Changeset 2297 for trunk/LMDZ.GENERIC/libf/phystd/radcommon_h.F90

Timestamp:

Apr 26, 2020, 7:45:46 PM (5 years ago)

Author:

jvatant

Message:

Add a generic n-layer aerosol scheme to replace the former buggy 2-layer scheme as well as the hard-coded NH3 cloud.

It can be called using 'aeronlay=.true.' in callphys.def, and set the number of layers (up to 4) with 'nlayaero'.
Then, the following parameters are read as arrays of size nlayaero in callphys.def (separated by blank space)

*aeronlay_tauref (Optical depth of aerosol layer at ref wavelenght)
*aeronlay_lamref (Ref wavelenght (m))
*aeronlay_choice (Choice for vertical profile - 1:tau follows atm scale height btwn top and bottom - 2:tau follows it own scale height)
*aeronlay_pbot (Bottom pressure (Pa))
*aeronlay_ptop (Top pressure (Pa) - useful only if choice=1)
*aeronlay_sclhght (Ratio of aerosol layer scale height / atmospheric scale height - useful only if choice=2 )
*aeronlay_size (Particle size (m))
*optprop_aeronlay_vis File for VIS opt properties.
*optprop_aeronlay_ir File for IR opt properties.

+Extra info :

+ In addition of solving the bug from 2-layer it enables different optical properties.
+ The former scheme are left for retrocompatibility (for now) but you should use the new one.
+ See aeropacity.F90 for the calculations

+ Each layer can have different optical properties, size of particle ...
+ You have different choices for vertical profile of the aerosol layers :

• aeronlay_choice = 1 : Layer tau is spread between ptop and pbot following atm scale height.
• aeronlay_choice = 2 : Layer tau follows its own scale height above cloud deck (pbot).

In this case ptop is dummy and sclhght gives the ratio H_cl/H_atm.

+ The reference wavelenght for input optical depth is now read as input (aeronlay_lamref)
+ Following the last point some comment is added in suaer_corrk about the 'not-really-dummy'ness of IR lamref..

File:

• trunk/LMDZ.GENERIC/libf/phystd/radcommon_h.F90 (modified) (1 diff)

trunk/LMDZ.GENERIC/libf/phystd/radcommon_h.F90

@@ -57 +57 @@
 ! 
 ! For the "naerkind" kind of aerosol radiative properties : 
-! QIRsQREF :  Qext / Qext("longrefvis")
+! QIRsQREF :  Qext / Qext("longrefir")
 ! omegaIR  :  mean single scattering albedo
 ! gIR      :  mean assymetry factor
+!
+!
+! Note - QIRsQREF in the martian model was scaled to longrefvis,
+! here it is scaled to longrefir, which is actually a dummy parameter,
+! as we do not output scaled aerosol opacity ...
+!
 
       REAL*8 BWNI(L_NSPECTI+1), WNOI(L_NSPECTI), DWNI(L_NSPECTI), WAVEI(L_NSPECTI) !BWNI read by master in setspi