492
Chemical Oceanography, 4th Edition
agreement with the theoretical values in the Cariaco Trench (Zhang and Millero, 1993a,b).
The ratio of ΔTA to ΔTCO 2 in anoxic waters of the Framvaren was 1.14 and equal to the
model value. The same value of ΔTA/ ΔTCO 2 (1.15) was found by Dyrssen (1989). This ratio
was higher (1.32) than the model value in the Black Sea. As suggested by Dyrssen (1985),
the differences between the observed ratios and the model values may be attributed to a
combination of high values of TA and TCO 2 and the low values of H 2 S. One possible explanation of low H 2 S concentration is that a small fraction of hydrogen sulfide is removed by
formation of Fe sulfides (Goyet, Bradshaw, and Brewer, 1991):
2 FeOOH + 3 H 2 S = FeS + FeS 2 + 4 H 2 O
(10.70)
without changing TA and TCO 2 . Goyet, Bradshaw, and Brewer (1991) argued that the
capacity of this reaction must be small because of the low dissolved iron concentrations
and cycling rates in the Black Sea.
Inflow of O 2 -containing water through the sill may cause the removal of H 2 S in the deep
waters of the Framvaren, such as the renewal of the bottom waters when the channel
was dug in 1850. To keep the same deviations in the ratios of TA/ H 2 S and TCO 2 /H 2 S,
the removal process of H 2 S should not change the TA. In other words, sulfide should be
removed from the water column as FeS 2 and elemental sulfur or polysulfides:
2 H 2 S + O 2 = 2 S 0 + 2 H 2 O
(10.71)
(x – 1) S 0 + HS – = HS X
–
(10.72)
again, without altering both TA and TCO 2 . If the product is SO 4
2– instead of S 0 , the alkalinity will be lowered. Millero et al. (1987) found that the main reaction product between H 2 S
and O 2 was SO 4
2– when O 2 was in excess. Experimental errors in the measurements of TA
and H 2 S, and the presence of organic acids that are not taken into account in this simple
chemical theory, are also likely to contribute to the difference.
The calculated pCO 2 in the surface water in June 1993 was 250 μatm. This low pCO 2
was mainly due to the low temperature (15.41°C) and low CO 2 content (875 μM due to low
salinity) in the surface water. If the pCO 2 in the atmosphere is assumed to be 350 μatm, the
surface water of the Framvaren serves as a sink of CO 2 for the atmosphere.
Ammonium started increasing below the interface because of the anaerobic decomposition of organic matter. Based on the model, a plot of NH 4
+ versus H 2 S should give a slope of
0.30. Our experimental result gave the same value, 0.30 (±0.01) for all data from 20 to 175 m
(Figure 10.82). A lower value of NH 4
+ /H 2 S = 0.25 (±0.02) was found for data from 30 to 80 m.
The ratio of NH 4
+ to H 2 S in the Cariaco Trench (0.31; Zhang and Millero, 1993a,b) agreed
with the model value, while it was lower (0.23) in the Black Sea (Table 10.11). Due to the
possible removal of H 2 S as discussed, the real ratio of NH 4
+ to H 2 S must be lower than the
model value. One explanation for this is that the C:N ratio in organic matter was higher
than 106:16. The detailed discussion follows.
Compared with nitrogen, phosphorus transformations are simple in terms of redox
chemistry because of its single oxidation state. However, phosphate can absorb onto particles such as reactive Fe and Mn hydroxides, which results in a lower recycling efficiency
than C and N, especially in upper oxidizing sediments. The rapid increase of concentrations of phosphate below the interface also results from the anaerobic respiration of
organic matter. The ratio of PO 4
3– to H 2 S should be 0.019. Our experimental result gave
exactly the same value, 0.0191 (±0.0004), for all data from 20 to 175 m (Figure 10.82). A
Chemical Oceanography, 4th Edition
agreement with the theoretical values in the Cariaco Trench (Zhang and Millero, 1993a,b).
The ratio of ΔTA to ΔTCO 2 in anoxic waters of the Framvaren was 1.14 and equal to the
model value. The same value of ΔTA/ ΔTCO 2 (1.15) was found by Dyrssen (1989). This ratio
was higher (1.32) than the model value in the Black Sea. As suggested by Dyrssen (1985),
the differences between the observed ratios and the model values may be attributed to a
combination of high values of TA and TCO 2 and the low values of H 2 S. One possible explanation of low H 2 S concentration is that a small fraction of hydrogen sulfide is removed by
formation of Fe sulfides (Goyet, Bradshaw, and Brewer, 1991):
2 FeOOH + 3 H 2 S = FeS + FeS 2 + 4 H 2 O
(10.70)
without changing TA and TCO 2 . Goyet, Bradshaw, and Brewer (1991) argued that the
capacity of this reaction must be small because of the low dissolved iron concentrations
and cycling rates in the Black Sea.
Inflow of O 2 -containing water through the sill may cause the removal of H 2 S in the deep
waters of the Framvaren, such as the renewal of the bottom waters when the channel
was dug in 1850. To keep the same deviations in the ratios of TA/ H 2 S and TCO 2 /H 2 S,
the removal process of H 2 S should not change the TA. In other words, sulfide should be
removed from the water column as FeS 2 and elemental sulfur or polysulfides:
2 H 2 S + O 2 = 2 S 0 + 2 H 2 O
(10.71)
(x – 1) S 0 + HS – = HS X
–
(10.72)
again, without altering both TA and TCO 2 . If the product is SO 4
2– instead of S 0 , the alkalinity will be lowered. Millero et al. (1987) found that the main reaction product between H 2 S
and O 2 was SO 4
2– when O 2 was in excess. Experimental errors in the measurements of TA
and H 2 S, and the presence of organic acids that are not taken into account in this simple
chemical theory, are also likely to contribute to the difference.
The calculated pCO 2 in the surface water in June 1993 was 250 μatm. This low pCO 2
was mainly due to the low temperature (15.41°C) and low CO 2 content (875 μM due to low
salinity) in the surface water. If the pCO 2 in the atmosphere is assumed to be 350 μatm, the
surface water of the Framvaren serves as a sink of CO 2 for the atmosphere.
Ammonium started increasing below the interface because of the anaerobic decomposition of organic matter. Based on the model, a plot of NH 4
+ versus H 2 S should give a slope of
0.30. Our experimental result gave the same value, 0.30 (±0.01) for all data from 20 to 175 m
(Figure 10.82). A lower value of NH 4
+ /H 2 S = 0.25 (±0.02) was found for data from 30 to 80 m.
The ratio of NH 4
+ to H 2 S in the Cariaco Trench (0.31; Zhang and Millero, 1993a,b) agreed
with the model value, while it was lower (0.23) in the Black Sea (Table 10.11). Due to the
possible removal of H 2 S as discussed, the real ratio of NH 4
+ to H 2 S must be lower than the
model value. One explanation for this is that the C:N ratio in organic matter was higher
than 106:16. The detailed discussion follows.
Compared with nitrogen, phosphorus transformations are simple in terms of redox
chemistry because of its single oxidation state. However, phosphate can absorb onto particles such as reactive Fe and Mn hydroxides, which results in a lower recycling efficiency
than C and N, especially in upper oxidizing sediments. The rapid increase of concentrations of phosphate below the interface also results from the anaerobic respiration of
organic matter. The ratio of PO 4
3– to H 2 S should be 0.019. Our experimental result gave
exactly the same value, 0.0191 (±0.0004), for all data from 20 to 175 m (Figure 10.82). A