Opened 5 months ago

Last modified 5 weeks ago

#2631 accepted task

JULES medlyn stomata simulation

Reported by: pmcguire Owned by: pmcguire
Priority: normal Component: JULES
Keywords: JULES medlyn stomata Cc: p.l.vidale@…
Platform: Other UM Version:


Can you get the JULES 5.2 medlyn stomata branch (from Doug Clark, CEH) working on JASMIN?
And compare the global sims to JULES 4.9 trunk and JULES 5.2 trunk?

Change History (5)

comment:1 Changed 5 months ago by pmcguire

  • Owner changed from um_support to pmcguire
  • Status changed from new to accepted

comment:2 Changed 5 months ago by pmcguire

We have updated and finished a global run with JULES 5.2 trunk on CEDA JASMIN.
The JULES 5.2 medlyn stomata branch is currently running on CEDA JASMIN and will be finished in a few days.
Then we will do an ILAMB model-benchmark benchmarking of the 3 models (JULES4.9, JULES5.2, and JULES5.2medlyn) over the whole globe, and some model-model comparisons. Attempts have been made to do the model benchmarking with ILAMB for only the UK region, but this is still work in progress.

comment:3 Changed 5 months ago by pmcguire

Omar Muller asked to have TRIP added to the Medlyn suite u-ba422. This is now done and being tested.

I determined that the default values for the g1 parameter for the different PFTs were set to 0.
I therefore reran the UK region Medlyn sims on CEDA JASMIN with g1 set to 5*0.5. And I repeated this run with g1 set to 5*10.

The actual values for the different values have been obtained from Doug Clark (CEH). Lina Mercado suggested that he look in Becky Oliver's runs. He said:
"But I have taken a look in an area that I know she used (for CEH people this is ~rfu/IMPALA/jules-vn4.6/farq_pn_acclim_adapt_medgs_ltdiags/nml) and “grep” has found a few instances:

WFDEI_5pft_1pt/pft_params.nml:g1_io=3.99, 2.22, 5.56, 1.10, 2.24,
WFDEI_5pft_regional_ltdiags/pft_params.nml:g1_io=3.99, 2.22, 5.56, 1.10, 2.24,

So I am now using g1_stomata_io=3.99, 2.22, 5.56, 1.10, 2.24.
Runs for the UK region and for the whole globe are being run right now with these values for g1 for the different PFTs.

comment:4 Changed 4 months ago by pmcguire

I have done Medlyn JULES runs for the UK region, the China region and for the whole globe. These runs have been put through the ILAMB model/benchmark comparison software.

comment:5 Changed 5 weeks ago by pmcguire

I have sent the ILAMB output webpages for the Medlyn JULES runs for the whole globe and for the China region to collaborators for evaluation. The whole globe ILAMB comparison has been evaluated, and the differences between the control study (Jacobs model) and the new run with Medlyn parameters have been found to be small but not insignificant.

Note: See TracTickets for help on using tickets.