**************************************************************************************************
V200511B wrt. V200511A
**************************************************************************************************

1 - Scripts
===========

$tmpdir is removed by default

Files : chimere-mm5


2 - make.log was removed
========================

make.log was removed when $tmpdir was removed. Now make.log becomes make.$$.log
and remains in /tmp directory.

Files : script_functions/chimere-mm5.exec.sh
        meteo/interf-mm5


3 - Pointer type in udunits
===========================

PTR is now of integer(kind=8) type. Allows udunits to work in 64 bits.

Files : toolbox/exachim/udunits.f90
        toolbox/exachim/mkdiff.f90
        toolbox/exachim/chm2coards.f90

4 - udunits compilers
=====================

It is not necessary to compile udunits with the same compiler as CHIMERE.

Files : makefiles/Makefile.hdr.*
        HOWTO-INSTALL-UDUNITS

5 - diagmet segfaults in debug mode
===================================

The array bounds test in debug mode fails under certain conditions.
correction : a complex "if ... thne" is splitted in two nested "if ... then"

Files : src/diagmet_science.f90


6 - outprint.o listed twice in Makefile
=======================================

Files : src/Makefile

7 - default mode = PROD
=======================

Files : makefiles.hdr/*


8 - "status" variable undeclared in subs.f90
============================================

Added declaration in #ifdef G95

Files : src/subs.f90

9 - g95 supports -fendian
=========================

Add -fendian in g95 compiler flags.

Files : makefiles.hdr/Makefile.hdr.g95*

10 - ifort bug with FC4
=======================

Added -no-ipo in ifort complier flags to allow for compiling with FC4 and 
recent versions of ifort.

Files : HOWTO-INSTALL-NETCDF
        makefiles.hdr/Makefile.hdr.ifort*


11 - mouse buttons reverted in README.EXACHIM
=============================================

Files : toolbox/exachim/README.EXACHIM


12 - exadiff.sh in toolbox
==========================

A script and a fortran routine are added in toolbox/exachim, to display file differences.

Files : toolbox/README.TOOLBOX
        toolbox/exachim/mkdiff.f90
        toolbox/exachim/exadiff.sh

        
**************************************************************************************************
V200511A wrt. V200510B
**************************************************************************************************

1 - Scripts
===========

Modification of scripts logic : The main script "chimere-mm5" contains
almost only parameter declarations, while execution is done by another script which is sourced
by the main script.

Files : chimere-mm5
	script_functions/calc_dates.sh
	script_functions/chimere-mm5.exec.sh


**************************************************************************************************
V200510B wrt. V200510A
**************************************************************************************************

9 - prepexmm5.f90
=================

When trying to correct bug #14, created a new bug which made precipitations
wrong for the first record.

Correction : follow exactly the logic of the "binary" version

Files : meteo/prepexmm5.f90

8 - toolbox
===========

In exachim.gs and examout.gs, define the auto-generated palette from the
last scene of the series.

Add a standard palette for "airm" in examout.gs

Files : toolbox/exachim.gs, toolbox/examout.gs


7 - Add version information
===========================

The model version which produced the output CHIMERE netCDF files is
now stored as an attribute in those file headers.

Files : chimere-mm5.exec, chimere_common.f90, iniread.f90, iniout.f90


6 - shorten findspec
====================

Add and "exit" statement when species found to shorten the do loop

Files : findspec.f90



5 - Species list when generating boundary conditions
====================================================

In previous versions, the species list was held separately from the concentration file.
Now the species list is included as a variable in the netCDF boundary files.
During nested runs, CHIMERE reads as many species as it can in the boundary condition
files. Species which are not present in boundary conditions files are grown from zero
by the model.

Files : chimere-mm5.exec, chimere-mm5-<*>.conf, iniboun.f90, preplat_chimere.f90,
        preplat_gocart.f90, preptop_chimere.f90, preptop_gocart.f90


4 - length of species string
=============================

The length of a species name string is now defined by the parameter
"splen" in chimere_consts.f90

Files : chimere_common.f90, iniboun.f90, preplat_chimere.f90, preptop_chimere.f90
        preplat_gocart.f90, preptop_gocart.f90 Makefile


3 - Buglet in HOWTO-INSTALL-UDUNITS
===================================

Discrepancy between udunits.tar.Z and  udunits.tar.gz

Correction : udunits.tar.gz becomes udunits.tar.Z

Files : HOWTO-INSTALL-UDUNITS



2 - Bug report (R. Vautard)
==============

A bug in interpolation routine wcal1.f90 makes location of nearest 
neighbours impossible when the points of the two grids coincide exactly.

Correction : Add a test for equality and derive the neighbours for
this case.

Files : meteo/wcal1.f90



1 - Bug report (R. Vautard)
==============

A bug in prepexmm5.f90 precluded the reading of the first record of
the MM5 file.

Correction : Read all MM5 time slots between dates "ids" and "ide"

Files : meteo/prepexmm5.f90



**************************************************************************************************
V200507A wrt. V200501Hjlm-1 
**************************************************************************************************

3 - pgf90 version 4 cannot compile prepexmm5.f90
================================================
Some f90 compilers are not aware of the NULL() function which is now
standard in f95.
Correction : delete line "xin => NULL()". It was in fact a no-op.

