The output problem

[LianGong98]

Hi, I‘m running a FATES single point simulation in CESM2.2.0 and the FATES version is sci.1.30.0_api.8.0.0. I chose two pfts (broadleaf_evergreen_tropical_tree, broadleaf_hydrodecid_tropical_tree), adjusted some parameters, and initialized with inventory. It could run successfully. However, some outputs always zero, such as AGB, Total biomass, Leaf biomass. Besides, the H2OSIO seems unreasonable, it was too dry in the beginning.

It is very difficult for me. Is it the problem of parameters? or the Initialization? and what should I do ?
Any advice would be greatly appreciated.

[rosiealice]

That does seem like a very low initial soil moisture? Did everything die of hydraulic failure? If you plot out the number of individual what happens? I wonder why the soil is starting so dry? I'm about to head out but it's worth trying to figure out if all the tree have died. What type of mortality they died from, and perhaps what H2OSOI we are initializing with here in the init_cold routines.

[rosiealice]

The mortality rates are output as

ih_m1_si_scpf
To
ih_m8_si_scpf

And (if they aren't labelled in the output), they are set here:

main/FatesHistoryInterfaceMod.F90
Around line 2561.
(Writing this on a train on my iPad so the line permalinks don't work)

[LianGong98]

Hi ,thank you for reply. The mortality rates from M1_SCPF to M8_SCPF are really weird, all were zero. And I initialized it with reference to https://fates-users-guide.readthedocs.io/en/latest/user/Model-Initialization-Modes.html
use_fates_ed_st3 = .true.
fates_inventory_ctrl_filename= ....
but the water in Patch file format (.pss) is zero Is soil initialization come from the Patch file format (.pss)? or should provide the initialization file? If not , how should I Initialize the H2OSOI?

[rosiealice]

Oh, its in St3 mode... I guess then you have static structure, by definition, and so my hypothesized cause isn't going to be applicable...

[rosiealice]

Sorry i didnt catch that from your post! What is the GPPand LAI? That would tell you if the physiology is working and whether this is an issue with the outputs or if the system isn't working correctly st all....

[LianGong98]

I am so soorry ,I attach the diagram of LAI and GPP now


and the GPP values like this :

and its units: gC/m^2/s

[rosiealice]

No worries! That looks to me like the model is running internally OK, but that it isn't printing our the biomass values etc. Perhaps because these are inputs not outputs in St3 mode, but I am not an expert on how that works Maybe @ckoven would know if we should be reporting those things?

You can just throw away the output until the soil moisture has stabilized here, as it doesn't look like it's having any major impact on the later outputs.

[LianGong98]

Thank you very much！I will try it and check my forcing data

[dlawrenncar]

We have seen something like this in the past with CLM if CO2 was inadvertently set to 0. Or, this could happen if there is something wrong with the forcing data (e.g., solar radiation is 0). Just a couple things that are worth checking.

…

On Thu, Oct 27, 2022 at 5:14 AM LianGong98 ***@***.***> wrote: I am so soorry ,I attach the diagram of LAI and GPP now [image: TLAI] <https://user-images.githubusercontent.com/97998383/198268296-aadd9cd1-79a3-4089-88a1-71a94d81fd8a.png> [image: GPP] <https://user-images.githubusercontent.com/97998383/198268326-e1f91094-27fd-4d94-95e1-4d62465d20dd.png> and the GPP values like this : [image: GPP] <https://user-images.githubusercontent.com/97998383/198269139-cdff0d3c-6fe1-4208-9fa2-8ae205a0de1b.jpg> and its units: gC/m^2/s — Reply to this email directly, view it on GitHub <#928 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFABYVFVRUSSY4CTAJSACOTWFJPYBANCNFSM6AAAAAARPVBKCQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[LianGong98]

OK .That's a good advice ,I will check my forcing data .Thank you very much!

[rosiealice]

I'm not sure anything is running incorrectly, those GPP numbers look fairly normal given their weird units.

