(back up to CurrentWork)
Interactive Carbon Cycle
Sep ‘08
for reasons similar to the Schmidt bounds capping (see
CurrentWorkHadOCC) issue, I’ve put a cap of ± 30 moleC/m2/yr on instantaneous CO2 flux exchange between the ocean and atmosphere - there’s a carbon flux spike/heating/positive feedback/explosion problem that turns up when running with the free CO2 system sometimes. I’ve only seen it when running with the standard Wanninkhof piston velocity formulation whilst climate states are spinning up with MOSES2.2 - a short run with the Liss+Mervelat piston velocities didn’t show this (although possibly it wasn’t run for long enough), and neither has a long run using a well spun-up MOSES1 version of the model. The UMUI recommends the Wanninkhof form, and I didn’t want to change a parameterisation that seemed to work fine for most uses of the model, so I’ve put a cap on rather than switch to Liss+Mervelat. Seems to work, so far.
- The problem:
NCAS-Climate has been employed to provide a representation of the carbon cycle in FAMOUS (to form FAMOUS-C). Now that all the pieces are in place, it’s time to get them to talk to each other.
- Roadbumps:
conservative coupling of CO2 fields
CO2 coupling has previously only happened in HadCM3LC, where the ocean and atmosphere grids were congruent. Coupling just involved copying an array. FAMOUS needs the same grid interpolation for CO2 as it has for all the other coupling fields.
spinup/drift of biogeochemical tracers
Spinning up the ocean properly takes a few thousand years for the biogeochemistry, even without sediments. Even with iceberg calving to approximately balance the ocean water budget, there’s still some water flux, so some drift. The vflux_drift correction calculation, normally used to match the global tracer drift to the global water drift, is being used to artificially zero the drift of all ocean tracers (regardless of the actual water flux) whilst spinup is going on. Using UTOPIA advection for the biogeochemical tracers greatly improves
Had OCC’s surface pCO2, and hence carbon fluxes for the spun up state - this is important.
inland seas and surface carbon fluxes
the lack of communication between the inland seas and the rest of the ocean mean that, over time, small imbalances in their local P-E+R can blow up to give huge positive or negative tracer concentrations. Whilst this is inconvenient for, say, salinity, it’s even less helpful for TCO2, where fluxes are exchanged with the atmosphere based on the tracer concentration. For this reason, all biogeochemical tracers and the surface carbon fluxes for these inland seas have been masked out using the same mask employed in the vflux drift calculation. I’ve expanded it to include the (also “inland”) Baltic and Hudson Bay regions, which are capped so they can’t fall below 0 in salinity and which see some pretty high CO2 uptake. I guess they’d sort themselves out w.r.t TCO2 due to this feedback, but it would be inconsistent with salinity, Alk and the water budget in general. I think
- The Climate (May ‘08)
The actual code running at this point (the QUEST job, at least) is in much better shape than it was over the winter. It’s now properly reproduceable over restarts and configs with different numbers of processors, a couple of bugs have been caught and the whole thing is generally tidier. We’re not totally free of bugs, however: all MOSES2.2 runs still hit occasional NaNs in the 10m wind diagnostics (see
CurrentWorkMOSES Nov07) and the fully coupled CO2 run just segfaults after ~200 years (see Sep08 above), although a quick look shows nothing unusual. There’re also some legacy STASH options with MOSES1 diagnostics still left in that fill up with junk which can upset xconv/ff2pp.
Valnotes of the current MOSES2.2 FAMOUS for both fixed CO2 and fully coupled carbon runs are available, compared against each other, MOSES1 FAMOUS and
Had CM 3.
notes for Fixed CO2 MOSES2.2 vs MOSES1
--------------------------------------
the really glaring issue is the warming of most land surface areas by a few degrees. The warm bias extends higher into the atmosphere. The valnote suggests it’s a summer (JJA) problem, even in South America and Africa. In DJF, only Canada sees anomalous warming, and Asia is cooler. North America, Asia and Australia show an increase in net downward SW in JJA, but this doesn’t seem to apply to Africa and South America. The only thing that correlates well with all the warming is a reduction in net downward LW during JJA, which is a reflection of the higher surface temperatures. The switch to MOSES2 does mean some quite large changes in soil moisture, which I think are probably significant (NB the soil moisture plots in the valnote aren’t directly comparable due to changes in this diagnostic between the versions of MOSES)