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

CFC paper, ferret scripts

Dear Jean-Claude,

I concur with Ray that you have done a marvelous job in pushing the CFC 
analysis forward and removing many of the hurdles that are associated with 
a project that grand in scale.  We owe you much for taking up this 
leadership and I hope other people within OCMIP will equally step forward 
and move the analysis forward in their field of interest.

The paper has improved impressively since I saw it the first time about 
half a year ago.  I am particularly delighted to see that you added 
vertical column integrals as well as some more crucial data comparisons in 
key areas.  The paper is nearly ready to go, but in my opinion would 
greatly benefit from a few additional analyses.  Ray Najjar provided 
already extensive comments. I can therefore be relatively brief.

Major comment:

As discussed a little bit more than a week ago during the OCMIP meeting, I 
would like to encourage you strongly to compute normalized CFC inventories 
and add this as a third panel in Figures 4, 6, and 12, so that the effect 
of temperature errors in the models can be separated from the effect of 
circulation errors.  I believe this is very important, because it is the 
latter that we are mostly interested in when using CFCS. Knowing that there 
are serious biases in the model simulated temperatures, one always has to 
second-guess how much of the differences are due to this effect and 
how much is due to circulation. Adding this panel removes this ambiguity
and gives a clear answer. 

I attach the ferret scripts that compute the solubility of the
CFCs and also normalized CFC concentrations. I hope you find
them useful.

Minor comments: (in order of appearence in text)

Title: The title is quite complicated. It might help to change 
   the substantives to verbs, i.e.
   Using CFCs to evaluate high-latitude ocean ventilation in the OCMIP-2
Abstract: (to be written)
   What are the critical results that we want to highlight?
   - large spread in global uptake
   - largely due to differences in the ventilation of the high-latitudes 
     in these models --> strong temperature dependence of CFC solubility
	 makes the CFC more sensitive to processes in the high latitudes (in
	 comparison to ant. CO2, which is more sensitive to low latitude
   - can we point to particular processes/parameterizations that cause
     these differences in the models. I believe you make a pretty strong
	 case for the surface buoyancy forcing (sea-ice model) and the
	 GM parameterization. 
Introduction: [p1] and elsewhere
   I suggest to use the word "sink" has few times as possible. You 
   use it most of the time in the sense of a flux, but often one thinks of
   it also in the sense of storage. In the example " [..] in the 
   Southern Ocean, which is the largest sink for anthropogenic CO2", I
   am left with the impression that the Southern Ocean is not only the
   region of uptake, but also the region of storage, which is not correct.
   This ambiguitiy can be avoided by writing: "[..] which is the region
   of largest anthropogenic CO2 uptake."
Introduction: [p2, 1st para] Interhemispheric transport
   Reasoning is unclear to me. While the interhemispheric transport
   problem certainly served as a focus point for the OCMIP simulations
   in general, CFC simulations will shed no light on this issue. I therefore
   found this discussion more confusing than helpful.
   I therefore suggest to remove this part.
Introduction: [p3, 2nd para] "sink"
   Again be careful with the word "sink"
Results: [Figure 1]
   It took me a long time to understand Figure 1c. This was probably 
   partially due to the lack of an x-axis label, and this plot being
   together with the Fig 1a and Fig 1b which have time as an x-axis. 
   I think you can show the same information by plotting the relative
   difference versus time, i.e. (y-y_mean)/y_mean, where y are the time
   evolving CFC inventories of the different models.
CFC Inventories: [p6, 2nd para] "sink"
   Again be careful with the word "sink". Furthermore, you mentioned
   this fact already several times before. Is it necessary to repeat it
   here again?

CFC Inventories: [p7, 1st para] SH vs NH
   Isn't this simply proportional to the ice-free ocean surface area in the
   two hemispheres?
CFC Inventories: [p8, 1st para] Inventories 
   I am extremely surprised 
   about the underestimation of the model simulated CFC inventories in the 
   North Atlantic along OC202.  I am particularly sceptical because I see 
   that PIUB predicts one of the lowest inventories.  When I computed the 
   ant.  CO2 inventories for the Princeton model and the PIUB model and 
   compared them against my ant.  CO2 inventories, I found that the
   PIUB overpredicted the inventory in the northern most latitudes by
   at least a factor of 2! The main reason for this difference was simply
   topography (PIUB has a uniform depth of 4000). Are you 100% sure that
   all the calculations have been done correctly?
Discussion: [p15, last para] uptake, transport, storage
   You mention here for the first time the connection between uptake,
   transport, and storage. I believe it would be very helpful to move
   a short description of this into the introduction, particulary when
   seen in the light of the "sink" problem, I mentioned above. This is
   important because people often get confused about this issue.
Discussion: [p17, end of 1st para] meso-scale eddies
   What is the basis for this conclusion? At least a reference is needed.

