|
The NPZ model for the CGOA consists of ten compartments: nitrate,
ammonium, small phytoplankton, large phytoplankton, large and small
microzooplankton, small copepods, large oceanic copepods, euphausiids and
detritus.
Nitrogen is modeled as both nitrate and ammonium so that the
proportion of new versus recycled production can be described, along
with the effect of the nitrogen source on the food web. Phytoplankton
is modeled as two size fractions which differ in both their use of nitrate
vs. ammonium, and their response to iron and light limitation.
The two phytoplankton sizes support different herbivore species
complexes. Small microzooplankton, represented by heterotrophic
nanoflagellates, ciliates and medium dinoflagellates, eat only small
phytoplankton. Large microzooplankton, represented by large heterotrophic
dinoflagellates, eat only large phytoplankton. A single compartment is
included for small copepods which reproduce in response to the spring bloom.
These are represented by Pseudocalanus, spp. Large oceanic copepods are
represented by Neocalanus, spp., and their behavior includes simulated diapause.
Euphausiids are represented by Euphausia pacifica.
|
 |
The model optionally includes iron limitation of phytoplankton growth.
Iron is initialized to a field based on data, is drawn down by phytoplankton
growth, and is nudged back to the initial field with a timescale of 30 days.
The iron field, which is high in coastal areas and freshwater runoff and
low offshore, is designed to test the hypothesis that iron limitation is a
main cause of the observed differences between coastal and oceanic ecosystems.
This model is configured as a Fortran subroutine in the Regional Ocean
Modeling System (
ROMS). As
such, it can be run in 1D or 3D, in a variety of grid resolutions, and with
realistic or idealized physical forcing.
We ran this model in both the NEP and
CGOA grids for the year 2001. Model output
were compared to data collected along the GLOBEC LTOP line.
Results were
presented at the 2004 Ocean Sciences Meeting.
Early Results
1-D Model Results
The model was initially developed in the C programming language and
used stored physical fields to drive the biology. The model was also
configured somewhat differently.
Here are some preliminary results from that model:
3-D Model Results
With a 3-D version of that model (the C version), we conducted an
experiment to test
whether downwelling could supply enough nutrients to the CGOA to drive
summer biological production. The results were presented at the 2001
PICES conference, and are being written up as a paper.
Comparison of 1-D Gulf of Alaska NPZ with 1-D NPZD model
A core hypothesis of the Northeast Pacific GLOBEC program is that the
ecosystem dynamics of the California Current System (CCS) and Coastal
Gulf of Alaska (CGOA) are connected by basin-scale physical processes.
One of the primary goals of our modeling is to test this hypothesis, but
how do you compare the results of differently structured biological models
in different physical regimes? In order to compare the biological models,
independent from physical conditions, we developed a 1-D test case of
ROMS. Into this test problem, we inserted both CGOA's NPZ model (10 boxes,
83 parameters) and CCS's much simpler NPZD model (4 boxes, 12 parameters).
The results were presented in this poster
at the 2002 Eastern Pacific Ocean Conference.
|
http://www.pmel.noaa.gov/~dobbins/glnpz/index.html