Files : prepexmm5.f90

2 - pgf90 cannot compile chimere2gmt.f90 
=========================================
Correction : Add a "#ifdef PGI" for use of iargc()
Declare iargc()

Files : chimere2gmt.f90

1 - Bug in ncmapit.sh
=====================
Error in arguments passed to chimere2gmt.e
Correction : replace $$gmt by $5

Files : ncmapit.sh


**************************************************************************************************
V200501Hjlm-1 wrt. V200501Hjlm 
**************************************************************************************************

6 - out file par.XXXXX does not change with time
=================================================
time dimension forgotten in scientific variables of par. file.
Correction : add time dimension to scientific variables in iniparam.f90

Files : iniparam.f90

5 - domain check error under g95 in concat.f90
==============================================
nf90_get_att() returns domain+garbage, thus breaking the
domain consistency test.
Correction : cut string at the useful length.

Files : concat.f90

4 - missing "''" in interf-mm5
===============================
Badly generated prepexmm5.nml
Correction : enclose ${chimere_dom} with "''"

Files : interf-mm5

3 - clean-up update_history routine in subs.f90
===============================================
When (history==""), update_history segfaults under pgf90.
Correction : upper bound of history >= 1 .

Files : subs.f90

2 - rename local variables in iniconc.f90
=========================================
idate and ibox renamed to iidate and iibox to avoid conflict with
global variables

Files : iniconc.f90

1 - Buglets in CHIMERE scripts
==============================
* chimere-mm5.exec :
    - rewrite test of existence of "gmake".
    - check existence of aerosols data base only if required.
    - add test for existence of MM5 file.
    - correct error message related to "make.log"
    - run meteo interface in ${chimere_tmp}/meteo

Files : chimere-mm5.exec


**************************************************************************************************
V200501Hjlm wrt. V200501Djlm 
**************************************************************************************************

4 - New netCDF training example readcdf.f90
===========================================
toolbox/ now includes a demonstration program to help reading CHIMERE
netCDF files in a fortran90 code

Files : readcdf/, readcdf/Makefile, readcdf/readcdf.f90, readcdf/README

3 - Backport of diffs from V200501D to V200501G
===============================================

Includes cosmetic diffs, and bioemissions :
file diagbio_science.f90
    nemisb has nominal value=11 (was 10)
    bioem(11) created (Aerosol marine water emission)
file iniemis.f90
    modif in reading emiprof from file fnprim
file chimere-mm5.exec
    modifs to accomodate new directory structure in data/

Files : diagbio_science.f90, iniemis.f90, chimere-mm5.exec


2 - Bug report (E. Minguzzi)
==============
From Bug report 200501G-b10.
In gridloc.f90 and wcal1.f90, a division by zero occurs systematically when
the coarse grid is rectilinear.

Correction : This bug has no consequence if the runtime Floating Point Exception handler 
lets the program run after a divide-by-zero exception. 
This is generally the case. However it is more elegant and portable to make
a test before the division.

Files :
gridloc.f90, wcal1.f90


1 - Bug report (E. Minguzzi)
==============
From Bug report 200501G-b11.
An out-of-bounds error may occur on arrays boundlat and boundtop if boundary
concentrations are initalized with a species not found in ACTIVE_SPECIES
but found in FAMILIES.

Correction : This condition should never happen if the database is consistent.
However we include a test with an exit condition if it occurs.

Files :
iniboun.f90

**************************************************************************************************
V200501Djlm : initial test release of the netCDF version
**************************************************************************************************






**************************************************************************************************
V200501Hjlm wrt. V200501Djlm 
**************************************************************************************************

4 - New netCDF training example readcdf.f90
===========================================
toolbox/ now includes a demonstration program to help reading CHIMERE
netCDF files in a fortran90 code


3 - Backport of diffs from V200501D to V200501G
===============================================

Includes cosmetic diffs, and bioemissions :
file diagbio_science.f90
    nemisb has nominal value=11 (was 10)
    bioem(11) created (Aerosol marine water emission)
file iniemis.f90
    modif in reading emiprof from file fnprim
file chimere-mm5.exec
    modifs to accomodate new directory structure in data/

Files
=====
diagbio_science.f90, iniemis.f90, chimere-mm5.exec


2 - Bug report (E. Minguzzi)
==============
From Bug report 200501G-b10.
In gridloc.f90 and wcal1.f90, a division by zero occurs systematically when
the coarse grid is rectilinear.

Correction
==========
This bug has no consequence if the runtime Floating Point Exception handler 
lets the program run after a divide-by-zero exception. 
This is generally the case. However it is more elegant and portable to make
a test before the division.

Files
=====
gridloc.f90, wcal1.f90


1 - Bug report (E. Minguzzi)
==============
From Bug report 200501G-b11.
An out-of-bounds error may occur on arrays boundlat and boundtop if boundary
concentrations are initalized with a species not found in ACTIVE_SPECIES
but found in FAMILIES.

Correction
==========
This condition should never happen if the database is consistent.
However we include a test with an exit condition if it occurs.

Files
=====
iniboun.f90

**************************************************************************************************
V200501Djlm : initial test release of the netCDF version
**************************************************************************************************
