528
ANDREAS OSCHLIES
impact of the physics on the marine biogeochemistry. Here, a potential
mapping of errors of the physical model onto the predicted ecosystem
fields makes the separate evaluation of the ecosystem model component
difficult. This is not necessarily a disadvantage: Because of the marine biology’s strong sensitivity and fast response time of the order of
days to changes in the light or nutrient supply, coupling ecosystem and
circulation models may actually help to identify deficiencies of physical transport processes, particularly in the upper ocean [e.g., Oschlies,
1999].
The following section will give a brief overview over the field of biogeochemical models and presently used marine ecosystem models. Section 3
discusses some aspects of observations that are relevant for data assimilation, and section 4 addresses the potential use of combining data and
biogeochemical models. Data assimilation methods are presented in section 5, and this article ends with a discussion of some achievements and
perspectives of data assimilation in the field of biogeochemical modelling.
2.
Biogeochemical modelling
Compared to numerical modelling of the ocean circulation, the field
of biogeochemical modelling is much less mature. In particular, there is
no known equivalent to the Navier-Stokes equations. In principle, these
describe the motion of sea water exactly, but an exact solution does not
(yet?) exist. The rules are thus clear for physical models, and different
numerical models basically attempt to find different approximations to
the unknown exact solution.
The situation is different in the field of biogeochemical modelling.
Although there are some reliable, albeit mainly empirical, laws that
scribe transformations among various inorganic compounds dissolved in
sea water, things become relatively shaky once life, and thereby transformations among organic and inorganic chemical compounds, comes into
play. In practice, biogeochemical models are generally composed of an
inorganic chemistry component and an ecosystem component, of which
the latter is the by far more complex, expensive, and problematic part.
In the following I will focus on this ecosystem model component and
often use the term ecosystem model as synonym for the entire biogeochemical model.
Marine ecosystem models usually attempt to describe life’s action
on marine biogeochemical tracers by partitioning the marine ecosystem
into a handful of boxes, often called compartments, such as phytoplankton (plants), zooplankton (animals), or detritus (dead organic matter).
Sometimes, a further distinction is made between particular and
e
d
ANDREAS OSCHLIES
impact of the physics on the marine biogeochemistry. Here, a potential
mapping of errors of the physical model onto the predicted ecosystem
fields makes the separate evaluation of the ecosystem model component
difficult. This is not necessarily a disadvantage: Because of the marine biology’s strong sensitivity and fast response time of the order of
days to changes in the light or nutrient supply, coupling ecosystem and
circulation models may actually help to identify deficiencies of physical transport processes, particularly in the upper ocean [e.g., Oschlies,
1999].
The following section will give a brief overview over the field of biogeochemical models and presently used marine ecosystem models. Section 3
discusses some aspects of observations that are relevant for data assimilation, and section 4 addresses the potential use of combining data and
biogeochemical models. Data assimilation methods are presented in section 5, and this article ends with a discussion of some achievements and
perspectives of data assimilation in the field of biogeochemical modelling.
2.
Biogeochemical modelling
Compared to numerical modelling of the ocean circulation, the field
of biogeochemical modelling is much less mature. In particular, there is
no known equivalent to the Navier-Stokes equations. In principle, these
describe the motion of sea water exactly, but an exact solution does not
(yet?) exist. The rules are thus clear for physical models, and different
numerical models basically attempt to find different approximations to
the unknown exact solution.
The situation is different in the field of biogeochemical modelling.
Although there are some reliable, albeit mainly empirical, laws that
scribe transformations among various inorganic compounds dissolved in
sea water, things become relatively shaky once life, and thereby transformations among organic and inorganic chemical compounds, comes into
play. In practice, biogeochemical models are generally composed of an
inorganic chemistry component and an ecosystem component, of which
the latter is the by far more complex, expensive, and problematic part.
In the following I will focus on this ecosystem model component and
often use the term ecosystem model as synonym for the entire biogeochemical model.
Marine ecosystem models usually attempt to describe life’s action
on marine biogeochemical tracers by partitioning the marine ecosystem
into a handful of boxes, often called compartments, such as phytoplankton (plants), zooplankton (animals), or detritus (dead organic matter).
Sometimes, a further distinction is made between particular and
e
d