, F.S., Lambin, E.F., Lenton, T.M., Scheffer, M., Folke, C., Schellnhuber, H.J., Nykvist, B., de Wit, C.A., Hughes, T., van der Leeuw, S., Rodhe, H., Sörlin, S., Snyder, P.K., Costanza, R., Svedin, U., Falkenmark, M., Karlberg, L., Corell, R.W., Fabry, V.J., Hansen, J., Walker, B., Liverman, D., Richardson, K., Crutzen, P., Foley, J.A., 2009. A safe operating space for humanity. Nature 461 (7263), 472-475. doi: 10.1038/461472a.

[47] Sanderson, B.M., Xu, Y.Y., Tebaldi, C., Wehner, M., O’Neill, B., Jahn, A., Pendergrass, A.G., Lehner, F., Strand, W.G., Lin, L., Knutti, R., Lamarque, J.F., 2017. Community climate simulations to assess avoided impacts in 1.5 and 2°C futures. Earth Syst. Dynam. 8 (3), 827-847. doi: 10.5194/esd-8-827-2017.

[48] Schellnhuber, H.J., 1999. ‘Earth system’ analysis and the second Copernican revolution. Nature 402 (S6761), C19-C23. doi: 10.1038/35011515.

[49] Seo, H., Lee, D., 2017. Reinforcement learning and strategic reasoning during social decision-making. In: Decision Neuroscience. Elsevier, Amsterdam, pp. 225-231. doi: 10.1016/B978-0-12-805308-9.00018-X.

[50] Shmelev, S.E., Speck, S.U., 2018. Green fiscal reform in Sweden: econometric assessment of the carbon and energy taxation scheme. Renew. Sustain. Energy Rev. 90, 969-981. doi: 10.1016/j.rser.2018.03.032.

[51] Shuvo, S.S., Yilmaz, Y., Bush, A., Hafen, M., 2022. Modeling and simulating adaptation strategies against sea-level rise using multiagent deep reinforcement learning. IEEE Trans. Comput. Soc. Syst. 9 (4), 1185-1196. doi: 10.1109/TCSS.2021.3122282.

[52] Silver, D., Singh, S., Precup, D., Sutton, R.S., 2021. Reward is enough. Artif. Intell. 299, 103535. doi: 10.1016/j.artint.2021.103535.

[53] Smith, S.R., Christie, I., Willis, R., 2020. Social tipping intervention strategies for rapid decarbonization need to consider how change happens. Proc. Natl. Acad. Sci. U.S.A. 117 (20), 10629-10630. doi: 10.1073/pnas.2002331117.

[54] Steffen, W., Rockström, J., Richardson, K., Lenton, T.M., Folke, C., Liverman, D., Summerhayes, C.P., Barnosky, A.D., Cornell, S.E., Crucifix, M., Donges, J.F., Fetzer, I., Lade, S.J., Scheffer, M., Winkelmann, R., Schellnhuber, H.J., 2018. Trajectories of the Earth system in the anthropocene. Proc. Natl. Acad. Sci. U.S.A. 115 (33), 8252-8259. doi: 10.1073/pnas.1810141115.

[55] Strnad, F.M., Barfuss, W., Donges, J.F., Heitzig, J., 2019. Deep reinforcement learning in World-Earth system models to discover sustainable management strategies. Chaos 29 (12), 123122. doi: 10.1063/1.5124673.

[56] Sun, A.Y., 2020. Optimal carbon storage reservoir management through deep reinforcement learning. Appl. Energy 278, 115660. doi: 10.1016/j.apenergy.2020.115660.

[57] Sun, W.J., Bocchini, P., Davison, B.D., 2020. Applications of artificial intelligence for disaster management. Nat. Hazards 103 (3), 2631-2689. doi: 10.1007/s11069-020-04124-3.

[58] Sutton, R.S., 1988. Learning to predict by the methods of temporal differences. Mach. Learn. 3 (1), 9-44. doi: 10.1023/A:1022633531479.

[59] Tagomori, I.S., Diuana, F.A., Baptista, L.B., Bertram, C., Dafnomilis, I., Drouet, L., Fosse, F., Fragkiadakis, D., Fricko, O., Hooijschuur, E., Iyer, G., Kikstra, J.S., Krey, V., Luderer, G., Ou, Y., Reis, L.A., Richters, O., Rochedo, P.R.R., Vrontisi, Z., Weitzel, M., Zwerling, M., van Ruijven, B., Schaeffer, R., van Vuuren, D., 2026. Promising climate progress from net-zero ambitions to the Paris agreement goal. Nat. Clim. Change 16, 550-557. doi: 10.1038/s41558-026-02615-y.

[60] van Vuuren, D.P., Doelman, J.C., Schmidt Tagomori, I., Beusen, A.H.W., Cornell, S.E., Röckstrom, J., Schipper, A.M., Stehfest, E., Ambrosio, G., van den Berg, M., Bouwman, L., Daioglou, V., Harmsen, M., Lucas, P., van der Wijst, K.-I., van Zeist, W.-J., 2025. Exploring pathways for world development within planetary boundaries. Nature 641 (8064), 910-916. doi: 10.1038/s41586-025-08928-w.

[61] van Vuuren, D.P., Kok, M., Lucas, P.L., Prins, A.G., Alkemade, R., van den Berg, M., Bouwman, L., van der Esch, S., Jeuken, M., Kram, T., Stehfest, E., 2015. Pathways to achieve a set of ambitious global sustainability objectives by 2050: explorations using the IMAGE integrated assessment model. Technol. Forecast. Soc. Change 98, 303-323. doi: 10.1016/j.techfore.2015.03.005.

[62] Weyant, J., 2017. Some contributions of integrated assessment models of global climate change. Rev. Environ. Econ. Policy 11 (1), 115-137. doi: 10.1093/reep/rew018.

[63] Roser, M., Ritchie, H., Mathieu, E., 2023. Technological Change. Our World in Data. https://ourworldindata.org/technology-adoption.

[64] Wolf, T., Nardelli, N., Shawe-Taylor, J., Perez-Ortiz, M., 2023. Can reinforcement learning support policy makers? A preliminary study with integrated assessment models. arXiv. doi: 10.48550/arXiv.2312.06527.

[65] Wong, A., Bäck, T., Kononova, A.V., Plaat, A., 2023. Deep multiagent reinforcement learning: challenges and directions. Artif. Intell. Rev. 56 (6), 5023-5056. doi: 10.1007/s10462-022-10299-x.

[66] Ye, C., Gao, L., Zhuang, L., Li, Y.H., Fanzo, J., Meadows, M.E., Tang, X., Baklanov, A., Wu, Y.M., 2026. Climate and society as one: a (de)coupling future for escalating climate challenges. Sci. Bull. doi: 10.1016/j.scib.2026.03.026.

[67] Yu, C., Velu, A., Vinitsky, E., Gao, J., Wang, Y., Bayen, A., Wu, Y., 2022. The surprising effectiveness of PPO in multi-agent games. In: 36th Conference on Neural Information Processing Systems (NeurIPS 2022). Curran Associates, Inc., pp. 24611-24624.

[68] Yuan, S.W., Li, X., Du, E.H., 2021. Effects of farmers’ behavioral characteristics on crop choices and responses to water management policies. Agric. Water Manag. 247, 106693. doi: 10.1016/j.agwat.2020.106693.

[69] Yuan, S.W., Li, X., 2026. Generalizing human-natural systems modeling and decision-making: a multi-agent deep reinforcement learning framework and its application to the tragedy of the commons. Sci. China Earth Sci. 69 (1), 109-124. doi: 10.1007/s11430-025-1738-8.