F. Große et al.: Looking beyond stratification 
2527 
www.biogeosciences.net/13/2511/2016/ 
Biogeosciences, 13, 2511-2535, 2016 
The analysis clearly shows vertical mixing is the only ef 
ficient gain term for O2 in the bottom layer. Benthic aero 
bic remineralisation constitutes the major driver for O2 defi 
ciency in the North Sea O2 minimum zone, although pelagic 
aerobic remineralisation still has a significant effect on bot 
tom O2, but accounts for a remarkably lower proportion than 
in the entire sub-MLD volume. The simulated benthic rem 
ineralisation rates are in the same order as those derived from 
observations giving a range from 7 to 25 mmol 02 m -2 d _1 
for a nearby station (station 3 in Upton et al., 1993), how 
ever, rather at the lower end of this range. 
3.6 Spatial variability in the North Sea bottom O2 
dynamics 
Figure 9a and b show the bottom O2 mass balances of re 
gions 4 (southern North Sea) and 5 (northern North Sea) in 
2002 (see Fig. 2 for location of regions). Both regions show 
different stratification periods than in region 3. In region 4, 
the period between the first and last day of stratification ac 
counts for only 163 days compared to 187 in region 3. Addi 
tionally, stratification is temporarily intermittent in late April 
and early July. In region 5, stratification lasts for 211 days 
without any interruptions. The water depths and bottom layer 
volumes also differ between the regions. Region 3 has an av 
erage water depth of 47.75 m and a bottom layer volume of 
about 14.4 km 3 , while region 4 is characterised by an aver 
age depth of 45 m and a bottom layer volume of 11.9 km 3 . 
Region 5 has an average bottom depth of 99.23 m and a bot 
tom layer volume of 16.3 km 3 . 
The O2 concentrations at the beginning of the stratified 
period in regions 4 (9.74mg02L _1 ) and 5 (9.46mg02L _1 ) 
are lower than in region 3. In contrast, both regions show 
higher concentrations at the end of stratification compared to 
region 3. These values reach 6.98 mgCUL -1 in region 4 and 
7.61 mgCUL -1 in region 5, compared to 6.76mg02L _1 in 
region 3. 
In region 4, intense mixing in late June/early July, indi 
cated by the steep increase in MIXo 2 , causes the breakdown 
of stratification and the complete replenishment of bottom 
O2. Integrated over the stratified period the effect of MIXo 2 
is almost 1.5 times higher than in region 3. In region 5, 
MIXo 2 represents a significantly lower supply of O2, which 
mainly relates to the significantly greater water depth favour 
ing more stable stratification. ADVo 2 has an opposite effect 
in regions 4 and 5 relative to region 3, however, showing only 
minor negative integrated effects. 
The integrated biological O2 consumption in region 4 is 
about 7 % less than in region 3. Referring to changes in 
O2 concentrations, the consumption even exceeds that in re 
gion 3 by about 6 %, due to the thinner bottom layer, and cor 
responds to an EXP org below 25 m depth (calculation analo 
gous to Table 1) in region 4 being 19% higher than in re 
gion 3. In the deeper region 5, EXP org accounts for 62 % of 
that in region 3, while biological O2 consumption accounts 
for only 35 % of that in region 3. 
Despite the differences in the overall biological O2 con 
sumption, the relative contributions of the different sink 
processes in region 4 are in the same order as in re 
gion 3. REM S ed represents the largest contributor with about 
54.6 % of the total biological consumption, while REM pe i ac 
counts for 27.2 %. Thus, the combined effect of REM se d and 
REMpei accounts for 81.8 % which is similar to region 3. Av 
erage daily benthic remineralisation rates are higher than in 
region 3 and yield 8.9 mmolCU m -2 d _1 . The relative contri 
butions for RES zoo and NIT result in 13.6 and 4.6 %, respec 
tively. 
The comparison of the relative contribution of REM pe i and 
REM S ed in region 5 reveals some changes compared to re 
gions 3 and 4. REM S ed accounts for about 70 % of the to 
tal biological O2 consumption, while REM pe i contributes to 
only about 20 %. This relates to the generally lower amount 
of exported organic matter reaching the model bottom layer 
(63 % of that in region 3). On the one hand, this causes lower 
O2 consumption due to REM pe i, and on the other hand, en 
hances REMsed relative to REM pe i as more organic matter 
reaches the bottom. 
Considering the combined effect of stratification and bio 
logical consumption in region 4 reveals that the shorter strat 
ification period and the strong mixing prohibit the evolution 
of low O2 conditions in this region, despite the highest bi 
ological consumption. The higher benthic remineralisation 
rate compared to region 3 underlines the high potential for 
low O2 conditions in the Oyster Grounds under persistent 
seasonal stratification. This is in good agreement with the 
findings by Greenwood et al. (2010), who observed bottom 
O2 concentrations less than 6mg02L _1 in this area. As in 
region 3, the simulated benthic remineralisation rate lies at 
the lower end of the range of 5.6 to 30.6 mmol02 m -2 d _1 
obtained from observational studies near this site (Lohse 
et al., 1996; Weston et al., 2008). This suggests that the 
model most likely underestimates benthic remineralisation 
rates. 
In region 5, the amount of exported organic matter reach 
ing the bottom layer (deepest pelagic model layer) is limited 
due to the great water depth. Thus, biological consumption 
in the bottom layer is low, preventing the evolution of O2 de 
ficiency, even though stratification lasts longer and is more 
stable than in the other regions. This suggests that region 5 
is unlikely to be affected by low bottom O2 concentrations. 
Jfowever, Queste et al. (2013) found bottom O2 concentra 
tions of about 6mg02L _1 near this area in 2010, which in 
dicates that this area can be affected by O2 deficiency. 
3.7 Interpreting observed bottom O2 at North Dogger 
The O2 observations at station North Dogger (Greenwood 
et al., 2010) shown in Fig. 3a and b showed similar O2 con 
centrations of about 9.5mg02L _1 at the beginning of the