The influence of storms on the CO2 flux in the Southern Ocean
Thesis / Dissertation
2026
Permanent link to this Item
Authors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
University of Cape Town
Publisher
University of Cape Town
Department
Faculty
License
Series
Abstract
The Southern Ocean has been one of the largest contributors to mitigating the rising greenhouse effect by acting as a sink and long-term storage for anthropogenic CO2 and excess heat. Despite its significance, consistent observations at the required temporal and spatial scales are still lacking due to its challenging environment. Hence, the mechanisms underlying the role of storms in driving the air-sea CO2 exchange (FCO2) are yet to be fully understood. An ongoing debate exists regarding the importance of observing the short-term (< 3 days) variability in the air-sea gradient in the partial pressure of CO2 (ΔpCO2) and its contribution to the FCO2. Satellite reanalysis-based studies have largely attributed the short-term FCO2 variability to the wind via changes in the gas transfer velocity (kw), with any contribution from the ΔpCO2 to be of secondary importance to the FCO2. On the other hand, studies that utilized high-resolution observations from autonomous gliders have revealed a potentially greater contribution of ΔpCO2 to the FCO2 during storms. The main hypothesis that shapes this thesis is thus that synoptic features, including storms and mesoscale induced variability, may play a more significant role in driving seasonal and annual FCO2 variability in the Southern Ocean than previously thought. Here, I examine this hypothesis by comparing the contributions of both kw and ΔpCO2 to the FCO2, through a series of storms experiments in the South Atlantic sector of the Southern Ocean using high-resolution glider observations and model simulations. From autonomous glider observations at hourly resolution, I show that ΔpCO2 may be as decisive to consider as kw when quantifying the impact of wind on FCO2 over storm synoptic cycles. The influence of the wind on the ΔpCO2 drove a feedback on the FCO2 uptake, slightly counteracting the enhanced wind-driven uptake (through the kw), resulting in lower FCO2 uptake than expected. This storm-pCO2 feedback on the FCO2 was established to be seasonal owing to the seasonal stratification and depth of the pycnocline. Storm-induced entrainment and mixing during winter and spring caused a weakening in the ΔpCO2 (by 22-37 μatm), reducing storm-driven FCO2 uptake by 6-27%. Contrary to what was previously understood, this negative feedback in the ΔpCO2 response occurred in phase with the rise and drop in wind stress. With stronger buoyancy forcing across summer, the sensitivity of ΔpCO2 to strong winds was much weaker, and changes in temperature and net primary productivity showed to be the more dominant forcing on ΔpCO2. Since observations only provided a one-point location assessment in the sub-Antarctic zone, a high resolution (± 3 km) forced mesoscale-resolving regional NEMO-PISCES model was used to explore the spatial sensitivity of the storm-pCO2 feedback over variable mesoscale gradients. This investigation highlighted that this feedback was highly spatially sensitive and strongest (up to 32%) in regions of strong horizontal density gradients. The model simulation also emphasized that the ΔpCO2 response to the storm was more synchronous with the wind in regions of pronounced density gradient, but a more delayed response (about two days) was noted in areas of weak gradients. Finally, because the storm-pCO2 feedback is enhanced in regions with strong mesoscale density gradients, two forced regional NEMO-PISCES models at different spatial resolutions (± 3 km and 40 km) were used to confirm whether fully mesoscale-resolving models are necessary to capture this feedback and to understand the potential impact on the seasonal to annual FCO2. The influence of capturing the intraseasonal variability in ΔpCO2 at higher resolution on the seasonal and annual mean FCO2 was examined, and the mechanistic responses of the FCO2 to storms in summer and winter were compared. Although the seasonal responses in ΔpCO2 to storms varied, the coarser resolution model (~ 40 km) consistently indicated that the variability in FCO2 was primarily attributed to kw, while in the mesoscale resolving model (~ 3 km), the spatial variation of FCO2 was more significantly influenced by the ΔpCO2 at one snapshot in time. Counterintuitively, the domain-averaged monthly mean FCO2 appeared to be more affected by the presence of mesoscale features during the summer and autumn months than during the winter-to-spring transitions, which enhanced the annual mean FCO2 by 24.61%. To conclude, this thesis highlighted that existing literature on the dynamics of FCO2 responses to storms primarily pertains to non-mesoscale resolving models, where high-frequency variability in kw alone dominates the intraseasonal FCO2 variability. However, strong mesoscale features, which can be captured using high-resolution observations and mesoscale resolving models, tend to reflect a more significant contribution of the ΔpCO2 to the FCO2, emphasizing further the need to observe synoptic scale variability at the correct temporal and spatial scale across the Southern Ocean.
Description
Reference:
Toolsee, T. 2026. The influence of storms on the CO2 flux in the Southern Ocean. . University of Cape Town. http://hdl.handle.net/11427/43593