Note that there is detailed discussion of these figure in this issue. Let's keep the discussion on that issue, and we can use this page as a summary of that discussion.
Fig 1 - experimental design - Julia
(a) Climatological wind speed of the control experiment. Blue contour marks the zero divide line (poleward of which the perturbation is applied), black contour marks the 1000m isobath, and yellow boxes mark the DSW formation regions where wind was masked in the additional experiment. Climatological (b) eastward and (c) northward wind velocities with 1000m isobath contour (black).
Fig 2 - summary figure - SSH, and cross-slope transport time series with 12 month running mean. - Julia
Fig 3 - Maps showing UP and DOWN bottom age and salinity anomalies. Maybe decide later if we want 4 panels, or combine with salt on the shelf and age in the abyss. -- Wilma
This figure shows the primary response to the UP winds is an increase in DSW formation, enhanced AABW transport and a consequent decrease in age in the abyssal ocean. Leads to the question: "Why"?
We formulate 3 potential hypotheses to explain the increased DSW formation:
- The UP experiment has enhanced katabatic winds, which increase the surface water mass transformation in specific locations where coastal polynyas allow production of DSW;
- The UP experimental forcing acts to increase Ekman driven upwelling in a broad region across the Southern Ocean (in the perturbation region), increasing along-isopycnal transport in CDW layers and thus bringing more salt onto shelf. This small change in the salt balance enhances transformation of water into denser classes;
- The UP experiment increases northward sea ice transport of fresh water, which increases the salinity on the shelf and thereby enhances transformation rates (as per option 2).
Schematic figure of different hypotheses here??
Fig 4 - show results from "no katabatics" simulation. Add error bar to control to show interannual (or 5 year avg) variability. - Adele
This rules out hypothesis 1.
This section will also list some headline numbers to show that the no katabatics case is the same as UP, but we will just add these to the text.
Fig 5 - show that CDW isopycnals don't (or only rarely) outcrop (ideas: density as function of depth along the 1000m isobath; or find the transect where CDW is shallowest; or ... ?). - Paul?
Place holder figure:
This rules out hypothesis 2 (or at least indicates it's unlikely, and we will return to it later on to more thoroughly rule it out).
Fig 6 - Changes in buoyancy fluxes to show that freshwater fluxes dominate over heat fluxes and that sea ice processes dominate the freshwater flux. - Andy
Fig 7 - Sea ice advection changes - maps of change in sea ice concentration and ice velocity. - Paul
Place holder figure:
The sea ice velocity plot shows that it's the meridional component of the winds that dominate the sea ice advection changes and therefore sea ice concentration changes.
Fig 8 - SSH and DSW cross-slope transport in the UP, UPzonal and UPmeridional experiments with a 12 month running mean- Also include a panel for sea ice export here (difference of melt and sea ice volume changes) Julia and Andy
This shows that the DSW increase in UP is driven by the meridional component of the winds, consistent with the sea ice mechanism explained above.
Fig 9 - Ekman driven upwelling change - using calculation of Ekman pumping binned into surface density bins. - Paul
Placeholder figure (just need top panel here):
This figure is included here to more conclusively rule out hypothesis 2. This plot shows that upwelling in CDW classes increases in UP, and that this is driven by the zonal component of the wind change not the meridional component. The argument here is that even if the CDW isopycnals do outcrop and experience Ekman upwelling (i.e. if our assumption above in Figure 5 was wrong due to stratification within the mixed layer and surface diabatic effects), this Ekman upwelling is coming from the zonal component, which results in no DSW change. So therefore hypothesis 3 is correct.
Fig 10 - Temperature changes - add in zonal and meridional simulations Wilma and Paul
Explanation of temperature changes:
- Warming occurs in UP at or downstream of DSW formation sites due to enhanced upward heat transport that occurs when lower overturning cell increases. We hypothesise this will come from the meridional component.
- Cooling occurs in UP elsewhere (most of West and East Antarctica) due to depression of isopycnals that occurs due to downward Ekman pumping at coast. We hypothesise this will come from the zonal component.
Possibly include a figure of surface temperature changes here or in supplementary? Or not?
Fig 11 - Schematic illustration draft: - Adele