Authors: Sung-Bin Park, Chang-Eui Park, Hanna Lee, Jin-Soo Kim
First published: 22 September 2026
Abstract
Stratospheric Aerosol Injection (SAI) is one of the alternative approaches to counteract anthropogenic warming by introducing reflective aerosols into the stratosphere. SAI-induced surface cooling and altered hydrological cycle have been widely investigated. However, the impact of SAI on terrestrial carbon retention, as represented by terrestrial ecosystem carbon turnover time (τ), remains unknown. Here, we quantify regional responses of τ to SAI, along with the relative importance of variables inducing τ variability, using simulations from four Earth System Models (ESMs) under the G6sulfur scenario. This scenario reduces the radiative forcing of Shared Socioeconomic Pathway 5-8.5 (SSP5-8.5) to SSP2-4.5 levels via stratospheric SO2 injection from 2020 to 2100. Results show that global τ generally increased under G6sulfur relative to SSP5-8.5 throughout the injection period, indicating an increase in ecosystem carbon retention. The largest regional increase in τ occurred in the northern high-latitudes, reaching up to about 6 years across all four ESMs during 2081–2100. In some tropical regions, such as Africa, the projected τ responses to SAI showed large inter-model spread, including opposing signs of change. These regional and model-dependent differences in τ responses mainly stem from variations in ecosystem carbon fluxes, specifically net primary production. The divergent regional impacts of SAI on τ highlight the need to account for regional differences in carbon retention capacity when evaluating the overall effectiveness of geoengineering strategies.
Plain Language Summary
Stratospheric Aerosol Injection is a proposed way to temporarily slow human-caused warming by placing reflective particles high in the atmosphere, reducing the sunlight that reaches Earth's surface. We know relatively well how this could change temperature and rainfall, but less about what it might do to how quickly carbon moves through land ecosystems. Using four Earth system models, we examined a case where this method is applied from 2020 to 2100 to offset warming under very high greenhouse gas emissions. We evaluated changes in terrestrial carbon turnover time, which summarizes carbon storage relative to carbon flows through vegetation and soils. Across the models, turnover time generally increases worldwide compared with the same high-emissions future without this intervention. The most consistent increase occurs in northern high latitudes, reaching up to about 6 years by the late century in all four models. In some tropical regions, including Africa, models disagree and even show opposite changes, mainly because they simulate different responses of plant growth. Our results show that this climate intervention alters land carbon cycling in strongly region-dependent ways; therefore, regional responses should be considered when judging its overall effectiveness.
Source: AGU