Changeset 6162 for LMDZ6/trunk/libf/phylmd/mo_simple_plumes.f90

Timestamp:

Apr 8, 2026, 7:26:15 PM (3 days ago)

Author:

asima

Message:

For aerosols Simple Plume :

• in mo_simple_plume, changing formulas for lfactor, ssa and asy to exactly correspond to equations 11 to 13 in Stevens et al. 2017.

Main motivation : the ssa expression is simpler.
Changes (slightly) the results.

• minor : removing an useless "i" variable in macv2sp, to avoid a warning at compilation.

File:

• LMDZ6/trunk/libf/phylmd/mo_simple_plumes.f90 (modified) (3 diffs)

LMDZ6/trunk/libf/phylmd/mo_simple_plumes.f90

@@ -579 +579 @@
     ENDDO
 
-    ! lfactor, eq(12) de Stevens et al 2017 inversed (LAMBDA=max(1,lambda/700. therein) 
-    lfactor   = MIN(1.0,700.0/lambda)
+    ! lfactor cf Stevens et al 2017, eq (12) 
+    lfactor = MAX(1.0,lambda/700.)
 
     ! sum contribution from plumes to construct composite profiles of aerosol optical properties
@@ -587 +587 @@
       !
       ! calculate wavelength-dependent scattering properties
-      !  Stevens et al 2017, eqs (11)&(13)
-!    NOTES ASima : ssa and asy only depend on iplume (via ssa550, asy550) and lfactor(lambda), don't need icol dimension.
-!      Also, why eq(11) was written in this more complex form than in the paper ?
-      !ssa(icol) = (ssa550(iplume) * lfactor**4) / ((ssa550(iplume) * lfactor**4) + ((1-ssa550(iplume)) * lfactor))
-      ssa = (ssa550(iplume) * lfactor**4) / ((ssa550(iplume) * lfactor**4) + ((1-ssa550(iplume)) * lfactor))
-      !ssa = ssa550(iplume) / (ssa550(iplume) + (1-ssa550(iplume)) / lfactor**3) 
-      asy = asy550(iplume) * SQRT(lfactor)
+      ! (ASima) Stevens et al 2017, eqs (11)&(13) :
+      !         ssa and asy only depend on iplume (via ssa550, asy550) and lfactor(lambda)
+      ssa = ssa550(iplume) / (ssa550(iplume) + (1-ssa550(iplume)) * lfactor**3) 
+      asy = asy550(iplume) / SQRT(lfactor)
       !
       ! distribute plume optical properties across its vertical profile weighting by optical depth and scaling for
@@ -602 +599 @@
       lextinct = EXP(-angstrom(iplume) * LOG(lambda/550.0))
 
-! NOTE : lextinct, ssa, asy ne dependent ni de icol, ni de k ; peut-on ptimiser ?
+! NOTE : lextinct, ssa, asy ne dependent ni de icol, ni de k ; peut-on optimiser ?
       DO k=1,nlevels
         DO icol = 1,ncol
-           aod_lmdz          = aod550(icol,k,iplume)     * lextinct
+           aod_lmdz         = aod550(icol,k,iplume)       * lextinct
            asy_prof(icol,k) = asy_prof(icol,k) + aod_lmdz * ssa  * asy
            ssa_prof(icol,k) = ssa_prof(icol,k) + aod_lmdz * ssa