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

Re: Returned mail: Private list 'ocmip-all'




Ernst,
 	Thanks for the feedback on our routines. You have brought up an 
interesting
 point in that the functions we generated include the full carbon chemistry to 
be totally
 consistent with the measured carbon parameters. There are a number of shortcuts 
one can
 take to simplify these equations, but there are finite errors associated with 
each of
 the simplifications. For example, I gather that you ignore all of the secondary 
species 
 that contribute to total alkalinity except for borate. Borate is by far the 
largest of 
 the secondary species and I think you agree that it needs to be included. As 
for sulfate
 and fluoride, their contribution is very small (<0.1 uatm in calculated fco2) 
and 
 could probably be ignored without a problem. I guess it comes down to a trade 
off 
 between being chemically correct versus the computer time for the 4 extra 
calculations
 (bt, ft, kb and and kf). Phosphate and Silicate are somewhere in the middle. I 
would
 disagree with your statement that the phosphate and silicate are "far below any
 precision of measurements"). Obviously in warm, low latitude surface waters 
where the
 concentrations are near 0 they are not significant, but they do certainly make 
a
 measurable difference in the Southern Ocean. If we take an example from the 
mixed
 layer in the South Pacific with: 
 
 alk=2275, TCO2=2150, Temp=1.5, Sal=34, phosphate=2, silicate=50
 
 With the full chemistry you would calculate an fCO2 of 384 uatm.
 
 If you ignore the phosphate and silicate, fco2=377, a difference of 7 uatm. 
About
 5 uatm of that error results from ignoring phosphate. The remaining 2 uatm 
results
 from the silicate. Since silicate can get as high as 200 umol/kg in the ocean 
that
 contribution can be larger. I can't say whether a 7 uatm error is significant 
for 
 the model calculations. That is a decision you guys need to make. If you would 
like
 us to remove some or all of the secondary species just let me know.
 
 	As for using CO3-- and hydrogen to calculate the chemistry, I am not 
sure
 I follow. Unfortunately, I was not able to completely follow the code you 
included 
 without knowing the definitions for some of the variables you use. With our 
code, 
 I believe you can easily converge on an htotal with only 4 iterations if you 
put
 a little thought into your choice of ph ranges. I am not sure what the 
differences
 are in our approach without a better understanding of your code. One thing to 
 remember is that the dissociation constants are strongly dependent on 
temperature
 and salinity, so the CO3-- can vary widely simply because of changes in 
temperature
 or salinity. The only carbon parameters that are not a function of temperature 
and
 salinity are total CO2 and total alkalinity (in mass units). Note that with the 
 accepted definition, carbonate alkalinity is not independent of Temp and Sal.
 
 	I hope this makes some sense. If you have further questions or would 
like
 us to modify the functions just let me know.
 
 Cheers,
 Chris
- ------------- End Forwarded Message -------------