Opened 8 months ago

Closed 8 months ago

#837 closed task (fixed)

Check sapiens_forestry.f90

Reported by: luyssaert Owned by: somebody
Priority: minor Milestone: ORCHIDEE 4.3
Component: Anthropogenic processes Version: trunc
Keywords: Cc:

Description

When working on the unmanaged forest (r7549, r7550, r7551) the following was noted:

  • in sapiens_forestry.f90, the parameter n_self_thin is used to decide whether the forest will be thinned or not. It is not clear what happens to the forest if this we are still above n_self_thin. It is possible that nothing happens which would explain why the calculation of st_dens in stomate_mark_kill.f90 for managed forests has so much impact. For managed forests st_dens should only deal with overstocking and given that sapiens_forestry is called every year, overstocking should be limited. If the model does not thin until n_self_thin is reached, there would be considerable overstocking resulting a reduction of the height growth (as observed in tests with r7549 when running without this overstocking line of code).

Test: comment the overstocking code and comment the n_self_thin criterion in sapiens_forestry.f90.

Change History (1)

comment:1 Changed 8 months ago by luyssaert

  • Resolution set to fixed
  • Status changed from new to closed

Comparison of r7551 with overstocking (in stomate_mark_kill.f90) and n_self_thin (in sapiens_forestry.f90) to the same version without these two blocks of code resulted in an overall slightly faster height growth (0.5 to 2 m over 5 years) for the model version without this code. Given that this code is a bit inconsistent in the sence that ifm_thin is dealt with in stomate_mark_kill.f90 which was reserved for ifm_none and that this code requires an additional empirical parameter (i.e., n_self_thin), it was decided to simplify the code and remove both blocks. This changes has been made in r7555.

Note: See TracTickets for help on using tickets.