The Paris Agreement aims to limit global mean temperature increase to well below 2°C, preferably 1.5°C. However, a significant emission reduction gap exists, prompting discussions around geoengineering as a potential intermediate solution. Geoengineering strategies are broadly categorized into Carbon Dioxide Removal (CDR) and Radiative Forcing Geoengineering (RFG). CDR directly addresses the root cause by removing CO2 from the atmosphere, while RFG alters the Earth's energy budget. RFG methods, such as Stratospheric Aerosol Injection (SAI), Marine Cloud Brightening (MCB), and Cirrus Cloud Thinning (CCT), offer the potential for rapid climate cooling, but their long-term impacts are uncertain. While several studies have investigated the efficacy of RFG in limiting global warming, fewer have comprehensively explored the potential side effects on high-latitude ecosystems, particularly in the Arctic, a region experiencing amplified climate change. This study addresses this knowledge gap by assessing the impacts of three commonly discussed RFG methods on Arctic temperatures, extreme events, and terrestrial system dynamics.

Numerous studies confirm RFG's potential in limiting future global warming, but also highlight undesirable side effects. These include altered precipitation patterns, monsoon periods, crop yields, ocean productivity, and dryland expansion. Previous research on the Arctic's response to RFG, primarily focusing on SAI, shows mixed results. Some studies indicate continued sea ice decline under SAI, while others suggest potential remediation with sufficient SO2 injection. The impact of SAI on permafrost warming is also debated, with some suggesting a slowing of warming but not to RCP4.5 levels. Existing research shows that equatorial SAI and MCB tend to cool the Arctic less than the global mean, but a thorough analysis of various RFG methods on high-latitude ecosystems, especially extreme temperatures, is lacking. The interplay between plant physiological forcing under high CO2 concentrations and RFG application is poorly understood, particularly in the Arctic, which experiences strong plant physiological forcing through heat transport from mid-latitudes.

This study employs the Norwegian Earth System Model (NorESM1-ME) to simulate the effects of three RFG methods (CCT, MCB, SAI). The model is configured to reduce radiative forcing from RCP8.5 to RCP4.5 levels by 2100. The three RFG methods are implemented in the model to achieve this reduction: (i) CCT by increasing the fall speed of ice crystals below -38°C; (ii) MCB by increasing sea salt aerosol emissions between 45°S and 45°N; (iii) SAI by prescribing the properties of an injected layer of sulfur dioxide in the stratosphere near the equator. The model simulates monthly boreal summer (JJA) maximum temperature (Txx), boreal winter (DJF) minimum temperature (Tnn), and mean temperature (Tmean) over land, both north of 50°N and 65°N. The study analyzes regional energy budgets, focusing on changes in transpiration, sensible heat, and latent heat. The impacts of RFG on Arctic terrestrial system dynamics, such as burned area and permafrost temperature conditions, are also evaluated using satellite data (MODIS burned area and ESA Climate Change Initiative Permafrost) and temperature data from the Climatic Research Unit TS4.01. Linear regression analysis compares temperature responses under RFG and RCP4.5 to assess the impact of plant physiological forcing and the effect of RFG on Arctic amplification. Three ensembles are run for each RFG scenario with slight perturbations to initial conditions, while one simulation each is carried out for RCP8.5 and RCP4.5.

While all three RFG methods successfully reduce global mean temperatures to near RCP4.5 levels by 2100, Arctic temperatures remain significantly higher. The Arctic is 0.7-1.1 °C warmer under all three RFG scenarios than under RCP4.5 by the end of the century, and more than 4 °C warmer compared to the 2006-2026 mean of RCP8.5. CCT and MCB show increased boreal summer maximum temperatures (Txx) north of 50°N compared to RCP4.5, while SAI exhibits an increase in boreal winter minimum temperatures (Tnn), particularly pronounced in the Arctic. All three methods show reduced transpiration compared to RCP4.5, indicating stomatal closure under high CO2 concentrations. The analysis of energy budget components reveals that per unit increase in global mean temperature, transpiration is significantly reduced in high latitudes under all three RFG methods, accompanied by an increased ratio of sensible to latent heat, supporting the hypothesis that plant physiological forcing contributes to Arctic temperature increase. CCT and MCB show decreases in albedo, potentially due to reduced snow cover, whereas SAI exhibits a decrease in incoming clear-sky shortwave radiation due to increased stratospheric aerosols. The study links these temperature changes to increased wildfire risk under CCT and MCB, as Txx anomalies become higher than Tmean anomalies in 2100, and lower soil moisture under MCB. The increased minimum temperatures under SAI lead to less favorable conditions for permafrost compared to RCP4.5, with a 7.8% decrease in permanently frozen area north of 65°N by 2100.

The findings demonstrate that the studied RFG methods, while effective in reducing global mean temperatures, do not sufficiently control Arctic temperatures and may exacerbate extreme events. The less efficient cooling of the Arctic under these RFG designs compared to the global mean is consistent with previous findings for SAI and MCB. The observed increase in Arctic extreme temperatures (Txx and Tnn) under different RFG methods points to the importance of considering regional impacts of geoengineering, even if global mean temperatures are successfully reduced. The role of plant physiological forcing in enhancing Arctic warming under RFG, highlighted by the reduced transpiration and shifted energy balance, is a critical aspect needing further investigation. Although the study cannot directly isolate the effect of physiological forcing, the shifts observed in the energy budget support the hypothesis that it plays a significant role. The observed changes in energy balance components under SAI and their differences compared to those under MCB and CCT highlight the method-specific impacts of RFG and the resulting differences in Arctic responses.

This study reveals that the three analyzed RFG methods, while potentially mitigating global warming, may increase the risk of Arctic wildfires and permafrost thaw compared to a mitigation scenario achieving the same global mean temperature reduction. The results underscore the need for comprehensive assessment of regional impacts of RFG and highlight the potential for unintended consequences in vulnerable ecosystems. Further research is essential to refine RFG strategies, explore alternative deployment methods, and improve the understanding of complex interactions within the Earth's climate system, particularly in the Arctic.

The use of a single Earth System Model (NorESM1-ME) limits the generalizability of the findings. Model-specific biases, and uncertainties in representing Arctic processes, could influence the results. The temporal resolution of the model output (monthly) limits the analysis of daily-scale extreme events, particularly for wildfires, and precludes a more detailed evaluation using fire weather indices. The study focuses on the response under a high-emission scenario (RCP8.5), and more research is needed to understand the impacts of RFG under mitigated emission scenarios.