Parallel Session 3.4: CE assessment metrics – Comparative, Integrative, Comprehensive

To enable fair, comprehensive and comparative decision-making on Climate Engineering, we need to foster a multidisciplinary and integrative selection process for assessment metrics. In this session we want to learn to what extent established climate-change assessment metrics are applicable for Climate Engineering assessment and what kind of extensions are needed.

 

This session aims to foster discussions about approaches to comparatively assess different climate engineering (CE) ideas, both among each other and in the context of mitigation. We encourage contributions that address the following questions:

The session will start with an introductory talk by Elnaz Roshan and Nadine Mengis and continue with talks by Peter Irvine, Yann Chavaillaz, Mohammad Khabbazan and Nils Matzner. The session is convened by Nadine Mengis, Elnaz Roshan, Sebastian Sonntag, Andreas Oschlies, Wilfried Rickels and Hermann Held.
Elnaz Roshan, Sebastian Sonntag and Nadine Mengis will also be giving a poster presentation during the poster session on Wednesday.

Note: Hermann Held and Wilfried Rickels, while playing a lead role in co-convening this session, are unfortunately not able to attend in person.

Talks:

• Peter Irvine
  • Solar geoengineering could offset some of the effects of climate change on key drivers of climate risks but it would not do so perfectly and could exacerbate some of the effects of climate change in some regions. Using the GFDL’s high-resolution tropical-cyclone permitting model HiFLOR and for the Geoengineering Model Intercomparison Project ensemble we quantify whether solar geoengineering reduces or increases the magnitude of climate trends at the gridcell level. We focus on 5 key variables: temperature, maximum annual daily max temperature, precipitation, precipitation minus evaporation, and 5-day maximum precipitation. We find that if solar geoengineering is deployed to offset all global mean temperature change from an increase in CO2 concentrations (100%Geo) the majority of regions see a reduction in hydrological changes but a significant fraction of the land area see greater hydrological change. However, we show that for 50%Geo many fewer areas are made worse off for these hydrological variables. Furthermore, we find limited consistency between those regions made worse off in terms of precipitation and precipitation minus evaporation. These results suggest that most regions would benefit from a moderate deployment of solar geoengineering and highlights the importance of a detailed hydrological impacts assessment of solar geoengineering.
     
• Yann Chavaillaz
  • Stratospheric aerosol injection (SAI) is discussed as an option for counteracting anthropogenic global warming. But most rather consider it as a portfolio measure holding the potential to avoid an overshoot of mean temperatures, and minimizing our vulnerability to climate change. But climate extremes might pose a higher threat on our societies, ecosystems and infrastructures. Therefore, we investigate if the implementation of SAI can efficiently mitigate extremes in daily temperature, precipitation and heat stress or if it only forces extremes to shift to new spatial or temporal regimes. We used simulations from the MPI-ESM-LR model based on the socio-economic assumptions of the RCP8.5 scenario, with an SAI scenario until 2100 targeting the radiative forcing of RCP4.5 values. These simulations are compared to the three existing members of both RCP4.5 and RCP8.5 scenarios. We show that, despite a significant decrease of the occurrence of current daily extremes due to SAI, the total number of extremes remains higher over the century than with RCP4.5 mitigation measures. The probability of their occurrence does also not evolve at a similar pace throughout the century. Overall, our results challenge the concept of implementing SAI as a ‘shaving-the-cap’ measure to avoid dangerous climate change.
     
• Mohammad Khabbazan
  • While solar radiation management (SRM) offers an option to ameliorate anthropogenic temperature rise, it is not simultaneously expected to perfectly compensate for anthropogenic changes in further climate variables. Here, considering different regional weights in precipitation disparities, we ask to what extent a proponent of the 2°C-temperature target would apply SRM in conjunction with mitigation. We utilize cost-risk analysis in ‘Giorgi’-regional-scale to evaluate the optimal mixture of SRM and mitigation under probabilistic information about climate sensitivity for global temperature-risk-only, regional precipitation-risk-only, and equally-weighted both-risks scenarios. We find that although SRM can almost perfectly substitute for mitigation in temperature-risk-only scenarios, it matters how Giorgi regions are weighed in regional precipitation-risk-only and both-risks scenarios. Giving the whole weight to one region, only six critical regions get welfare-loss under joint-mitigation-SRM portfolio with perfect substitution of SRM for mitigation in all cases except Amazonia and Central America. Considering all specific trade-off parameters from previous evaluation divided by the number of regions, SRM saves about 1/3 of welfare loss under joint-mitigation-SRM portfolio. Under equal weights, SRM can save 1/3 to 2/3 of welfare-loss. However, considering only critical regions’ precipitation risks, SRM can save only about 1/10 to 1/5 of welfare loss in precipitation-risk-only and both-risks scenarios.
     
