Compute the iron sediment forcing (fedsedflux) supplied to the model

Implements an approach to computing fesedflux originally in an IDL routine by J. K. Moore.

fesedflux includes two components:

• fesedflux_oxic: a constant low background value; increased in regions of high bottom horizontal current speed (sediment resuspenion) by up to a factor of 100.
• fesedflux_reduce: source everywhere linearly related to the sinking POC flux by coef_fesedflux_POC_flux, except:
  • source is zero below POC_flux_gCm2yr_min (3 gC m$^{-2}$ yr$^{-1}$ in CESM2), and
  • constant above POC_flux_gCm2yr_max.
  • This puts a source on the shelf, and along productive slope/margins, but has little source in the deep ocean, where almost all the remineralization is oxic right on the sediment surface.

fesedflux is computed on subgrid-scale bathymetry, using the fraction of each cell that is ocean bottom at each depth: fesedfrac. fesedfrac is computed from ETOPO1 bathymetry and modified as follows:

• a minimum percentage of each grid cell that is sediments (land_adj_sedfrac_min) is applied to all land-adjacent grid cells.

Arbitrary modification to this objective scheme:

• fesedflux_reduce is multiplied by 10 in the western equatorial Pacific (135-200E, 15S-15N, above 504 m).

Procedure

1. Prepare fesedfrac: sedfrac_compute.ipynb;

2. Take time mean of POC_FLUX_IN, UVEL, and VVLEL from previous model solution: _poc_flux_bottom_velocity_inputs.ipynb

3. Compute fesedflux_reduce:

   • Read fesedfrac, determine land-adjascent points;
   • Create sedfrac_mod by applying land_adj_sedfrac_min.
   • Read POC_flux and convert units;
   • Where POC_flux < POC_flux_gCm2yr_min, POC_flux = 0.;
   • Where POC_flux > POC_flux_gCm2yr_max, POC_flux = POC_flux_gCm2yr_max
   • fesedflux_reduce = POC_flux * coef_fesedflux_POC_flux * sedfrac_mod
   • Apply ad hoc scaling in to select regions.

4. Compute fesedflux_oxic:

   • Read UVEL and VVEL and compute current_speed
   • Where current_speed < 1.0: current_speed = 1.0
   • Where current_speed > 10.0: current_speed = 10.0
   • fesedflux_oxic = coef_fesedflux_current_speed * sedfrac * current_speed**2

5. Output fesedflux_oxic and fesedflux_reduce in model units: µmol/m^2/d