Author: Taylor Cason
University of California, Los Angeles
Mentors: Professor Andrew Thompson, Mar Flexas, and Lily Dove
Resnick Sustainability Institute, California Institute of Technology
Editor: Bertha Mireles

Introduction

The ocean plays a vital role in the climate system because of its ability to store up to sixty times the carbon present in the atmosphere, exhibiting an exchange between the atmosphere and ocean that is of great interest as anthropogenic atmospheric CO₂ levels increase. The Southern Ocean is of particular importance due to the presence of the Antarctic Circumpolar Current (ACC), the world’s only global current flowing from west to east around Antarctica. The ACC is a major current that drives the majority of transport due to its strong flow rates and large range of density classes [1]. Wind forcing at the surface of the Southern Ocean brings dense and carbon-rich water to the surface, where its properties are susceptible to modification by surface heat fluxes, precipitation, ice melt, and air-sea gas exchange. This water is then injected back into the interior of the ocean along the northern edge of the ACC, making the Southern Ocean a key site for carbon sequestration. Thus, the Southern Ocean exerts great influence over the world’s carbon and nutrient cycling [2]. As ice sheets begin to melt at a rapid rate due to climate change, adjustments in sea density and level are expected to alter the Southern Ocean’s carbon transport [3]. 

Despite the Southern Ocean’s well-known control on the Earth’s atmosphere and CO₂ storage, quantifications of exchange rates between surface and interior waters have been minimal and difficult to obtain on a large scale. Regions known as mixing hotspots or areas having the strongest current interactions with underwater topography are vital to observe and describe [2]. One such region known as Drake Passage is located between South America and the Antarctic Peninsula. The land masses north and south of the passage make the latitudinal extent relatively small compared to the rest of the Southern Ocean. This narrow shape makes the transport of ACC and its features easier to constrain. Previous expeditions often acquired data on the “mesoscale” resolution or within 20-100 km. This project provides novel data from gliders on the “submesocale” or within 1-10 km. Measurements taken from this region were extrapolated and paired with altimetry data to visualize Drake Passage. These results are then able to be used for circumpolar extrapolation and climate modeling, making them highly influential on the interpretation of the Southern Ocean’s carbon budget and transport mechanisms [2]. 

Methods

In an effort to quantify and analyze the ocean’s carbon sequestration potential and its role in the global climate, autonomous platforms such as gliders were deployed to resolve small spatial scales of variability in biogeochemical and physical tracers that cannot be obtained by traditional observation methods. These observations provide 4D information (with respect to latitude, longitude, depth, and time) that can improve model simulations and forecasts of ocean conditions. Gliders utilize an inflatable, oil-filled bladder to change its buoyancy and to profile vertically between the surface and 1000-m depth. To obtain horizontal motion, gliders are equipped with wings that convert vertical lift into horizontal motion. Sensors are attached to the gliders to obtain data at varying depths and positions. Using the Iridium global telecommunications network, gliders receive user feedback for future actions and transmit data [3].  

This project analyzes 500 dives from Seaglider 539 (SG539) to identify anomalies in temperature, salinity, and dissolved oxygen. These anomalies can be paired with satellite altimetry data recorded during the expedition to explore the effects of fronts, distinct boundaries between water masses of different properties, on mixing and carbon transport.

Data and The Glider Track 
Seaglider 539 was deployed in Drake Passage from May 10 to August 20, 2016 covering the latitudes from 55.00°S to 62.30°S and longitudes from 62.30°W to 65.84°W. The expedition included 500 dives with a total distance traveled of 2832 km and a maximum dive depth of 1000 m. Three fronts of the ACC were crossed: the Subantarctic Front (SAF), Southern Antarctic Circumpolar Current Front (SACCF), and Polar Front (PF) as shown in Figure 1. The sensors on SG539 included pressure (dbar), temperature (℃), dissolved oxygen (µmol/L), salinity (g/kg), chlorophyll fluorescence, and backscatter. Using temperature and salinity measurements, potential density was calculated using the Glider Tools python package [4]. Chlorophyll fluorescence and backscatter were only recorded for the first 240 dives due to short glider battery life.

Glider Data Processing
Analysis of physical and biogeochemical properties were done through use of three datasets: raw data, vertically binned data, and horizontally interpolated data. The raw dataset was put through a quality control function using the Glider Tools python package and then used for initial observations and temperature-salinity and other property-property plots [4]. 

