Abstract:
This data set is the Radiocarbon Tracer Ocean Model Data
Appendix 1, Guilderson et al. (2000):
GEOSECS and model comparison of the penetration and uptake of bomb-radiocarbon.
For definitions see Broecker et al., Global Biogeochem. Cycles, 9, 263-288,
1995.
Penetration depth in meters, DD amplitude is in per mil ("), and specific
column inventory (CI: 10E9 atoms-cm-2) of bomb 14C. These are
... equivalent model
grid-box station comparisons.
MODEL DESCRIPTION:
Ocean model results are from the Lawrence Livermore National Laboratory's
enhanced variant of the Geophysical Fluid Dynamics Laboratory Modular Ocean
Model [Pacanowski et al., 1991]. In the runs presented here, the model was
configured with 23 layers in the vertical, 7 of which were in the upper 300 m.
The model includes the equation of state as well as equations for momentum,
continuity, and tracer transport. Convection is represented by an adjustment
scheme which mixes vertically adjacent grid cells when the potential density of
the overlying cell exceeds that of the underlying one. The model has lightly
smoothed bathymetry at a resolution of 2? latitude x 4? longitude. It
represents flow through all major straits except for the Strait of Gibraltar,
which is accounted for with a source of salt at the appropriate depth horizon.
The simulations presented here include the "Gent-McWilliams" eddy
parameterization [Gent and McWilliams, 1990]. Coefficients of vertical
diffusivity are prescribed and depend on depth. Diffusivities increase from
0.2 cm2-sec-1 at the ocean surface to 1.3 cm2-sec-1 at the ocean bottom. In
addition to treating the physical ocean circulation, the model also calculates
concentrations and fluxes of the individual carbon isotopes. The model
contains a simple "Redfield" biology model [Najjar et al., 1992] or more
appropriately a biological chemical flux model with phosphate as the limiting
nutrient. Fixation of silica by opal producers (diatoms) is allowed to
outcompete calcium carbonate fixation (coccolithophorids), and as a
consequence, the organic carbon to calcium carbonate rain ratio is not fixed a
priori [e.g., Maier-Reimer, 1993]. Alkalinity is conserved and deep-sea
carbonate dissolution occurs in waters that are undersaturated with respect to
calcite or aragonite [Archer, 1991; Maier-Reimer, 1993]. Following the Ocean
Carbon Model Intercomparison Project (OCMIP) standardization, gas exchange uses
the Wanninkhof [1992] wind speed dependence and the solubility of Weiss [1974]
and the model is forced with monthly climatological winds [Hellerman and
Rosenstein, 1983]. This version of the model does not include an interactive
sea-ice model. Instead, we used the climatological distribution of sea ice and
sea ice inhibition of gas exchange [Zwally et al., 1983]. Surface salinities
and temperatures are relaxed to the observed monthly climatology [Levitus and
Boyer, 1994] with a time constant of 60 days. The model is spun-up in an
accelerated mode; it was run for 3300 surface years with an acceleration factor
of 7.5 (equivalent to ~25,000 years) in the deepest model level. We used a
constant zero permil atmosphere and PCO2 of 280 ?atm, to allow for deep ocean
14C equilibration. This seemingly long spin-up is necessary because waters
outcropping in the Southern Ocean must be returned to the deep ocean prior to
complete isotopic equilibration with the atmosphere. After being spun-up, the
model was then run with evolving atmospheric PCO2 and D14C starting in 1765 as
documented by archives and observational networks [Boden et al., 1993].
Similar to Toggweiler et al., [1989], three zonal atmospheric bands are used
for the postbomb atmospheric forcing. 
