Integrating global and regional climate modeling for robust future projections
Over the past few decades, the international modeling community has made significant strides in advancing our understanding of the Earth system and its response to human-induced climate change. Through collaborative efforts, such as the Coupled Model Intercomparison Project (CMIP) and the Coordinated Regional Climate Downscaling Experiment (CORDEX), scientists have developed increasingly sophisticated global and regional climate models that can simulate the complex interactions between the atmosphere, ocean, land, and biosphere.
As we approach a new round of these modeling activities, including the start of CMIP7, it is an opportune time to reflect on the progress made and identify the key priorities that can further enhance our ability to provide robust and actionable climate information to support international climate policy and adaptation planning.
Advancing Earth system modeling for policy-relevant insights
One of the overarching goals for the coming decade is to deliver a coordinated and internally consistent set of scenarios, projections, and impact assessments to support the Intergovernmental Panel on Climate Change (IPCC) assessments and the United Nations Framework Convention on Climate Change (UNFCCC) Global Stocktake. This will require a tighter integration between the various modeling communities, including integrated assessment models (IAMs), global and regional Earth system models (ESMs), and impact models.
Improved representation of the coupled Earth system
A key priority is to enhance the representation of the coupled Earth system in climate models, particularly with respect to the carbon cycle and its interaction with the physical climate. Many ESMs now include a full representation of the carbon cycle, allowing for more realistic simulations of future climate-carbon-cycle feedbacks. This progress has motivated calls for CMIP7 to focus more strongly on CO2-emission-driven simulations, where these feedbacks can be fully captured.
Beyond the carbon cycle, models also need to improve their representation of other critical Earth system processes, such as nutrient limitation, interactive methane cycles, nitrogen and iron cycles, permafrost dynamics, interactive fires, and fully coupled atmosphere-ocean-ice sheet interactions. Incorporating these processes will enable more thorough assessments of plausible emission pathways and global warming pathways that realize the Paris Agreement targets, including the potential for temporary overshoots and the associated risks and consequences.
Exploring the feasibility and impacts of negative emissions
A key focus will be on investigating the feasibility and impacts of negative CO2 emissions, a crucial component of many mitigation pathways aiming to limit global warming to 1.5°C or 2°C above pre-industrial levels. This includes assessing the efficacy of proposed land-based and marine-based carbon dioxide removal (CDR) techniques, such as bioenergy with carbon capture and storage (BECCS) and ocean alkalinization, in reducing atmospheric CO2 and driving global cooling.
Equally important is the need to evaluate the potential impacts of these large-scale CDR interventions on the natural environment, water resources, and other human activities. The full Earth system response to such interventions remains highly uncertain and requires careful assessment using the latest generation of ESMs.
Addressing the risks of overshooting climate targets
Another critical area is the investigation of plausible pathways that overshoot the 1.5°C or 2°C warming targets, before eventually returning to these levels. This includes assessing the rate and magnitude of the overshoot, the efficacy of mitigation measures (such as negative emissions) in reducing atmospheric CO2 and driving global cooling, and the potential risks and consequences of exceeding critical tipping points in the Earth system during the overshoot phase.
Enhancing regional climate information for adaptation planning
While global-scale modeling is crucial for understanding the overall Earth system response, regional-scale information is essential for informing climate change adaptation and resilience planning at national and local levels. The integration of global and regional climate modeling approaches is, therefore, a key priority for the coming decade.
Improving regional climate model fidelity
One important focus is to enhance the fidelity of regional climate models (RCMs) in simulating key regional processes and phenomena. This includes improving the representation of regional-scale features, such as orography, land-sea contrasts, and land-use patterns, as well as better capturing the influence of large-scale modes of variability (e.g., the El Niño-Southern Oscillation) on regional climate.
Increasing the resolution of RCMs, including the use of convection-permitting models, can significantly improve the simulation of regional climate, particularly for extremes such as heavy precipitation events. Closer collaboration between the global and regional modeling communities, as well as the use of advanced statistical and machine learning techniques for downscaling, can further enhance the reliability of regional climate projections.
Addressing the role of natural variability
Another key priority is to better understand and quantify the role of natural climate variability, particularly at regional scales and on timescales relevant for adaptation planning (e.g., the next 10-40 years). Large initial condition ensembles, combining global and regional models, can provide critical insights into the relative contributions of forced climate change and internal variability to regional climate impacts.
This information is crucial for adaptation decision-making, as natural variability can often dominate the forced climate change signal, especially at local scales and in the near-term. Improving the accessibility and usability of this information for end-users, such as policymakers and climate service providers, is a key challenge that needs to be addressed.
Advancing model development and evaluation
Alongside the priorities for enhancing Earth system representation and regional climate information, continued efforts are needed to improve the overall performance and evaluation of climate models. This includes:
Improved simulation of the historical climate
Ensuring climate models can accurately reproduce the observed historical evolution of the climate system, including global and regional warming trends, is a fundamental requirement for increasing confidence in future projections. Advances in model parameterizations, the representation of key processes (e.g., cloud feedbacks), and the use of observational constraints can help address persistent biases in models.
Enhanced model resolution and process realism
Increasing the resolution of global and regional climate models, as well as improving the representation of key processes and their interactions, can lead to more realistic simulations of climate variability and extremes. Closer collaboration between the global model development community, the regional modeling community, and experts in machine learning and hybrid modeling approaches can help optimize these advancements.
Improved uncertainty quantification
Developing and applying a hierarchy of models and methods, including large ensembles, perturbed parameter ensembles, and statistical emulators, can help more comprehensively explore the range of uncertainties in future Earth system projections. This is particularly important for assessing the risks of high-impact, low-likelihood outcomes, such as the exceedance of critical tipping points in the Earth system.
Enabling effective science-policy interfaces
To realize the ambitious goals outlined in this article, a robust and interconnected computational and data infrastructure ecosystem is essential. This includes:
Efficient data management and sharing
Developing an integrated data infrastructure that supports the efficient co-production, co-exploitation, and rapid sharing of model simulations, data, and analysis tools among the various modeling communities and end-users. This will help ensure internal consistency and enable the timely delivery of scientific knowledge to support policy decisions.
Collaborative model development and evaluation
Fostering closer collaboration between modeling groups, observational data providers, and end-users to co-design experiments, co-produce datasets, and co-develop evaluation frameworks. This will help ensure the scientific outputs are tailored to the needs of policymakers, adaptation planners, and climate service providers.
Capacity building and global participation
Strengthening the involvement of the Global South in the development and application of climate models, as well as the use of climate information for decision-making. This will help ensure that the science is truly global in its scope and relevance, and that the benefits of improved Earth system understanding are shared equitably across the world.
By addressing these priorities, the international modeling community can deliver the enhanced scientific support needed to tackle the urgent challenges posed by climate change and inform effective mitigation and adaptation strategies at global, regional, and local scales.
