Comparison of UKESM1 and CESM2 Simulations Using the Same Multi-Target Stratospheric Aerosol Injection Strategy

Abstract

Abstract. Solar climate intervention using stratospheric aerosol injection (SAI) has been proposed as a method which could offset some of the adverse effects of global warming. The Assessing Responses and Impacts of Solar climate intervention on the Earth system with Stratospheric Aerosol Injection (ARISE-SAI) set of simulations is based on a moderate greenhouse gas emission scenario and employs injection of sulphur dioxide at four off-equatorial locations using a control algorithm which maintains the global-mean surface temperature at 1.5 K above preindustrial conditions (ARISE-SAI-1.5), as well as the latitudinal gradient and inter-hemispheric difference in surface temperature. This is the first comparison between two models (CESM2 and UKESM1) applying the same multi-target SAI strategy. CESM2 is successful in reaching its temperature targets, but UKESM1 has considerable residual Arctic warming. This occurs because the pattern of temperature change in a geoengineered climate is determined both by the structure of the climate forcing (mainly greenhouse gases and stratospheric aerosols) and the climate models’ feedbacks, the latter of which favour a strong Arctic amplification of warming in UKESM1. Therefore, research constraining the level of future Arctic warming would also inform any hypothetical SAI deployment strategy which aims to maintain the interhemispheric and equator-to-pole near-surface temperature differences. Furthermore, despite broad agreement in the precipitation response in the extratropics, precipitation changes over tropical land show important inter-model differences, even under greenhouse gas forcing only. In general, this ensemble comparison is the first step in comparing policy-relevant scenarios of SAI, and will help in the design of an experimental protocol which both reduces some known negative side effects of SAI and is simple enough to encourage more climate models to participate.

Climate Intervention using marine cloud brightening (MCB) compared with stratospheric aerosol injection (SAI) in the UKESM1 climate model

“…The latitudinal distribution of aerosol optical depth in G6sulfur peaks in tropical regions which is due to the specified injection strategy of injecting between 10° N and 10° S. Significant work has been done examining the utility of alternative strategies using latitudinally variable injections (e.g., Kravitz et al, 2017;Bednarz et al, 2023;Visioni et al, 2023a;Henry et al, 2023) that reduce the tropical AOD peak and the associated over-cooling of tropical regions with continued warming at high latitudes (Fig. 7c).…”

“…While G6sulfur shows the expected maximum zonal mean residual warming for 2081-2100 between 60-90 °N which has been evident in GeoMIP simulations which inject aerosol at Equatorial latitudes (e.g.,Kravitz et al, 2013aKravitz et al, , 2015, Fig.7(c)shows that G6MCB has another maximum at 30-60 °N. Just as for SAI, where considerable research has been performed into strategies to ameliorate residual temperature impacts by injecting at latitudes outside of the tropics (e.g.,Kravitz et al, 2017;Henry et al, 2023), these details are likely to be a function of the deployment strategy. Increasing relative emissions in the NP region which spans 30-50 °N might rectify this issue.…”

Assessing the Impact of Stratospheric Aerosol Injection on US Convective Weather Environments

“…Future work could examine how model-specific biases impact future projections of convective weather environments with and without SAI. For instance, simulations similar to the ARISE-SAI scenarios examined here were recently completed using the first version of the UK Earth System Model (Archibald et al, 2020;Henry et al, 2023;Sellar et al, 2019). It would also be insightful to examine Earth-system model simulations with different SAI deployment goals and timelines (e.g., MacMartin et al, 2022), as well as simulations under different climate change scenarios, such as the Stratospheric Aerosol Geoengineering Large Ensemble Project that utilized the high-end RCP8.5 emissions scenario (Tilmes et al, 2018).…”

G6-1.5K-SAI: a new Geoengineering Model Intercomparison Project (GeoMIP) experiment integrating recent advances in solar radiation modification studies

“…Since 2015, many studies of SAI have shown that strategies that move away from equatorial injections, as was used for G6sulfur, might be preferable. Recently, Henry et al (2023) have compared two models using a controller (Kravitz et al, 2016) to manage four injection locations (30ºN, 15ºN, 15ºS, 30ºS). However, this would be hard to achieve for models which have not implemented a feedback controller.…”

“…(2022), so that, given the same starting date, all models can start "ramping up" with the SRM amount independently of how fast they were in the historical period at warming. As noted by Henry et al (2023), the choice of both 2035 and of defining 1.5ºC compared to the model PI period may mean relatively rapid deployment of SAI in models that have already exceeded the 1.5ºC target. If the start date were also changed in each model dependent on when that model reached 1.5 ºC, that may result in implausible start-dates, as well as making intermodel comparisons more difficult.…”