[LianGong98]

Emmm.. maybe my incident solar radiationsomething wrong? but I am not sure:

[dlawrenncar]

That looks right to me. As Rosie said, GPP is actually not that low, so I would agree with her perspective that this doesn't seem like a forcing issue.

…

On Thu, Oct 27, 2022 at 8:31 AM LianGong98 ***@***.***> wrote: Emmm.. maybe my incident solar radiationsomething wrong? but I am not sure: [image: FSDS_in_2003-02] <https://user-images.githubusercontent.com/97998383/198315030-a1563fe5-114d-4d51-8987-beddce71cac7.png> [image: incident solar radiation] <https://user-images.githubusercontent.com/97998383/198315098-a8eb0d52-61a1-4879-9e9d-ad06f864c3fe.jpg> — Reply to this email directly, view it on GitHub <#928 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFABYVFT4FWNYIIRLHYXZHTWFKG3ZANCNFSM6AAAAAARPVBKCQ> . You are receiving this because you commented.Message ID: ***@***.***>

[rgknox]

@LianGong98 I saw that you are using the "st3" mode. This does require an initial condition of ecosystem structure. It appears that you have LAI, so it looks like you successfully initialized. Could you describe the initial condition, ie is it from census data, or something else?

[LianGong98]

Thank you very much for your reply ,the initial data derived from census data, just like this

I tried to change the value of water to 0.3, but it seemed to have no effect on results

[mpaiao]

The water column is just a legacy column from the original ED model (which generated pss and css files as outputs).
I don't think this is used by FATES, actually I think the only columns that are read in by FATES are trk, age, and area.

[mpaiao]

Am I interpreting it wrong? It seems GPP is going to zero after the beginning of the simulation, and then bouncing back, so I don't think it's an output issue, it may be a problem with initial conditions...

A couple of thoughts based on previous runs that caused massive dieback in the tropics:

1. Long wave radiation. @LianGong98 if you are running at BCI, the meteorological forcing does not have incoming long wave radiation. If FLDS is not removed from <path_to>/run/datm.streams.txt.CLM1PT.CLM_USRDAT, the model runs with zero long wave radiation and you end up with an ice age that kills all the plants. I don't think this is your case because plants eventually grow, but it may be worth checking the long wave radiation output (Variable FLDS).
2. Soil moisture. When I run with hydro deciduous, I ended up editing the initial soil moisture in the host land model. In the case of CLM, I replaced use_fates_planthydro with use_fates in this if block in biogeophys/WaterStateType.F90 (and also in the use statement at the beginning of the file). This had more of an effect on deciduous, not much in the evergreens.

[LianGong98]

yeah I did remove the FLDS
./preview_namelists
cp run/datm.streams.txt.CLM1PT.CLM_USRDAT user_datm.streams.txt.CLM1PT.CLM_USRDAT
sed -i '/FLDS/d' user_datm.streams.txt.CLM1PT.CLM_USRDAT
but I'm not sure whether I need to run ./preview_namelists again.
If I want to change the initial soil moisture, shoud I replaceduse_fates_planthydro with use_fates just like you did?

[mpaiao]

This is the correct setting for skipping long wave radiation from site met drivers, so you probably don't have the zero FLDS problem (worth checking the output just in case).

Regarding the initial soil moisture, yes, just replace use_fates instead of use_fates_planthydro, this should increase the initial moisture for every FATES run (not only when FATES-Hydro is enabled).

[LianGong98]

Thank you so much for your assistance , I'll do it.

Best wishes

[LianGong98]

@mpaiao Hi, I'm sorry to bother you again, there was a new error in plant hydraulics when I ran

but this error disappeared when the hydraulic module was turned off (use_fates_planthydro = .false. )
It's really annoying, I do not know why it happened.
Any advice would be greatly appreciated !

[rosiealice]

I am still not totally sure what proble you ar trying to solve here. As I see it, the soil moisture starts off too low, but then equilibrates and GPP seems to be normal after that. Because we are in st3 mode there isn't a feedback to the low productivity, and so I think you can just throw out this spinup phase output (it is normal practice to throw away the output before the soil moisture equilibrates in prescribed LAI mode(s)).

Is there some other problem I'm missing?