By Hassaan Sipra, Director of Global Engagement
Last month, I participated in the annual meeting of the Geoengineering Model Intercomparison Project (GeoMIP), hosted at Cornell University this year. This project, now in its 14th year, has been a pioneer in increasing our understanding of how solar geoengineering (SG) might impact climate change. Although climate models simulated SG techniques like stratospheric aerosol injection (SAI) and marine cloud brightening (MCB) prior to GeoMIP, this project offers standardized experiments and scenarios for climate modeling groups across the world to use for simulations and comparisons. This process allows for the various groups to better determine whether the results of their modeling exercises are due to the SG intervention placed in the model, or due to the model’s own intricacies.
This model intercomparison approach is used for conducting climate assessments by international and national bodies, including the Intergovernmental Panel on Climate Change (IPCC), the foremost scientific authority on the subject. Most notable of the various climate modeling projects is the Coupled Model Intercomparison Project (CMIP), currently getting ready for its seventh phase – CMIP7, which will be a crucial assessment tool for the IPCC’s Seventh Assessment Report, due in 2029. GeoMIP will continue to be a part of these assessments as well, as its outputs have helped inform the global climate modeling community on many aspects of the global climate system. They have improved simulations and projections, and reduced uncertainties across various scenarios that are relevant for advancing scientific knowledge and aiding decision making on climate action.
As the conversation around SG continues to gain momentum among scientific, policy and civil society sectors, the work of GeoMIP stands out as an important success story thus far. The project continues to develop new experiments, analyses and publications that sheds some light onto a critical question: if emissions are not mitigated and climate change worsens, what might be the climate system response to SG strategies?
A GeoMIP meeting is one part research conference, where climate scientists gather to share their work; one part discussion forum on how the SG field and this project is developing; and one part planning committee, where the community deliberates on what to do next. Within this format, the GeoMIP community has evolved over the past decade and a half. When I went to my first GeoMIP meeting back in 2019, I was new to SG. Compared to last month’s gathering, it was smaller, the discussions felt highly technical, and representation from the Global South was limited. But this recent meeting showed me that as the community has grown, so have the kinds of discussion taking place among its members.
New Proposals Relevant to Global Discussions
While several new modeling experiments were discussed during the meeting, two stood out to me as highly relevant to the global discussions on SG. The first proposed SAI experiment would focus on the Arctic, with a high latitude injection, which may provide insights into SG’s potential around slowing sea level rise. At present, the potential effectiveness of such ‘localized’ interventions are ill understood, and are unlikely to be effective under high emissions scenarios. While it merits exploring impacts at regional levels, the discussions taking place at GeoMIP made clear that without deep decarbonization, such an SAI strategy would likely not be favorable.
Another proposed modeling experiment focuses on large scale MCB (G6-1.5K-MCB), and seeks to serve as a complement to a previously proposed GeoMIP SAI experiment that simulated keeping global temperatures stable at 1.5oC warming. MCB has been challenging for GeoMIP modeling teams since there is a large diversity in cloud cover and locations among models, making multimodel comparisons difficult. Cloud physics modeling is hard, and there is a lot more heterogeneity in how MCB is implemented as compared to SAI, making it a tough challenge to address. As such, GeoMIP discussed benefiting from focusing on more idealized MCB experiments, which are simpler and help in characterizing climate responses, ahead of more policy relevant experiments. As there have been relatively fewer studies on MCB thus far, the debate on how to design experiments that tackle this intervention are ongoing.
Deliberating Core Questions
In previous GeoMIP meetings, and throughout the climate modeling and assessment community, data has always been a struggle. Beyond the need for stable internet connections to download these large datasets, maintaining the servers and data nodes at the various modeling facilities is an arduous and expensive task. The impact on Global South researchers is fairly high as a result. Resolving the resource constraints facing both the modeling and the modeling output assessment groups is urgently needed to deliver better understanding of how these technologies could potentially be integrated into the global climate response portfolio.
One solution that has been under debate for some time is to develop emulators – a piece of software that ‘imitates’ an earth system model (ESM), or some component of it, using statistical modeling techniques, rather than physics-based ones. They represent the climate system at a lower level of complexity, and require much less computing power. These could be useful in multiple ways: scenario generation, quick exploration of novel scenarios, understanding small-scale processes, tuning parameters, etc. However, a great deal of care and thought needs to go into their design, and prioritizing their use. Some core questions that the meeting deliberated on were:
- What ESM data is needed to train emulators and how can GeoMIP help?
- How should various emulator groups coordinate?
- What should be the priority areas for emulators development and use? Is there a taxonomy of needs?
- Who will benefit from this knowledge development and how can GeoMIP best facilitate their use while reducing misinterpretation?
- Is there a possibility of misuse?
The answers to these questions about emulators are crucial not just for improving climate models but also for ensuring that the tools and scenarios developed within GeoMIP are robust, reliable, and serve the broader goals of climate science and climate policy.
The Evolving Role of GeoMIP
While the first questions in the previous section are somewhat technical in nature, the last two have a sentiment in them that echoed beyond the emulator discussion at this meeting. Two particular statements* made me realize the important role GeoMIP plays in the debate around SG.
- “The scenarios we pick have a lot of power, because they set the stage and the analysis boundaries for future discussions.”
- “We need to do a better job of communicating the good and the bad.”
The reason I bring these statements up is that it has always been necessary for GeoMIP (and the scientific community at large) to study and inform. But being cognizant of the responsibility that comes with making such decisions is key to ensuring that GeoMIP is seen as a fair, balanced, trustworthy resource beyond just the climate community – for policy makers, civil society and the general public.
Keeping such global stakeholders in mind tends to lead to another fascinating debate that is continuing at GeoMIP – the development of policy relevant scenarios. There is a push-and-pull within the GeoMIP community between 1) working towards ensuring that models and experiments provide the best scientific information (intricacies notwithstanding) versus 2) being more policy relevant. The problem with developing policy relevant scenarios is that they may not be future proof, and might lose their relevance fast, meaning resources are not efficiently utilized. Of course, the right answer is that the two types of experimental designs should be run in parallel, but finding that balance and coordinating between modeling groups is a large task. Hence, why these meetings are important to keep the discussion moving forward.
In line with the discussions on the types of scenarios to develop and the role they might play in informing scientists and decision makers on how seriously to consider these technologies, there is growing interest in studying the impacts of GeoMIP experiments on factors beyond climate variables (e.g. temperature and precipitation), such as human health, biodiversity, geopolitics, food security, etc. Although scientific studies have looked at impact assessments based on GeoMIP outputs, the analysis is currently insufficient to provide meaningful insights for many communities around the world. In particular, optimizing (downscaling) these global level models to regional and national levels is rarely done within the context of GeoMIP. There is a need to invest in more resources to tackle this challenge, so that more diverse, nuanced conversations can take place about how such climate interventions may be beneficial or harmful to diverse sets of communities around the world. GeoMIP is working to resolve these aspects by thinking through experiments that are designed for impact assessment.
As the field considering SG is growing to include more than the scientific community, with new players, new financing, and new motivations, what the GeoMIP community wants to do will also evolve. Many of the challenges facing GeoMIP are part-and-parcel of doing research within the climate science domain; it highlights the need to mainstream SG science into the broader climate research and conversation, for the benefit of both.
*Note: I am not naming individuals, only illustrating discussion points that resonated with me.