Opened 8 years ago

Closed 7 years ago

#229 closed defect (fixed)

Bug with explicit snow scheme after inclusion of implicit coupling

Reported by: jgipsl Owned by: cottle
Priority: critical Milestone: IPSLCM6.v1
Component: Physical processes Version: trunc
Keywords: Cc:

Description

There is a bug in the trunk introduced in revision [3059] when activating the explicit snow scheme. The snow mass is increasing to much and it seems like the snow melt is insufficient.

Attachments (2)

ncview.snowmelt.3269.png (39.9 KB) - added by maignan 8 years ago.
ncview.snowmelt.3059.png (26.5 KB) - added by maignan 8 years ago.

Download all attachments as: .zip

Change History (7)

comment:1 Changed 8 years ago by maignan

This bug has been corrected in revision [3269]. The important facts are:

  • The correct computation of snowmelt requires that the computation of the snow temperature profile is done before in explicitsnow (as in revision [3051].
  • Snow coefficients are still computed before the soil coefficients in thermosoil for the implicit scheme (as in revision [3059]).
  • enerbil_fusion is called after thermosoil_main for a correct setting of temp_sol_new.

Validation was done at global scale 2-degree. See snowmelt in the attached files after 19 years for revisions 3059 and 3269.

Last edited 8 years ago by maignan (previous) (diff)

Changed 8 years ago by maignan

Changed 8 years ago by maignan

comment:2 Changed 8 years ago by maignan

As noted by Josefine, results are modified even if explicit snow is not activated. This is due to the displacement of enerbil_fusion, which modifies temp_sol_new for both snow schemes.
This modification needs maybe to be done at different locations depending on the snow scheme. The ticket stays open till this question has been answered.
Also this revision solves the snowmelt problem in the explicit snow scheme, but there are still on-going discussions regarding its implicit status.

comment:4 Changed 8 years ago by jgipsl

Done in the trunk in commit [3313] :

  • ok_explicitsnow=y : moved the part from enerbil_fusion into the end of thermosoil_main/thermosoilc_main. This do not change the order of calculations, the results should be the same as before this commit.
  • ok_explicitsnow=n : change back as it was done before commit 3269 : call enerbil_fusion is done after hydrol_main. The resultats are now the same as before 3269 for this case.

comment:5 Changed 7 years ago by maignan

  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.