DATA ASSIMILATION IN BIOGEOCHEMICAL MODELLING
531
ological production is simulated by restoring to either zero or
served nutrient concentrations in the light-lit surface layer, and
instant sinking and remineralisation are accounted for by a prescribed remineralisation profile. Examples are the models of Bacastow and Maier-Reimer [1990], Najjar et al. [1992] and the models used during phases 1 and 2 of the Ocean Carbon Model Intercomparison Project (OCMIP, [Orr, 1999]). Depending on whether
or not dissolved organic matter is explicitly resolved, these biogeochemical/ecosystem models typically have 2 to 4 parameters.
They are most widely used in models that address time scales
much longer than a year, and applications to seasonal or shorter
time scales will be problematic because of the absence of any particulate organic-matter storage pools.
NPZD-type models. Although NPZD stands for Nutrient, Phytoplankton, Zooplankton, and Detritus, such models may contain
a few more prognostic variables like bacteria or dissolved organic
matter. Most of these models are descendants of a configuration
proposed by Fasham et al. [1990] and they explicitly simulate
the cycling of either nitrogen or phosphorus. They have been
plied to general ocean circulation models ranging from coarse resolution [Sarmiento et al., 1993; Fasham et al., 1993; Chai et al.,
1996; McCreary et al., 1996; Six and Maier-Reimer, 1996] to eddypermitting [Oschlies and Gar¸ con, 1998, 1999] and eddy-resolving
resolution [Oschlies, 2002]. Typically, these ecosystem models have
10 to 30 parameters.
Functional-group type models. Going beyond the NPZD-type structure, these recently emerging models attempt to resolve different
species or groups of phytoplankton and zooplankton. According
to their special ecological function (e.g., nitrogen fixation, calcification) these are often called functional groups. These different groups require (and allow) to explicitly resolve the cycling
of different biogeochemical elements. Examples are the models
described by Moore et al. [2002] and Aumont et al. [2004], as
well as the European Regional Seas Ecosystem Model (ERSEM,
Ebenh¨ oh et al. [1997]) and the evolving Dynamic Green Ocean
Model [Le Qu´ er´ e et al., Ecosystem dynamics based on plankton
functional types for global ocean biogeochemistry models, submitted to Global Change]. They typically have far more than hundred
parameters.
Bi
ob
,
,
,
,
,
ap
531
ological production is simulated by restoring to either zero or
served nutrient concentrations in the light-lit surface layer, and
instant sinking and remineralisation are accounted for by a prescribed remineralisation profile. Examples are the models of Bacastow and Maier-Reimer [1990], Najjar et al. [1992] and the models used during phases 1 and 2 of the Ocean Carbon Model Intercomparison Project (OCMIP, [Orr, 1999]). Depending on whether
or not dissolved organic matter is explicitly resolved, these biogeochemical/ecosystem models typically have 2 to 4 parameters.
They are most widely used in models that address time scales
much longer than a year, and applications to seasonal or shorter
time scales will be problematic because of the absence of any particulate organic-matter storage pools.
NPZD-type models. Although NPZD stands for Nutrient, Phytoplankton, Zooplankton, and Detritus, such models may contain
a few more prognostic variables like bacteria or dissolved organic
matter. Most of these models are descendants of a configuration
proposed by Fasham et al. [1990] and they explicitly simulate
the cycling of either nitrogen or phosphorus. They have been
plied to general ocean circulation models ranging from coarse resolution [Sarmiento et al., 1993; Fasham et al., 1993; Chai et al.,
1996; McCreary et al., 1996; Six and Maier-Reimer, 1996] to eddypermitting [Oschlies and Gar¸ con, 1998, 1999] and eddy-resolving
resolution [Oschlies, 2002]. Typically, these ecosystem models have
10 to 30 parameters.
Functional-group type models. Going beyond the NPZD-type structure, these recently emerging models attempt to resolve different
species or groups of phytoplankton and zooplankton. According
to their special ecological function (e.g., nitrogen fixation, calcification) these are often called functional groups. These different groups require (and allow) to explicitly resolve the cycling
of different biogeochemical elements. Examples are the models
described by Moore et al. [2002] and Aumont et al. [2004], as
well as the European Regional Seas Ecosystem Model (ERSEM,
Ebenh¨ oh et al. [1997]) and the evolving Dynamic Green Ocean
Model [Le Qu´ er´ e et al., Ecosystem dynamics based on plankton
functional types for global ocean biogeochemistry models, submitted to Global Change]. They typically have far more than hundred
parameters.
Bi
ob
,
,
,
,
,
ap