Changeset 2050 for trunk/LMDZ.TITAN/README

Timestamp:

Nov 30, 2018, 12:55:49 AM (7 years ago)

Author:

jvatant

Message:

Major radiative transfer contribution : Add the 'corrk_recombin' option that allows to use
correlated-k for single species instead of pre-mix and enables more flexiblity for variable species.
-> Algorithm inspired from Lacis and Oinas 1991 and Amundsen et al 2016

+ Added 'recombin_corrk.F90' - Important improvements in 'sugas_corrk.90' and 'callcork.F90'

• Must have the desired variable species as tracers -> TBD : Enable to force composition even if no tracers

• To have decent CPU time recombining is not done on all gridpoints and wavelenghts but we calculate a gasi/v_recomb variable on the reference corrk-k T,P grid (only for T,P values who match the atmospheric conditions ) which is then processed as a standard pre-mix in optci/v routines, but updated every time tracers on the ref P grid have varied > 1%.

READ CAREFULY :

• In case of 'corrk_recombin', the variable L_NREFVAR doesn't have the same meaning as before and doesn't stand for the different mixing ratios but the different species.
• Input corr-k should be found in corrkdir within 'corrk_gcm_IR/VI_XXX.dat' and can contain a 'fixed' specie ( compulsory if you include self-broadening ) that MUST have been created with correct mixing ratios, or a variable specie for which mixing ratio MUST have been set to 1 ( no self-broadening then, assume it's a trace sepecie ) -> You can't neither have CIA of variable species included upstream in the corr-k

File:

• trunk/LMDZ.TITAN/README (modified) (1 diff)

trunk/LMDZ.TITAN/README

@@ -1496 +1496 @@
 == 24/09/18 == JVO
 Fix the outdated value of surface CH4 in the chemistry (0.048 Niemann 2005 -> 0.0565 Niemann 2010)
+
+== 28/11/18 == JVO
+Add a simple seasonal+latitudinal variation model (season_haze.F90) for the vertical mean profile
+of haze. Based on Karkoscha,2016 . Activable through "seashaze" key (default=T).
+
+== 29/11/18 == JVO ( r2050 )
+Major radiative transfer contribution : Add the 'corrk_recombin' option that allows to use
+correlated-k for single species instead of pre-mix and enables more flexiblity for variable species.
+-> Algorithm inspired from Lacis and Oinas 1991 and Amundsen et al 2016
+
++ Added 'recombin_corrk.F90' - Important improvements in 'sugas_corrk.90' and 'callcork.F90'
+
+- Must have the desired variable species as tracers 
+  -> TBD : Enable to force composition even if no tracers
+
+- To have decent CPU time recombining is not done on all gridpoints and wavelenghts but we
+  calculate a gasi/v_recomb variable on the reference corrk-k T,P grid (only for T,P values
+  who match the atmospheric conditions ) which is then processed as a standard pre-mix in
+  optci/v routines, but updated every time tracers on the ref P grid have varied > 1%.
+  
+READ CAREFULY :
+- In case of 'corrk_recombin', the variable L_NREFVAR doesn't have the same meaning as before
+  and doesn't stand for the different mixing ratios but the different species.
+- Input corr-k should be found in corrkdir within 'corrk_gcm_IR/VI_XXX.dat' and can
+  contain a 'fixed' specie ( compulsory if you include self-broadening ) that MUST have
+  been created with correct mixing ratios, or a variable specie for which mixing ratio
+  MUST have been set to 1 ( no self-broadening then, assume it's a trace sepecie )
+  -> You can't neither have CIA of variable species included upstream in the corr-k