Vertical binning creates data that is evenly spread across the depth axis, allowing for gradient calculations to be straightforward. This was done by treating each cast profile as a single position in latitude and longitude, allowing properties to only change with depth for each profile. A bin size of 5 m was chosen to maintain accuracy in the resolution of the data. Thus, every 5 m in depth, properties within each 5 m section were averaged to a single value, and every profile was assigned an average position and surface position. The vertically binned dataset was first used to calculate the mixed layer depths and properties. This was done by calculating the potential density using temperature and salinity measurements. Mixed layer depth was then defined as the depth at which the difference in potential density from the surface value was greater than 0.03 kg/m³. All values prior to this depth value were averaged to find mixed layer properties for each profile. To assess the vertical stratification, the vertically binned dataset was used to calculate buoyancy frequency (N²), which is a measurement of the stability of the water to vertical displacements. The following equation was used to calculate buoyancy frequency, where g = 9.8 m/s² (acceleration due to gravity) and 𝜌₀ = 1025 kg/m³ (reference potential density),

Given that data was not distributed as evenly in the horizontal direction as the vertical, gridding properties over the horizontal direction could not simply be averaged. Instead, the Glider Tools package was used to interpret each property as a function of position and interpolate 500 evenly spaced points [4]. This dataset was used to create 3D section plots for front identification and horizontal stratification (M²), which measures the variance of density across latitude and longitude. Horizontal stratification was calculated using the following gradient equation:

Altimetry
Satellite data can be used to pair interior ocean measurements with surface data like sea level. Satellite data was obtained through the dates of the expedition. This dataset included absolute dynamic topography (ADT), sea level anomaly (SLA), zonal geostrophic velocity (ugos), and meridional geostrophic velocity (vgos). ADT measures the sea surface height above the geoid while SLA measures the sea surface height above the mean sea surface referenced in 1993-2012.  Thus, SLA outlines smaller scale features such as flow patterns less than 100km and timescales on the order of a month, while ADT captures larger patterns. Since each property in the altimetry data set is a function of latitude, longitude, and time, to create a 3D array for a section plot, all values were averaged in time across the entire span of the expedition. The altimetry data was used to quantitatively define the locations of the three fronts and compare to the 3D section plots used to qualitatively define the fronts.

Results and Discussion

Identification of Frontal Structures
Fronts on potential density figures are identified as regions of sharp horizontal gradients in potential density. Five main frontal crossings, indicative of multiple crossings of the PF and SAF can be seen in Figure 2 as well as other smaller fronts. 

Reduced vertical stratification as shown in Figure 3a occurs at the frontal crossings due to the injection of lighter water into depths. Additionally, mixed layer depth was found to increase primarily at the frontal crossings beyond 2000km. Buoyancy frequency was highest at the base of the mixed layer, while drastically dropping to nearly zero at the frontal crossings (Figure 3b) indicating the potential for increased mixing and injection of surface waters into the interior.

Ventilation
Dissolved oxygen is an indicator of the level of oxygen available for marine organisms, and its value with respect to the saturation solubility can be an indicator of primary productivity, the rate at which marine photosynthetic producers obtain energy and take up CO₂. Higher oxygen concentrations are found in the interior waters at the fronts suggestive of stronger ventilation or enhanced carbon exchange between surface and interior waters as seen in Figure 4. This correlates with the sharp boundary in density seen in Figure 1, which forms steeply sloped isopycnals, or surfaces with a constant density, that facilitate the subduction of surface waters and the upwelling of interior waters [6]. Furthermore, lower surface concentrations of dissolved oxygen are also found at these frontal crossings, likely due to the higher temperatures further north which results in a decrease in the solubility of oxygen into surface waters.       As seen in Figure 4, dissolved oxygen is less vertically stratified at the third frontal crossing indicating higher ventilation at this specific frontal crossing. Lower surface temperatures increase the density of water. Given that this crossing is later in the expedition and later in the winter season, lower surface temperature due to seasonality may have allowed for densification of the surface and interior waters.

Altimetry 
Ocean properties south of the Polar Front showed negative ADT values paired with colder mixed layers, while north of the Polar Front ADT values were mostly positive and paired with higher surface temperatures. Fronts show regions of large ADT change and the T-S diagram showed that north and south of the PF, density variations were primarily dominated by changes in salinity due to ice melt and coastal processes while the majority of temperature variation occurred at the front (ADT of approximately -0.61m) as shown in Figure 5.

Implications 
This Drake Passage dataset gives insight into the ocean’s ability to exchange and sequester atmospheric carbon and how this process is dominated in regions with strong fronts and pressure gradients. This was done through capturing the three main frontal features of the Southern Ocean and their surface and interior properties. Gathering these measurements provides implications for future extrapolation and estimation of carbon fluxes between the atmosphere and the Southern Ocean. Understanding the subduction of CO₂ and the hydrographic features that influence this process is critical to monitoring and modeling marine productivity in interior oceans as ventilation is a key component in photosynthesis and carbon cycling. As climate change continues to contribute to the warming of the ocean, ice melt begins to freshen the Southern Ocean, altering the density distribution and sea surface heights [3]. As a result, Southern Ocean fronts will likely shift position, potentially modifying the carbon transport and the distributions of nutrients and phytoplankton which are the base of the trophic web. This data in conjuction with expeditions in future years will begin to construct a temporal framework on fronts and carbon budgets. 

