Opened 4 years ago

Last modified 16 months ago

#411 accepted defect

Problem with cim

Reported by: maignan Owned by: maignan
Priority: major Milestone: ORCHIDEE 3.0
Component: Biogeochemical processes Version:
Keywords: Cc:


Initial mail from S. Denvil reporting the problem

il y a un plantage orchidee dans IPSLCM6.0.13 qui arrive année 7 et
année 10.

Le souci vient de cette variable. On la sortait déjà avant. Le problème
est donc nouveau, sûrement lié au changement de PFT. Je vais désactiver
la sortie de cette variable et je vais relancer.

field_def_orchidee.xml:    <field id="cim" name="cim"
long_name="Intercellular CO2 over nlai" unit="ppm" grid_ref="grid_nvm" />

CM6013-pi-01_17070101_17071231_out_execution_error:with errorcode -1.
CM6013-pi-01_17070101_17071231_out_xios.x.err.431:> Error
[CNc4DataOutput::writeFieldData_ (CField*  field)] : In file
line 2445 -> On writing field data: cim
CM6013-pi-01_17070101_17071231_out_xios.x.err.431:Error when calling
function ncPutVaraType(ncid, varId, start, count, data)

oasis_init_comp: Calling MPI_Init
> Error [CNc4DataOutput::writeFieldData_ (CField*  field)] : In file
line 2445 -> On writing field data: cim
In the context : orchidee_server
Error when calling function ncPutVaraType(ncid, varId, start, count, data)
NetCDF: Numeric conversion not representable
Unable to write data given the location id: 65536 and the variable whose
id: 172 and name: cim
  • A first (non directly related) problem has been identified in diffuco_trans_co2. Under:
IF ( ABS( assimi(icinic,jl) + Rd(icinic,jl) ) .LT. min_sechiba ) THEN

we should have for both C3 and C4:

   leaf_ci(icinic,jv,jl) = Ca(icinic)

The comment:

!leaf_ci keeps its initial value (Ca).

was incorrect because leaf_ci was initialized only for pixels doing no carbon assimilation.

  • The very high values of cim possibly causing the problem reported by S. Denvil occur for C4 plants. cim is high because leaf_ci is high, leaf_ci is high because Cc is high, these high Cc values seem to be associated with 0 values of b.

Change History (6)

comment:1 Changed 4 years ago by maignan

  • Milestone set to ORCHIDEE 2.0
  • Owner changed from somebody to maignan
  • Status changed from new to assigned

While we're looking for a scientific correction, the variable cim will be disabled for version 2.

comment:2 Changed 4 years ago by jgipsl

Variable cim is temporary desactivated [4769]

comment:3 Changed 4 years ago by maignan

  • Milestone changed from ORCHIDEE 2.0 to ORCHIDEE 3.0
  • Status changed from assigned to accepted

comment:4 Changed 17 months ago by jgipsl

Following was said during the ORCH meeting today:

  • Fabienne asked Patricia per mail to see if she tested in coupled mode (see minutes from meeting 04/06/2019).
  • FM, NVuich, NVio, PP, Matt(or SL) will meet to discuss further. Same code is used in CN-CAN so it would be good to do this before merging CN-CAN.

comment:5 Changed 16 months ago by jgipsl

Said today during ORCH meeting:

Patricia made test in coupled mode but with a more recent version of the code, the model did not crash. No need to make further tests.

NViovy will look at the issue tomorrow and PP thursday this week.

comment:6 Changed 16 months ago by nicolasviovy

looking to the end of the comment, it indicate that cim can reach a high value because leaf_ci and cc is high which can be problematic and should be solved.
A simple solution is to define leaf_ci=MIN(leaf_ci,ccanopy) after calculation of leaf_ci. Indded if we can imagine that in rare case leaf_ci could be slightly higher than Ca it should not be very much higher and only in case where there is no assimilation. so limit leaf_ci to Ca seems reasonable

Note: See TracTickets for help on using tickets.