Investigating the impact of CO 2 on low-frequency variability of the AMOC in HadCM3

• Published in: Journal of Climate
• Main Authors: Armstrong, Edward, Valdes, Paul, House, Jo, Singarayer, Joy
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2017
• Subjects: Atmosphere-ocean interaction; Climate models; Climate variability; Ocean dynamics; Spectral analysis/models/distribution; Thermohaline circulation; Arctic; Greenland; Climate change; Iceland
• Online Access: https://hdl.handle.net/1983/471988cb-fe5e-407b-aa5b-11044972588d; https://research-information.bris.ac.uk/en/publications/471988cb-fe5e-407b-aa5b-11044972588d; https://doi.org/10.1175/JCLI-D-16-0767.1; https://research-information.bris.ac.uk/ws/files/143721677/Full_text_PDF_final_published_version_.pdf; http://www.scopus.com/inward/record.url?scp=85028958778&partnerID=8YFLogxK

This study investigates the impact of CO 2 on the amplitude, frequency, and mechanisms of Atlantic meridional overturning circulation (AMOC) variability in millennial simulations of the HadCM3 coupled climate model. Multichannel singular spectrum analysis (MSSA) and empirical orthogonal functions (EOFs) are applied to the AMOC at four quasi-equilibrium CO 2 forcings. The amount of variance explained by the first and second eigenmodes appears to be small (i.e., 11.19%); however, the results indicate that both AMOC strength and variability weaken at higher CO 2 concentrations. This accompanies an apparent shift from a predominant 100-125-yr cycle at 350 ppm to 160 yr at 1400 ppm. Changes in amplitude are shown to feed back onto the atmosphere. Variability may be linked to salinity-driven density changes in the Greenland-Iceland- Norwegian Seas, fueled by advection of anomalies predominantly from the Arctic and Caribbean regions. A positive density anomaly accompanies a decrease in stratification and an increase in convection and Ekman pumping, generating a strong phase of the AMOC(and vice versa). Arctic anomalies may be generated via an internal ocean mode that may be key in driving variability and are shown to weaken at higher CO 2 , possibly driving the overall reduction in amplitude. Tropical anomalies may play a secondary role in modulating variability and are thought to be more influential at higher CO 2 , possibly due to an increased residence time in the subtropical gyre and/or increased surface runoffdriven by simulated dieback of the Amazon rain forest. These results indicate that CO 2 may not only weaken AMOC strength but also alter the mechanisms that drive variability, both of which have implications for climate change on multicentury time scales.