I am looking forward to a very nice paper on the CFC results. Thanks for
making it happen!

-- best regards



! GO TOOL for calculating solubility of freons
!                         CFC11, CFC12 and CCl4
!      NG 21. June 2000
!         on the basis of the coefficients used in OCMIP

\! run silently
\cancel mode verify
! description
!  $1 : name of temperature array [deg C]
!  $2 : name of salinity array    [psu]
!  $3 : number of dataset
! check input
\set mode/last verify
\query/ignore $1"< use as: go calc_cfcsolub temp sal dataset" 
\cancel mode verify
! define coefficients
!     Solubility coefficients in in mol/l/atm
!      _f11 for CFC11, (2) for CFC12, (3) for CCl4 
!     after Warner and Weiss (1985) DSR, vol 32 for CFC11 and CFC12
!     after Bullister, J.L. and Wisegavger, D.P. (1998) 
!        The solubility of carbon tetrachloride in water and 
!        seawater, (Deep Sea Res, 45, 1285-1302).
let d1_f11 = -229.9261
let d2_f11 =  319.6552
let d3_f11 =  119.4471
let d4_f11 =   -1.39165
let e1_f11 =   -0.142382
let e2_f11 =    0.091459
let e3_f11 =   -0.0157274
let d1_f12 = -218.0971
let d2_f12 =  298.9702
let d3_f12 =  113.8049
let d4_f12 =   -1.39165
let e1_f12 =   -0.143566
let e2_f12 =    0.091015
let e3_f12 =   -0.0153924
let d1_ccl4 = -148.247  
let d2_ccl4 = 227.758  
let d3_ccl4 = 62.5557   
let d4_ccl4 = 0.       
let e1_ccl4 = -0.400847 
let e2_ccl4 = 0.265218 
let e3_ccl4 = -0.0446424

!     calculate solubilities
!       use Warner and Weiss (1985) DSR, vol 32, final result
!       in mol/m3/pptv (1 part per trillion 1e-12)
!       use Bullister and Wisegavger for CCl4
define variable/d=$3 tk = ($1 + 273.15)*0.01

define variable/d=$3 alpha_f11 = exp(d1_f11+d2_f11/tk+d3_f11*ln(tk)+d4_f11*tk^2+$2*((e3_f11*tk+e2_f11)*tk+e1_f11)) * 1.0e-9
define variable/d=$3 alpha_f12 = exp(d1_f12+d2_f12/tk+d3_f12*ln(tk)+d4_f12*tk^2+$2*((e3_f12*tk+e2_f12)*tk+e1_f12)) * 1.0e-9
define variable/d=$3 alpha_ccl4 = exp(d1_ccl4+d2_ccl4/tk+d3_ccl4*ln(tk)+d4_ccl4*tk^2+$2*((e3_ccl4*tk+e2_ccl4)*tk+e1_ccl4)) * 1.0e-9 

! restore verify
set mode verify
! output: arrays alpha_f11, alpha_f12, and alpha_ccl4 ready to use
 ! Program FERRET (V500beta1.1)
 ! Version 5.00 - 06/03/99
 ! 23. June 2000
 ! compute normalized CFC concentrations
 ! Nicolas Gruber 
cancel mode verify
! ----------------------------------------------------------------------
! get data
! ----------------------------------------------------------------------
use "your data file"
! compute CFC solubilties
! ----------------------------------------------------------------------
go calc_cfcsolub temp[d=1] salinity[d=1] 1
! ----------------------------------------------------------------------
! define normalized quantities
! ----------------------------------------------------------------------

let alpha_f11_mean = 1.207E-11  ! sfc mean
let alpha_ccl4_mean = 4.060E-11 ! sfc mean

define var/title="norm. CFC-11"/unit="pmol/l"/d=7 cfc11_n = cfc11/alpha_f11*alpha_f11_mean*1e15
define var/title="norm. CCl4-stable"/unit="pmol/l"/d=7 ccl4_stable_n = ccl4_stable/alpha_ccl4*alpha_ccl4_mean*1e15
define var/title="norm. CCl4-decay"/unit="pmol/l"/d=7 ccl4_decay_n = ccl4_decay/alpha_ccl4*alpha_ccl4_mean*1e15
define var/title="pCFC-11"/unit="patm"/d=7 pcfc11 = cfc11/alpha_f11*1e6
define var/title="pCFC-12"/unit="patm"/d=7 pcfc12 = cfc12/alpha_f12*1e6
define var/title="pCCl4-stable"/unit="patm"/d=7 pccl4_stable = ccl4_stable/alpha_ccl4*1e6
define var/title="pCCl4-decay"/unit="patm"/d=7 pccl4_decay = ccl4_decay/alpha_ccl4*1e6