Changes between Initial Version and Version 1 of WorkingGroups/ConfigurationManager/2015_td

Timestamp:

2015-10-23T09:11:36+02:00 (9 years ago)

Author:

jpaul

Comment:

--

WorkingGroups/ConfigurationManager/2015_td

@@ +1 @@
+
+=== Working Group, 2015 tasks ===
+
+CONFIGMAN-1 Agree and test SIREN code 
+
+Some members of the working group have been contact to test Siren code.
+Unfortunately it seems they not found the time to test it, and  we are still waiting their feedback.
+
+However some Mercator Ocean user have try to use SIREN code.
+It seems that there is a need to improve readibility of the documentation.
+Moreover they ask for a tutorial or some examples.
+
+Furthermore some bugs fix, and changes have been commited (others will come
+soon).
+
+main changes:
+- add documentation
+- use mpp structure instead of file structure to read input
+- create one module for each interolation method
+- extrapolate all land points (_FillValue) + change on weight function
+- create math module
+- change dichotomy method to compute closest point on ORCA grid
+- improve way to find the offset between fine and coarse grid
+- allow to change unit of a variable (a scale factor could be applied)
+- new format in namelist to get information on additional request
+- new format for variable configuration file
+
+main bugs fix:
+- use function to copy structure instead of overload assignment operator (to avoid memory leak)
+- boundary indices on velocity point
+- correct issue with "disorder" dimension
+- define local structure to avoid mistake with pointer when deleting embedded structure
+- reinitialise detect array for the interpolation of each level
+- wrong array size in intrerp linear
+- check value, before use scale factor and offset for restart dimg (case no0d...)
+- in extrap, force coef indice to 1, for dimension length equal to 1
+
+
+
+CONFIGMAN-2 & 3 BDY configuration tool + Merge BDY configuration tool and SIREN
+
+Following advices of the NEMO developpers comitee, concernong the BDY tool.
+we decided to use the NERC python tool instead of the CMCC Fortran tool.
+Indeed, the NERC tool seems already able to do more things.
+
+Tests should have been done by CMCC on the NERC tool. 
+Unfortunately there were still a few outstanding pieces to address, and we
+missed time to achieve all of it. So tests could not start.
+
+
+