1. Lauren Santi and Alison Gray
Understanding oxygen in the Southern Ocean from models and observations
Lauren Santi and Alison Gray

2. Motivation High variability of oxygen levels
seasonal, interannual, regional
Limited knowledge of the Southern Ocean
historical observations are sparse, biased towards summer
Importance of Southern Ocean in global oxygen cycle
Annual mean zonally averaged oceanic uptake of oxygen for period (Matear et al. 2000)
There are a variety of reasons to look at the seasonal cycle in greater depth:
The Southern ocean experiences substantial variability, whether it be seasonal, regional, or interannual (from year to year).
S.O. important for oxygen uptake and also changes the most according to this model
green = change between climate forcing(red) and control(black) simulations
climate change
control
difference

3. Research Questions
How well do current models represent the variability of oxygen in the Southern Ocean?
How will oxygen concentrations in the Southern Ocean change with increased CO2?
Approach: Use observations and coupled climate models

4. ARGO Profiling Floats
Floats take profile measurements of top 2000 meters every ten days
A subset have oxygen sensors
17,521 profiles
Oxygen data has recently been recalibrated using World Ocean Atlas
floats go under the ice which is how we get this winter data

5. GFDL climate model simulations
CM2-1deg and CM2.6:
Pre-industrial and climate change CO2 simulations
Coupled to miniBLING biogeochemical model
Subsampling of model data
Allows for more representative comparison between the models and observations
Biology Light Iron Nutrient and Gas Model (miniBLING)
Biogeochemical model with oxygen, carbon and phosphate tracers
CM2-1deg
coupled climate model

6. Zonally and time-averaged depth profiles
both models underpredict the amount of dissolved oxygen, but CM2-1deg is generally worse
increased CO2 simulation showed less o2 in the water column
oxygen minimum zone tends to get shallower moving south
CM2.6 is better at modeling oxygen profiles in the pacific ocean but shows no clear improvement in the Indian and Atlantic Oceans.
region of maximum oxygen change with increased CO2 gets shallower moving south

7. Zonally averaged seasonal cycle comparison (depth = 50 meters)
lots of variability, not a strong seasonal cycle when looking at the subsampled data
data is incomplete- the Pacific Ocean at 45deg only has data from 1 month!
can’t say what the seasonal cycle is given the results of the subsampled data
contrast with strong seasonal cycle seen before the data was subsampled
for the most part, both models overlap with the observational data within one standard deviation

8. 20 year time-mean change in oxygen (µmol/kg)
(Control - climate change)
most similar at the surface
both show broad decreases in oxygen with climate change regardless of depth
biggest difference between the two models is arguably in the Pacific

9. Percent change between subsampled control and climate change simulations
CM2-1deg
Highest change in oxygen between 1026 and 1028 density surfaces
Generally positive changes mean that oxygen levels are predicted to decrease on these specific density surfaces
similar changes around 55 degrees south, around the region of the ACC
CM2.6 (higher resolution) is able to resolve eddies
eddys have an effect on stratification which could narrow the change of oxygen to smaller density ranges.
could be why CM2.6 shows sharp peaks where oxygen decreases by a large amount

10. Summary
CM2-1deg tends to underpredict the amount of dissolved oxygen compared to observations
Higher resolution CM2.6 is somewhat better at modeling oxygen concentrations but still underpredicts oxygen levels
High variability in the seasonal cycle
Decrease in dissolved oxygen in climate change simulation
Maximum decrease between density surfaces and kg/m 3

11. Increased ventilation/ less respiration Decreased ventilation/
more respiration
AOU changes greater under the thermocline
AOU change due to stratification change (time since water was at the surface)and respiration change
consumption is shifted in the water column
solubility decreases, is especially important near the surface
also near the NADW, which gets warmer according to the model
AOU negative near bottom
convection in wedell seas renewing bottom waters so H2o is newer?
Water temperature
increases
Little or no change
in water temperature