134 F. Schütte et al.: Hidden vortices: near-equatorial low-oxygen extremes driven by high-baroclinic-mode vortices 
[al 
IE 
26 
= 26.1} 
> 
> 
2 
€ 26.2 
„ 26.3 
>: 
26.4 
26.5 nn 
-0.1 -0.08 -0.06 -0.04 0.02 u 0.02 
Absolute Salinity [g kg] 
© 
a 
NOx [UMmol/kg] 
10 15 20 25 an 
da al 
515 
2, 
O10 
263 
DA 
h 
L 
„oo 35:5 
SA [g kg] 
100 
5 1007 
300 
24°-21°W, all 
— 24°.21°W, <60 ymol Kg % ‚mean 
+ “24°-21°W, >60 ymol kg”, mean 
*#» 24°-21°W, WOA23 
—21°-18°W, all 
— 21°-18°W, <60 ymol kg 77 mean 
* =21°-18°W, >60 umol kg”!, mean 
an» 21°-18°W, WOA23 
1»=« 18°W-African coast, 8°-10°N,WOA23 _. ı 
EA 
500 
0.04 0.06 0.08 0.1 
Nnınımum 02 
02<60umol/kg (N=323) 
02>=60uUMol/kg (N=1696) 
NOx (where 02<60umol/kg, N=35) 
NO, (where 02 >=60umol/kq, N=150) 
— 
HOHS 
oO: 
2 
[p! 
mol/kı 
gl 
vv 
Figure 8. (a) Large panel: anomalies of absolute salinity as a function of potential density in the eastern tropical North Atlantic for two 
different box regimes (Red: 24-21° W, 6-12° N/Blue: 21-18° W, 612° N). The boxes are highlighted in Fig. 4. The anomalies are referenced 
:o the mean profile of absolute salinity that was calculated from all hydrographic profiles found in both boxes. Thin solid lines denote all 
individual profiles and thick solid (dashed) lines show the average of the profiles, that are related to minimum DO concentrations below 
(above) 60 umol kg7! in the upper 200 m. Shadings to the average profiles illustrate the respective standard errors (see text for details). 
Blue and red dotted lines denote climatological profiles for the two boxes. Black dotted line shows the climatological profile for a third 
50x (18° W-African coast, 8-10° N), which defines the near-coastal regime off West-Africa. Inlet panel: mean characteristics of absolute 
salinity versus conservative temperature for the box 24-18° W, 6-12° N, taken from all CTD-O observations in this regime. Thick black line 
denotes the characteristics in the potential density range 25.9 to 26.5kg m”? and is the reference profile for the anomalies shown in the 
ıarge panel. (b) The blue curve shows the median of all oxygen CTD profiles with a minimum below 60 umol kg! in the upper 200 m. The 
red stars indicate the depths and dissolved oxygen concentration of these minima. Orange curves represent profiles with a minimum above 
60 umolkg !. Shaded areas indicate the standard deviation. The turquoise line depicts the mean nitrate profile for the profiles with oxygen 
minima below 60 umol kg” 1 and the yellow line shows the mean nitrate profile for the profiles with minima above 60 umol kg! 
tion that is found in the region 30-24° W, 8-12° N is in good 
qualitative agreement (though located further west) with the 
observed DO distribution (Fig. 4b versus Fig. 4a). In the lon- 
gitude range 24-21° W, low-oxygen events are less likely. It 
should be noted that Fig. 4a and b compare individual ship- 
board observations with the 20-year model climatology. Ob- 
servations represent snapshots of specific events, whereas the 
model averages over a longer temporal period. Consequently, 
apparent differences in the zonal distribution of low-DO ex- 
tremes are expected and do not necessarily indicate a system- 
atic model bias. 
From the GFDL CM2.6 model, we identified a HBV with a 
low-DO core in the near-equatorial open ocean as exemplar- 
ily shown at the position 10° N/28° W (Fig. 6g-1). The spatial 
extent is comparable to our observational results (Fig. 6a—- 
f). A meridional cross section through the simulated HBV 
reveals the low-DO core at 80m depth (isopycnal surface 
26.5 kg m”®) with a lateral extent of about 1° in latitude and 
a vertical extent between about 50 and 150m (Fig. 6h). The 
minimum DO is lower than 60 umolkg”!, whereas DO out- 
side the HBV is at values above 150 umolkg”!. Distribu- 
tions of conservative temperature and potential density show 
Ocean Sei... 22. 119-143. 2026 
shallowing and deepening isopycnal surfaces above and be- 
low the DO minimum, respectively, indicating a weakened 
stratification and consequently low PV at the low-DO core 
(Fig. 6i). The HBV’s velocity signature is strongly con- 
fined to subsurface depths and vanishes above 50 m (Fig. 6j 
and k). In particular, surface velocity does not show any co- 
herence with the subsurface velocity field at depth of the 
HBV (Fig. 6g). This substantiates our observational results 
that these HBVs can hardly be identified from the surface 
geostrophic velocity field obtained from satellite observa- 
tions. The HBV core exhibits low PV water, where mini- 
mum PV is found slightly deeper than the DO minimum 
(Fig. 61). This low PV water core is laterally isolated from 
the surrounding high PV water, but also separated from the 
deeper low PV water through an intermediate PV maximum 
along the isopycnal surface 26.7 kg m7*. This isolation is 
the prerequisite for a persistent eddy with a long-life time. 
The model tends to slightly underestimate PV and associated 
O>» anomalies, indicating somewhat weaker eddy coherence 
compared to observations. At the same time, due to reduced 
dissipation in the circulation model, we expect lifespans of 
the eddies to be slightly prolonged. Additionally, the MiniB- 
https://doi.org/10.5194/o0s-22-119-2026