We list here the simulations done for testing ORCHIDEE-CN at global scale, maintaining the ratio LEAF_M_c/LEAF_M_n fix (hereafter CN_ratio)

1) refCN3238Palmi01

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [-85] ; LON = [33]
• PERIOD LENGTH = 100 years
• FREQUENCY OUTPUT = yearly
• START TIME = 1901-01-01
• COMMENTS: CN_ratio is not stable for all PFTs in the grid cell. Depending on PFTs, the departure from the fixed value happens in different moment of the simulation.

2) refCN3238Palmi02

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• Imposing the PFT C3Ag (PFT10) for the whole grid.
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [-85] ; LON = [33]
• PERIOD LENGTH = 50 years
• FREQUENCY OUTPUT = yearly
• START TIME = 1901-01-01
• COMMENTS: The CN_ratio is not stable for the C3Ag PFT and stops exactly at the same time of the refCN3238Palmi01 simulation.

3) refCN3238Palmi03

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• Imposing the PFT C3Ag (PFT10) for the whole grid.
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [-85] ; LON = [33]
• PERIOD LENGTH = 50 years
• FREQUENCY OUTPUT = daily
• START TIME = 1901-01-01
• COMMENTS: The CN_ratio is not stable, we better investigate why it diverges. Probably some problems in stomate_growth_fun_all.f90 where the "impose_cn" condition is treated.

4) refCN3238Palmi04

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• Imposing the PFT C3Ag (PFT10) for the whole grid.
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [-85] ; LON = [33]
• PERIOD LENTH = 20 years
• FREQUENCY OUTPUT = daily
• START TIME = 1901-01-01
• COMMENTS: Some changes in stomate_growth_fun_all.f90 to optimize the part of the code linked to "impose_cn" condition. Things are not really improved. The origin of problems is more likely that N_UPTAKE_NO3 becomes negative and N_UPTAKE_NH4 becomes zero.

5) refCN3238Palmi05

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• Imposing the PFT C3Ag (PFT10) for the whole grid.
• CyclicBegin = 1911 ; CyclicEnd = 1911
• LAT = [-85] ; LON = [33]
• PERIOD LENTH = 5 years
• FREQUENCY OUTPUT = daily
• START TIME = 1911-01-01
• RestartPath = /ccc/scratch/cont003/dsm/p24palm/IGCM_OUT/OL2/TEST/refCN3238Palmi04
• RestartJobName = FG1low.CN.Palmi04
• COMMENTS: We investigate when CN_ratio starts to change. Printed many variables in stomate_soilcarbon.f90 into the stardand output at each time step (every 30 minutes) to see where the problem comes from. We see that plant_uptake(:,m,initrate) is negative.

6) refCN3238Palmi06

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• Imposing the PFT C3Ag (PFT10) for the whole grid.
• CyclicBegin = 1911 ; CyclicEnd = 1911
• LAT = [-85] ; LON = [33]
• PERIOD LENTH = 5 years
• FREQUENCY OUTPUT = daily
• START TIME = 1911-01-01
• RestartPath = /ccc/scratch/cont003/dsm/p24palm/IGCM_OUT/OL2/TEST/refCN3238Palmi04
• RestartJobName = FG1low.CN.Palmi04
• COMMENTS: If plant_uptake(:,m,initrate) is negative, we impose plant_uptake(:,m,initrate) = 0. If we impose it the CN_ratio is stable.

7) refCN3238Palmi07

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• Imposing the PFT C3Ag (PFT10) for the whole grid.
• CyclicBegin = 1911; CyclicEnd = 1911
• LAT = [-85] ; LON = [33]
• PERIOD LENTH = 5 years
• FREQUENCY OUTPUT = daily
• START TIME = 1911-01-01
• RestartPath = /ccc/scratch/cont003/dsm/p24palm/IGCM_OUT/OL2/TEST/refCN3238Palmi04
• RestartJobName = FG1low.CN.Palmi04
• COMMENTS: Printed many more variables in stomate_soilcarbon.f90 in the stardand output at each time step (every 30 minutes), to see why plant_uptake(:,m,initrate) becomes negative. We find that when immob(:,:) is bigger than the nitrogen in the two soil pools (soil_n_min pools (nitrate and ammonium) and soil_n_min(:,:,initrate)), plant_uptake(:,m,initrate) becomes negative.

8) refCN3238Palmi08

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [-179.0, 179.0] ; LON = [-89.0, 89.0]
• PERIOD LENTH = 30 years
• FREQUENCY OUTPUT = monthly
• START TIME = 1901-01-01
• COMMENTS: Global simulation imposing leaching = 0 in stomate_soilcarbon.f90. We test the impact of leaching on CN_ratio. This simulation gives weird results. The CN ratio change very little in comparison to reference simulations (refCN3238Palmi09) and performs worse that refCN3238Palmi10 (see below).

9) refCN3238Palmi09

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [-179.0, 179.0] ; LON = [-89.0, 89.0]
• PERIOD LENTH = 30 years
• FREQUENCY OUTPUT = monthly
• START TIME = 1901-01-01
• COMMENTS: Global simulation, with no code change compared to refCN3238Palmi07 simulation. It is the reference run to compare with refCN3238Palmi08 and refCN3238Palmi10 simulations.

10) refCN3238Palmi10

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [-179.0, 179.0] ; LON = [-89.0, 89.0]
• PERIOD LENTH = 30 years
• FREQUENCY OUTPUT = monthly
• START TIME = 1901-01-01
• COMMENTS: Global simulation, leaching is calculated after n_adsorbed in stomate_soilcarbon.f90. The CN_ratio is more stable, but anyway starts to change.

11) refCN3238Palmi11

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [-179.0, 179.0]
• LON = [-89.0, 89.0]
• PERIOD LENTH = 200 years
• FREQUENCY OUTPUT = yearly
• START TIME = 1901-01-01
• COMMENTS: in stomate_soilcarbon.f90 we add nitrogen when immob(:,:) is bigger than the nitrogen that can be given by the soil pools (nitrate and ammonium) and so soil_n_min(:,:,initrate) become negative. We take memory of the added nitrogen in N_support. We calculate leaching after n_adsorbed calculation (as in Zaehle version), in other words the ammonium available to be leached is decreased. This makes sense as the ammonium in less soluble than nitrate.

12) refCN3238Palmi12

• STOMATE_IMPOSE_CN = y
• Set-up as Nicolas run FG1low.Choi.CN3238
• Resolution = 2 degree
• CyclicBegin = 1901 ; CyclicEnd = 1910
• LAT = [47]
• LON = [83]
• Imposing the PFT BoBrDe (PFT08) for the whole grid.
• PERIOD LENTH = 200 years
• FREQUENCY OUTPUT = daily
• START TIME = 1901-01-01
• COMMENTS: Corrected a bug in stomate_phenology.f90 related to the calculation of labile biomass calculation. There are still some points that do not maintain a fixed CN_ratio in refCN3238Palmi11. We investigate in one of these points (China) to see if now it is fixed.