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

Re: ICBP-Newsletter misunderstandings about the AWI-model

Dear Marie France,

Thanks for the clarifying things about what was said about the AWI
model in the recent OCMIP-2 article in the GAIM Newsletter.

Please do keep in mind though that objective of that short article was
only to send a signal to scientists and program managers in IGBP as
well as other international organizations that OCMIP is alive and
well.  There was no intent to provide a scientific article.

In any case, after reading your response I still believe that the two
lines of text which you quote contain NO untruth, contrary to what you
imply.  They do lack detail, however.  Perhaps it would have been
better to add an intermediate sentence such as "The AWI model then uses
that derived circulation (calculated without using CFC data) in a
forward manner to compute oceanic distributions of CFC-11 and CFC-12."
In defense of the quoted text though, my thinking was that most
modelers would be able to figure out that missing sentence for
themselves, whereas many others would become lost and would skip to
the next sentence as soon as they saw the word "adjoint".  

In short, writing for a wide audience has its drawbacks. Next time
around, I'll be sure to ask for early comments from all those
concerned.  Such would have been done for this article as well, had
the submission deadline not been so urgent.

Best regards,

- -- Jim

On Tue, 31 Aug 1999, Marie France Weirig wrote:

> Dear all,
> in the latest IGBP GAIM newsletter, James and Jean-Claude presented
> some OCMIP-CFC results. There, it was written that

> "The AWI-model is an adjoint model, which uses hydrographic and
> tracer data in an inverse fashion to compute a mean annual ocean
> circulation.  The other models are of the more forward type"

> This is not true.  To prevent further misunderstandings and to
> clear up things I will explain briefly what our model really is
> doing.

> First of all, we compute CFC concentrations like anybody else in a
> "forward way."  The model we use to perform the OCMIP-simulations is
> neither doing any "adjoint" nor "inverse" calculations.  We
> calculate OCMIP-tracers with a stationary current field which was
> originally established using the adjoint technique.  Data of
> temperature, salinity, alkalinity, PO4, CO2, O2, SiO2, and NO3 were
> used to determine the mean, annual circulation.
> NO OTHER TRACER DATA were involved, which means that the velocity field
> has never been adjusted in order to reproduce CFC data.
> This current field is NOT changed during OCMIP simulations.
> The CFC-simulations are calculated strictly forward using the flow
> field calculated from INDEPENDENT measurements (and so any future
> OCMIP experiments will be)!  The very good results could not be
> expected because there is absolutely no tuning towards measured
> CFC-distributions.  I hope things are a bit clearer now.  Best
> regards Marie -France 

   James Orr                       Laboratoire des Sciences du Climat 
   LSCE, CEA Saclay                             et de l'Environnement    
   Bat. 709 - Orme                 Unite Mixte de Recherche CEA-CNRS    
   F-91191 Gif-sur-Yvette Cedex    Phone:   (33) (0)1 69 08 77 23      
   FRANCE                          Fax:     (33) (0)1 69 08 77 16
                                             |    |     
   mailto:  orr@cea.fr            (country code)  (when in France)