• Nils Matzner
  • Policy advisors and research institutions increasingly emphasize the “need to understand the possibilities, limitations, and potential side effects” of so-called climate interventions. Understanding the impacts of climate engineering (CE) depends to a large extent on choosing the right indicators and metrics for assessing CE simulations. Although it is clear that they often differ from conventional climate change research metrics due to changes in prevailing correlations (e.g. the breakdown of a correlation between global mean temperature and other indicators with a solar CE deployment), only a few studies have attempted to identify the metrics required for a fair and comprehensive CE assessment. This study utilizes the available CE literature and aims to identify and map currently used indicators and metrics to highlight research gaps. Applying quantitative text analysis to a comprehensive corpus of scientific CE literature allows us to efficiently map scientific applications of metrics. Furthermore, we aim to track how the metrics have changed over time, and if there are predominant metrics that appear in certain types of studies. By pointing to research gaps our results will foster research on insufficiently studied metrics and help to select the most appropriate ones, which have both scientific and societal relevance.

• Nadine Mengis
  • Climate engineering (CE) alters prevailing correlations between Earth system variables, hence an appropriate assessment of CE must include a reevaluation of the chosen assessment indicators. A fair comparison of CE methods presents an additional challenge, since they aim at manipulating different components of the Earth system.
    This study systematically identifies changes in correlation patterns introduced by three idealized Climate Engineering (CE) scenarios: Large-scale Afforestation (LAF), Ocean Alkalinity Enhancement (OAE) and Solar Radiation Management (SRM). Firstly, we investigate changes in prevailing correlations between Earths system variables of the single CE scenarios compared to two future emission scenarios, and the implications of such changes on chosen assessment indicators. Secondly, we evaluate a common correlation matrix and identify a set of 14 indicators for a comprehensive comparison of the three CE scenarios. The evaluation of the CE scenarios relative to a defined reference climate state shows, that each CE method can be found to show a good performance, depending on the given indicator. It is beyond the scope of this study to give a value judgement on which of these variables is of higher importance for society, but here we aim to provide the natural science knowledge to enable such a discussion.
     
• Sebastian Sonntag
  • We assess atmosphere-, ocean-, and land-based climate engineering (CE) measures with respect to their effects and unintended consequences consistently within one comprehensive model. We use the Max Planck Institute Earth System Model (MPI-ESM) with prognostic carbon cycle to compare solar radiation management (SRM) by stratospheric sulfur injection with carbon dioxide removal methods: afforestation and ocean alkalinization. We show that the CE methods differ vastly in terms of their effects on different Earth system components. We find that mitigating feedbacks emerge: for example, as a response to SRM temperatures are reduced leading to a reduction of atmospheric CO2 due to enhanced land carbon uptake. We also identify challenges arising in a comparative assessment of CE methods: the quantitative results depend on details of the CE scenarios and on the underlying models, and an interpretation of relative efficiency depends on the choice of variables that are analyzed. Furthermore, we show that normalisations allow for a better comparability of different CE methods. For example, we find that despite different amounts of global surface cooling achieved, local amplification factors compared to the global mean temperature changes are generally similar in the CE scenarios, with the exception of Arctic amplification, which is strengthened in SRM.
     
• Elnaz Roshan
  • Following the Paris Agreement, the research question has been raised about the role of climate engineering in compliance with 1.5°C-temperature target. Comparing target-based decision frameworks for 1.5°C- and 2°C- temperature targets in both global- and ‘Giorgi’-regional-scale analyses, here we ask for the optimal mix of SGE and mitigation while considering global and regional temperature and precipitation anomalies. Using a deterministic cost-effectiveness analysis (CEA) for climate sensitivity (CS) of 3°C, our simulations find no feasible solution to comply with 1.5°C-temperature target by only mitigation. By altering the target from 2°C to 1.5°C under CEA, SGE usage either does not change in the regional analysis or decreases in the global analysis. Applying cost-risk analysis (CRA) with probabilistic information on CS ranging between 1.01°C and 7.17°C, the cooling effect of SGE increases by 0.4°C for median CS of 3°C, when the temperature target is reduced. SGE can save approximately 2/3 to 3/4 of the welfare-loss in the mitigation-only portfolio in the regional setting of 1.5°C-temperature target analysis, respectively in precipitation-risk-only and both-risks scenarios. Moreover, regardless of the temperature target and decision framework, SGE usage is restricted by regionalizing the model.