Inclusion of Iron
in a Biophysical Model of the Gulf of Alaska -
What Have we Learned About Cross-Shelf Transport?

S. Hinckley (AFSC), A.J. Hermann (PMEL/JISAO),
& E.L. Dobbins (PMEL/JISAO)

Abstract:

One of the main questions about the coastal ecosystem in the Gulf of Alaska is how nutrients and ecosystem components from the central Gulf make their way onto the shelf, fueling productivity in what is mainly a downwelling system. Our previous modeling efforts employed an NPZ model (with nitrate, ammonia, large and small phytoplankton, large and small microzooplankton, two copepods, euphausiids and detritus) embedded within a physical model (the Regional Ocean Modeling System (ROMS)). A centeral issue in our biophysical modeling is the need to follow parcels of High Nutrient Low Chlorophyll (HNLC) oceanic water between the open basin and the coastal regions, while preserving the representation of these two distinct ecosystems within one model. We have incorporated iron into our model as a way to address the underlying hypothesis that iron deficiency in the basin drives the differences between the oceanic and coastal ecosystems. We compare results with and without iron limitation to data collected by the GLOBEC program on a line offshore of Seward, Alaska.

Methods:

Model Overview:
  • 10 compartment NPZ model plus iron limitation
  • 2 grids: 11 km NEP and 3 km CGOA
  • Realistic bathymetry, 30 vertical levels
  • Realistic forcing:
    • Atmospheric variables from MM5 nested atmospheric model
    • Monthly river and Alaskan linesource freshwater input
    • 8 tidal components (CGOA)
NPZ flowchart
Map of the Gulf of Alaska grid with bathymetry. Flowchart of the 10 compartment NPZ model.

Iron:

iron profile

To follow parcels of nutrient rich waters (HNLC) from the open Gulf of Alaska to coastal waters, at surface and at depth, the model must represent both ecosystems.

Iron limitation employs a Michaelis-Menton function (as in Fennel et al. 2003) as a multiplicative factor affecting growth of both small and large phytoplankton. It affects large phytoplankton more strongly.

  • Iron field initialized with onshore/offshore values based on data
  • Iron depleted, but not followed through ecosystem
  • Nudged back to initial profile with 30 day timescale
Section of iron along the Seward Line.

Onshore, the main nitrogen pathway is from nitrate through large phytoplankton to copepods.

Offshore without iron limitation, the main nitrogen pathway is also through large phytoplankton to copepods, which is not what we see in the open ocean.

flowcharts showing iron's effect

If we add iron limitation, which mostly affects the oceanic areas due to lower levels of iron in those areas, the largest part of the fluxes now goes through small phytoplankton to small microzooplankton. We are better able, therefore, to reproduce the main characteristics of the HNLC areas.

Flowcharts of model with and without iron limitation
NO without iron limitation
  • With no iron limitation term, during late May/ early June (after spring blooms), nitrate is depleted in the surface in all areas (left). Offshore areas should retain high surface nutrients at all times.
  • With the iron limitation scheme implemented (right), the oceanic areas no longer show iron depletion in the surface waters because lower iron concentrations there inhibit the bloom of phytoplankton.
NO with iron limitation
Surface nitrate in NEP grid without iron limitation in late May Surface nitrate in NEP grid with iron limitation in late May

Comparisons with Data:

chlor data vs model animation

In the movie comparing measured chlorophyll from SEAWIFS and model output:

  • March: very little activity.
  • April: some activity beginning on the shelf, particularly in the Seward area and Southeast Alaska.
  • May: somewhat more activity in the data around Kodiak Island, whereas the model shows more activity in the southeast.
  • June: the model shows nutrient depletion beginning in the nearshore areas. Both show activity around Kodiak Island and Prince William Sound.
  • July: more depletion on the coast in the model. Interesting shelf-break front of high production.

Lots of mesoscale activity (eddies) in May and June. These are often seen in snapshots of chlorophyll

Animation of chlorophyll from SEAWIFS and model.
a)nutrients along Seward b)phyto along Seward

Computed the average over the top 100 m of model output in three areas: the inner shelf (represented by GAK stations 1 and 2), the middle shelf (GAK stations 6 and 7) and the outer shelf (stations 11, 12, and 13).

  • Nitrate (upper left): the model does a pretty good job
  • Chlorophyll (lower left): magnitudes OK but miss timing of bloom
  • Copepods (upper right): magnitudes reasonable but miss timing of bloom
  • Neocalanus (middle right): underestimated by model
  • Euphausiids (lower right): look good
c)copepods along Seward d)neocalanus along Seward e)neocalanus along Seward
Depth averaged model output (black line) and data (red stars) for 3 regions along the Seward Line, a) nutrients, b) chlorophyll.

 Depth averaged model output and data for 3 regions along the Seward Line, c) Copepods, d) Neocalanus, and e) Euphausiids.

Nutrient Supply:

chlor isosurface

To the left is a picture of the 2 mg/m3 chlorophyll isosurface in quasi-3D from the region just to the northeast and east of Kodiak Island on DOY 209 (July 28th). Bathymetry is gray, and the chlorophyll isosurface is green. The solid black line is the location of the transect in the right figure.

Note how Portlock and Albatross Banks have chlorophyll "holes".

To the right, is a side view of the same chlorophyll isosurface with a transect of nitrate shaded from orange (high nitrate) deep in the water column to blue (depleted) at the surface. Topography is white, shaded with gray. Note:

portlock bank
Chlorophyll isosurface.
  • The column of water over Portlock Bank in the center, marked by an intermediate value of nitrate, probably due to enhanced mixing over the bank
  • A possible, localized front around the bank marked with slanted arrows, where perhaps nitrate is mixing up from the deeper depths
  • The chlorophyll ³hole² in the surface waters, perhaps because
    • Vigorous mixing over the bank itself mixes the chlorophyll too deep
    • Higher chlorophyll around edges responding to enhanced nutrient levels from the deeper waters at the front
Nutrient cross-section and Chlorophyll isosurface over Portlock Bank.
nutrient flux

This is a picture of the nitrogen flux at 50 m - the velocity times the nitrogen concentration at each location.

  • "Rivers" of nutrients
  • The cross-shelf transport produced by a 200 km scale eddy southeast of Kayak Island pulls deep water onto the shelf.
  • We see this at other times of year as well.
  • Some evidence of influx at Amatuli Canyon.
Nutrient flux.

Next Steps:

  • More analysis
  • Add other salmon food
  • Model experiments designed to address GLOBEC NEP hypotheses

Funding support provided by GLOBEC