[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

carbon constants



Hi Ernst,
	I am sorry if I am being dense, but I am still not following all
of your objections. Before I get into the specifics let me just remind
everyone that the point was to provide the Merbach carbon constants for
those that would like to use them. If you don't like our approach for 
solving the carbon chemistry than simply cut out the pieces you don't
like and use what you do want. There are several ways of getting at the
same thing and our intention was not to force our approach on anyone else.

1) Your suggested use of CO3-- to derive Alk as a function of DIC and H
	is equivalent to what we did. If you look in ta_iter_1.f the
	function "fn" is alkalinity defined with DIC and H as the only
	unknowns (same as yours). The function you gave is equivalent to 
	the first two terms in fn, but we have included additional terms 
	for the secondary species. TA is subtracted from fn as the last 
	term because the fitting routine works to get fn as close to zero
	as possible.
	
2) I don't understand why you were irritated by the use of 6 and 9 as initial
	bounds for the fitting routine. Both of our approaches require an
	iterative routine to determine the H concentration. Yours requires
	a single initial guess which you set to 10**-9 which converges in 8
	iterations according to your first message. We use the Newton-Raphson
	technique which requires a high and low initial guess. All we were
	saying was that if you use 10**-6 and 10**-9 as your initial guesses
	it will take 12 iterations to converge. If you can narrow that range
	just a little you can get convergence in 4 iterations which was a goal
	stated at the OCMIP meeting in Boulder. I think we are simply talking
	about two slightly different approaches to do the same thing. I am by
	no means saying that our approach is better...that is simply the way
	we approach it.
	
3) Your third point is, I think, the most important and interesting. The issue 
	of ignoring the secondary species is a matter of cost (in terms of 
	computer time) versus being chemically complete. The secondary species
	do influence the total alkalinity. I am not a modeler. Bob and I were 
	asked to put together a routine that is in line with what we, as a CO2 
	measurement community, are doing. We included the species that we feel
	are important from a measurement point of view, but may have too small
	of an effect to worry with in the models. I think that is a 
	decision that the modeling community needs to make. As you said in your
	last message, borate contributes about 20%. In the example I gave in
	my last message, phosphate and silicate made a difference of about 7 ppm
	in the calculated fCO2 in the Southern Ocean. You can relate this back
	to a change in DIC after equilibration with the atmosphere using the
	Revelle factor. It is not very large, but is it significant? People 
	often make the point that the anthropogenic uptake of 2 PgC/yr could
	be achieved with only a 7ppm delta pCO2 over the oceans. 
	
	You are probably right that the species will be partly mimicked in the
	tuning of inventories with the preindustrial atmosphere. I guess it
	depends on what you feel is important to get right. We have compared 
	estimates of preindustrial DIC based on modern DIC measurements and 
	C* anthropogenic CO2 calculations with the Princeton OBM and found that 
	the model had much more DIC than we estimate. By tuning the carbon 
	inventories to compensate for non-carbon species you are complicating 
	any direct comparison with measurements. If you create an artificial 
	7 ppm delta pCO2 that causes the model oceans to take up CO2 before 
	reaching steady state you may change the alkalinity:DIC ratio which in
	turn may effect anthropogenic CO2 uptake estimates. Again, I cannot
	honestly say whether any of this is significant or not. I think someone
	needs to try it both ways in the model as see.
	
I hope these comments have helped to clear up any confusion. I would like to 
hear what others think about the inclusion of the secondary species. If OCMIP
decides not to include them we can change the routines in about 10 minutes. 
I look forward to continuing the discussion.

Chris Sabine