445
12.5
Applications
available as a grid of discrete image points and
cover extensive surface areas of the ocean. Other
examples are data sets, which contain global
information on topographic height or bathymetric
depth (e.g. http://www.ngdc.noaa.gov/mgg/
global/seltopo.html)
Contrary to computer-aided programs used in
map construction, GIS enables a linkage between
various mapping levels and levels of information.
For instance, set operations are performed to
blend and calculate intersections, or enable sectional map coupling procedures. A bathymetric map
can be, for instance, combined with the distributive pattern of benthic particle fluxes. As a result,
the transport across the sediment/water interface
is balanced for specific regions, such as the deepsea, the continental slope, and the shelf area. As
an example, Figure 12.19 described in Section
12.5.2 could not have been produced without a
GIS program.
12.5 Applications
Following the theoretical explanations of the preceding sections, the different conceptual approaches of regionalization and balancing will now be
demonstrated on several practical examples.
12.5.1 Balancing the Diffusion Controlled
Flux of Benthic Silicate in the South
Atlantic - Applications of Kriging
Apart from the benthic oxygen respiration (cf.
next section), there are no globally valid functions of correlation between transfer rates of
other pore water solutes and their possible control parameters. Ultimately, all processes of early
diagenesis are in fact affected by the supply of
organic substance, although other control
parameters are also of relevance for the reflux
rates of most nutrients (e.g. C:N:P - ratio of the
remineralized material, cf. Section 6.2). The
complexity of the benthic system demands a
treatment of spatial differentiation. Like the
biogeochemical provinces applied for calculating
primary production or the TOC-based benthic
provinces mentioned before, correlative dependencies between benthic flux rates and ratelimiting control parameters only possess regionally restricted validity. However, a greater
number of reliable analytical data exist for a
limited number of parameters. In some oceanic
regions, the available database, as well as its
internal geographical resolution, already permit
the application of the kriging method. The following example of the regional distribution of
benthic silicon fluxes in the South Atlantic will
make both aspects evident.
On the basis of comparative pore water
measurements carried out at 76 locations in the
eastern part of the South Atlantic, Zabel et al.
(1998) demonstrated the distribution pattern of
rates of benthic silicate release. As explained in
Section 12.4.2 this extrapolation method always
averages the measured values. From this it follows
that the estimated result of regionalization may
strongly differ from the real values in regions
showing great differences between neighboring
locations (especially in regions of the intensively
studied continental slope). This effect depends on
the resolution of the chosen grid (clustering) and
can require a subsequent manual adaptation of the
kriging results to the database. Figure 12.14a
shows the corrected result of kriging-regionalization.
To regionalize the entire South Atlantic,
Hensen et al. (1998) used an extended database,
which contained 180 single measurements. Contrary to the more detailed map of the eastern parts,
the distribution pattern obtained was not manually corrected subsequently (Fig. 12.14b). In spite
of minor differences in the overall pattern, the
results obtained from both procedures reflect the
distribution of surface-water activity.
There are, however, differences in two areas:
a) in the eastern equatorial Atlantic and b) the
western part of the Argentine Basin. Despite a
high level of production, the benthic silicon
release is quite low in the Guinea and the northern
part of the Angola Basin. Although high
productivity prevails in the western part of the
Argentine Basin, export production is not supposed to supply sufficient amounts of biogenic opal
to match the high flux rates observed in this area.
These specific deviations would never have been
detected by simply applying empirical functions.
Nevertheless it is a fundamental task to explain
such regional deviations from normal ones.
As already described in Section 12.3.3, the
solubility rate of biogenic opal depends on a large
number of processes and environmental conditions. The quality of the various opal phases
and their alteration over time are crucial in this
regard (Archer et al. 1993; Van Cappellen et al.
12.5
Applications
available as a grid of discrete image points and
cover extensive surface areas of the ocean. Other
examples are data sets, which contain global
information on topographic height or bathymetric
depth (e.g. http://www.ngdc.noaa.gov/mgg/
global/seltopo.html)
Contrary to computer-aided programs used in
map construction, GIS enables a linkage between
various mapping levels and levels of information.
For instance, set operations are performed to
blend and calculate intersections, or enable sectional map coupling procedures. A bathymetric map
can be, for instance, combined with the distributive pattern of benthic particle fluxes. As a result,
the transport across the sediment/water interface
is balanced for specific regions, such as the deepsea, the continental slope, and the shelf area. As
an example, Figure 12.19 described in Section
12.5.2 could not have been produced without a
GIS program.
12.5 Applications
Following the theoretical explanations of the preceding sections, the different conceptual approaches of regionalization and balancing will now be
demonstrated on several practical examples.
12.5.1 Balancing the Diffusion Controlled
Flux of Benthic Silicate in the South
Atlantic - Applications of Kriging
Apart from the benthic oxygen respiration (cf.
next section), there are no globally valid functions of correlation between transfer rates of
other pore water solutes and their possible control parameters. Ultimately, all processes of early
diagenesis are in fact affected by the supply of
organic substance, although other control
parameters are also of relevance for the reflux
rates of most nutrients (e.g. C:N:P - ratio of the
remineralized material, cf. Section 6.2). The
complexity of the benthic system demands a
treatment of spatial differentiation. Like the
biogeochemical provinces applied for calculating
primary production or the TOC-based benthic
provinces mentioned before, correlative dependencies between benthic flux rates and ratelimiting control parameters only possess regionally restricted validity. However, a greater
number of reliable analytical data exist for a
limited number of parameters. In some oceanic
regions, the available database, as well as its
internal geographical resolution, already permit
the application of the kriging method. The following example of the regional distribution of
benthic silicon fluxes in the South Atlantic will
make both aspects evident.
On the basis of comparative pore water
measurements carried out at 76 locations in the
eastern part of the South Atlantic, Zabel et al.
(1998) demonstrated the distribution pattern of
rates of benthic silicate release. As explained in
Section 12.4.2 this extrapolation method always
averages the measured values. From this it follows
that the estimated result of regionalization may
strongly differ from the real values in regions
showing great differences between neighboring
locations (especially in regions of the intensively
studied continental slope). This effect depends on
the resolution of the chosen grid (clustering) and
can require a subsequent manual adaptation of the
kriging results to the database. Figure 12.14a
shows the corrected result of kriging-regionalization.
To regionalize the entire South Atlantic,
Hensen et al. (1998) used an extended database,
which contained 180 single measurements. Contrary to the more detailed map of the eastern parts,
the distribution pattern obtained was not manually corrected subsequently (Fig. 12.14b). In spite
of minor differences in the overall pattern, the
results obtained from both procedures reflect the
distribution of surface-water activity.
There are, however, differences in two areas:
a) in the eastern equatorial Atlantic and b) the
western part of the Argentine Basin. Despite a
high level of production, the benthic silicon
release is quite low in the Guinea and the northern
part of the Angola Basin. Although high
productivity prevails in the western part of the
Argentine Basin, export production is not supposed to supply sufficient amounts of biogenic opal
to match the high flux rates observed in this area.
These specific deviations would never have been
detected by simply applying empirical functions.
Nevertheless it is a fundamental task to explain
such regional deviations from normal ones.
As already described in Section 12.3.3, the
solubility rate of biogenic opal depends on a large
number of processes and environmental conditions. The quality of the various opal phases
and their alteration over time are crucial in this
regard (Archer et al. 1993; Van Cappellen et al.