Changes between Version 7 and Version 8 of ticket/0972_merge_PISLOB

Timestamp:

2012-08-06T14:45:19+02:00 (12 years ago)

Author:

cetlod

Comment:

--

ticket/0972_merge_PISLOB

@@ -31 +31 @@
 '''(2)''' Improvment of PISCES ( Done by Olivier Aumont )
 
-I spent some time on the code the last few days to check several things as well as to debug Kriest. I finally found some motivation to work on that "unmaintained" part of the code, which was a shame. I should say that the Kriest code was not so buggy and so the debugging was pretty short. The problem was more that the results I got with this version were more or less crap ... So I did some tuning, which is by no means over. During this debuging/tuning procedure, I found some problems in the standard code which explains why there are not only changes concerning Kriest but also some other changes. I should say that these changes include also some bugs/inappropriate codings I changed some months ago.
+* debbuging/tuning procedure
+Kriest code was not so buggy and so the debugging was pretty short. The problem was more that the results I got with this version were more or less crap ... So I did some tuning, which is by no means over. During this debuging/tuning procedure, I found some problems in the standard code which explains why there are not only changes concerning Kriest but also some other changes. I should say that these changes include also some bugs/inappropriate codings I changed some months ago.
 
-So, in the attached files, you will find a list of all the routines that have been modified. No major parameterizations have been changed. It's only debugging/slight changes. After the mercator meeting, I will send another package with real changes in the parameterization :
-1) denitrification in the sediments
-2) iron from seaice
-3) Al's iron model following their paper his paper with Christof (Al can you send me your code if possible ?)
-4) The new river package
-5) Potentially zoo in DEB (should discuss with THomas about that. Ok he is not that far away from me !).
-....
+Here is a list of all the routines that have been modified. No major parameterizations have been changed. It's only debugging/slight changes :
+ - denitrification in the sediments
+ - iron from seaice
+ - Al's iron model following their paper his paper with Christof (Al can you send me your code if possible ?)
+ - The new river package
 
-That's it for now. Of course, Christian don't hesitate to call me if you need more explanations or if you want to talk about the timing, etc ...
 
-Olivier
+List of the changes/bugfixes (alphabetical order) :
 
-List of the changes/bugfixes (alphabetical order)
-
-p4zbio.F90 : correction made on the update of trn to avoid marginal negative concentrations which happened with the previous coding.
-p4zlim.F90 : slight change on the minimum iron concentration
-                      small changes on the half-saturation constants for N limitation (x3 instead of x2)
-                      Bug correction on xlimbac. zlim1 should be limitation by N and not only by NH4 (bacteria can use                       both nh4 and NO3).
-p4zmicro : the multiplication by 2 in zrespz has been removed
-                    Changes in the threshold computation for grazing. Grazing is now possible even for very small                             values of food which avoids loosing microzooplankton when food is very scarce.
-p4zmeso (ok I know, it is before p4zmicro but it was easier for the comments ) : same change as in                                 p4zmicro for the threshold.
-p4zprod.F90 : various bug corrections (nothing serious).
-p4zrem.F90 : zbactfer has been completely rewritten
-p4zsed.F90 : P deposition has been added and is computed from dust.
-                        7.6 has been changed to 7.625
-                        The computation of the total supply by the different sources has been corrected to compute                             the real average. So the sums should be divided by the number of timesteps of the input file.                             There was no problem for the standard version but when the number of timesteps are changed                          in the input files (for instance monthly fields for rivers), then the budget is wrong !
-p4zsink.F90 : bug corrections for Kriest,
-                        I have added a new xkr_nmicro for the number of elementary particles in micro end products for the kriest parameterization. This means that the namelist has changed.
-                        doc aggregation for Kriest has been changed
-                        Computation of the standard sinking speed has been changed. For this point, in fact, the correct parameterization is commented out in the standard code. I changed that to uncomment the appropriate lines.
+ - p4zbio.F90 : correction made on the update of trn to avoid marginal negative concentrations which happened with the previous coding.
+ - p4zlim.F90 : slight change on the minimum iron concentration small changes on the half-saturation constants for N limitation (x3 instead of x2). Bug correction on xlimbac. zlim1 should be limitation by N and not only by NH4 (bacteria can use both nh4 and NO3).
+ - p4zmicro : the multiplication by 2 in zrespz has been removed. Changes in the threshold computation for grazing. Grazing is now possible even for very small values of food which avoids loosing microzooplankton when food is very scarce.
+ - p4zmeso (ok I know, it is before p4zmicro but it was easier for the comments ) : same change as in p4zmicro for the threshold.
+ - p4zprod.F90 : various bug corrections (nothing serious).
+ - p4zrem.F90 : zbactfer has been completely rewritten
+ - p4zsed.F90 : P deposition has been added and is computed from dust.
+                7.6 has been changed to 7.625 ( rno3 parameter )
+                The computation of the total supply by the different sources has been corrected to compute the real average. So the sums should be divided by the number of timesteps of the input file. There was no problem for the standard version but when the number of timesteps are changed in the input files (for instance monthly fields for rivers), then the budget is wrong !
+ - p4zsink.F90 : bug corrections for Kriest, I have added a new xkr_nmicro for the number of elementary particles in micro end products for the kriest parameterization. This means that the namelist has changed. doc aggregation for Kriest has been changed Computation of the standard sinking speed has been changed. For this point, in fact, the correct parameterization is commented out in the standard code. I changed that to uncomment the appropriate lines.
 
 trcnam_pisces, trcsms_pisces, ... : 7.6 -> 7.625 and perhaps some corrections. Furthermore, trcnam_pisces.F90 includes a new parameter that should be read in the namelist which is the N content of the aggregate.
@@ -73 +63 @@
 4) Refonte du time splitting scheme pour la sédimentation verticale.
 5) Ajout d'une solubilité variable pour le Fer controlée par un booleen dans la namelist.
-6) Ajoût d'une variable dust qui représente la concentration de dust dans l'océan. Ca sert pour le scavengign du fer.
+6) Ajoût d'une variable dust qui représente la concentration de dust dans l'océan. Ca sert pour le scavenging du fer.
 7) retouche de quelques paramètres du modèle
 8) refonte de la variation des constantes de demi-saturation avec refonte de la namelist correspondante.