The glider data provides the unique ability to relate interior properties with depth and surface data gathered from altimetry, rendering insight into the spatial variability in surface-interior exchange. Unlike mooring and float data, we are able to measure properties at varying latitudes and longitudes with user feedback. Though Drake Passage is one the most frequently occupied regions used for Southern Ocean data collection due to its accessibility relative to the rest of the polar ocean, this analysis presents novel measurements at the 1-10 km scale, providing higher-resolution data than the typical ship based methods.

Conclusions

Sea Glider 539 was able to collect high resolution (1-10 km scale) salinity, temperature, dissolved oxygen, chlorophyll, and backscatter observations at varying positions and depths through Drake Passage. This data was paired with altimetry datasets taken during the expedition period. These findings from Drake Passage are useful for further extrapolating and estimating the Southern Ocean’s carbon sequestration potential as it constrains the ACC and the three main fronts. The effects of the SAF, SACCF and primarily the PF on carbon uptake were shown through increased interior dissolved oxygen content, increase mixed layer depths, and decreased vertical stratification. Increased mixed layer depths and decreased vertical stratification show that lighter waters with densities similar to the surface are injected and mixed into deep water. For this to occur, these regions must contain dynamics such as eddies and vortices that result in high mixing. It is expected that due to the change in sea surface height at fronts and the resulting horizontal density gradients that fronts exhibit, the generation of eddies and consequently vertical mixing is enhanced. This shows that fronts reveal high potential for carbon sequestration as they are able to store absorbed CO₂ from surface waters and transport or store these gases at greater depths than non-frontal regions. This was also evident with the higher dissolved oxygen content. With increased CO₂ at lower depths, photosynthetic activity is higher, resulting in higher dissolved O₂ and greater marine productivity. Ultimately, this work investigated the mixing characteristics in Drake Passage, demonstrating the importance of fronts in carbon subduction. It supplements and enhances previous research on ocean fronts with novel glider methodologies and resolution as well as gives insight into Southern Ocean dynamics, a region critically in need of more data to better understand ocean heat, carbon, and nutrient budgets.

Future Directions

Further studies on the effect of frontal structures on marine productivity can be done by analyzing chlorophyll and nutrient distribution. Though chlorophyll and backscatter sensors were included on SG539, due to battery issues, biogeochemical data was only collected for the first frontal crossing. It would also be important to explore the relationship of ventilation to seasonality and local wind stress, which is of interest to understand the sensitivity and persistence of frontal structures and their ability for carbon sequestration. Additionally, the effect of local wind stress should be compared to the meridian and zonal velocity components which are indicative of mesoscale dynamics (eddies and vortices on scales less than 100 km) and consider which dominates the vertical mixing in Drake Passage.

Acknowledgments

Thank you to Andrew Thompson, Lily Dove, and Mar Flexas for the support and mentorship during this research. We are grateful for the financial support from the Resnick Sustainability Institute and the support of WAVE/SURF organizers, Carol Casey and Candace Rypisi.

References

[1] Smith, R., Desflots, M., White, S., Mariano, A. J., & Ryan, E. H. (2005). The Antarctic CP Current. The Cooperative Institute for Marine and Atmospheric Studies website (http://oceancurrents.rsmas. miami. edu/southern/antarctic-cp. html). 

[2] Viglione, G. A., Thompson, A.F., & Sprintall, J. Controls on wintertime subduction in southern Drake Passage. Geophysical Research Letters  (2018).

[3] Chapman, C.C., Lea, MA., Meyer, A. et al. Defining Southern Ocean fronts and their influence on biological and physical processes in a changing climate. Nat. Clim. Chang. 10, 209–219 (2020). https://doi.org/10.1038/s41558-020-0705-4

[4] Gregor, L., Ryan-Keogh, T. J., Nicholson, S.-A., du Plessis, M., Giddy, I., & Swart, S. (2019). GliderTools: A Python Toolbox for Processing Underwater Glider Data. Frontiers in Marine Science, 6(December), 1–13. https://doi.org/10.3389/fmars.2019.00738

[5] Kim, Y. S., & Orsi, A. H. (2014). On the Variability of Antarctic Circumpolar Current Fronts Inferred from 1992–2011 Altimetry, Journal of Physical Oceanography, 44(12), 3054-3071. 

[6] Cason, T., Thompson, A., Dove, L., Flexas, M. Carbon Transport in Drake Passage: Observations from Ocean Gliders. Poster presented at: California Geophysical Fluid Dynamics Conference; Aug 19 2022; Pasadena, CA.