How to configure a transient climate simulation with constant surface data?

[KeerZ]

Hello, I am configuring a transient climate simulation based on SSP5-8.5. For this case, I'd like to use a constant surface data at 2015 which I generated using a new urban dataset. The compset I used is : BSSP585.

To turn off the transient landuse and use my own surface data, in the user_nl_clm, I added：
check_dynpft_consistency =.false.
do_transient_crops=.false.
do_transient_pfts=.false.
flanduse_timeseries=' '
fsurdat ='glade/work/keerzhang/surfacedata/2015.nc' (my new surfdata)

I am wondering whether it is the correct way to do this?


Also, I set CLM_BLDNML_OPTS='-bgc sp'. But I am now wondering is it proper to turn off BGC-CROP and use SP (satellite phenology) for a transient climate simulation? Or what will be the difference between using BGC-CROP and SP in my case?

I'm currently using CESM2.1.2, CLM5.0 on Cheyenne. You can find my testcase in folder /glade/work/keerzhang/testcases/

Any suggestions will be appreciated!
Thanks!
Keer

 

[KeerZ]

Another question about runoff variables I intend to output:

I've read about the definitions of various runoff outputs (QRUNOFF, QOVER, QH2OSFC and QRGWL) in the CLM5 tech note. I'd like to confirm that, if my main research focus is urban inland inundation, should I use QOVER (surface runoff)+QH2OSFC (surface water runoff)+QRGWL (liquid runoff from glaciers/lakes) at urban landunit? It seems that I cannot use QRUNOFF_U (total runoff at urban) because it contains QDRAI (subsurface runoff) while no QH2OSFC (surface water runoff).

Thank you!

 

[erik]

Yes, your setup looks correct to me. Setting the landuse.timeseries file to blank means it won't try to find one, and explicitly setting the options to off will ensure they aren't used. You probably just need to set fsurdat and flanduse_timeseries, but no harm done in setting the others.

Running SP mode means that LAI will be prescribed according to LAI observed from Satellite data from your surface dataset. Running in BGC-Crop mode means plant phenology will be prescribed by the BGC model. Crop means crops will be managed by the model, rather than just using generic crops (which essentially act like grasses).

 

[erik]

Another question about runoff variables I intend to output:

I've read about the definitions of various runoff outputs (QRUNOFF, QOVER, QH2OSFC and QRGWL) in the CLM5 tech note. I'd like to confirm that, if my main research focus is urban inland inundation, should I use QOVER (surface runoff)+QH2OSFC (surface water runoff)+QRGWL (liquid runoff from glaciers/lakes) at urban landunit? It seems that I cannot use QRUNOFF_U (total runoff at urban) because it contains QDRAI (subsurface runoff) while no QH2OSFC (surface water runoff).

Thank you!

Your reasoning here sounds correct to me. If you can't get the "_U" output variables that you want, you'll either need to make code changes to add them in. By imitating how it's done elsewhere. Or you'll need to output your history data as a vector and then do data processing on it to visualize it. Your turn off the 2D history averaging to gridcell and then select what type of averaging you want to have, which could be to LANDUNIT or at PFT level.

 

[KeerZ]

Your reasoning here sounds correct to me. If you can't get the "_U" output variables that you want, you'll either need to make code changes to add them in. By imitating how it's done elsewhere. Or you'll need to output your history data as a vector and then do data processing on it to visualize it. Your turn off the 2D history averaging to gridcell and then select what type of averaging you want to have, which could be to LANDUNIT or at PFT level.

Yes! I intend to output landunit-level QOVER, QH2OSFC and QRGWL as a vector, and then get the urban results for these variables. Thank you very much!

 

[oleson]

Right, although the surface water formulation isn't implemented for urban landunits. So I think QH2OSFC will be zero.

 

[ykp990521]

Yes, your setup looks correct to me. Setting the landuse.timeseries file to blank means it won't try to find one, and explicitly setting the options to off will ensure they aren't used. You probably just need to set fsurdat and flanduse_timeseries, but no harm done in setting the others.

Running SP mode means that LAI will be prescribed according to LAI observed from Satellite data from your surface dataset. Running in BGC-Crop mode means plant phenology will be prescribed by the BGC model. Crop means crops will be managed by the model, rather than just using generic crops (which essentially act like grasses).

Hello, could you tell me from which year's observation is the satellite phenology derived from? Thanks a lot

 

[lawrencepj1]

Hello, Yes the CLM5 PFT LAI values are derived from MODIS MCD15A2 version 5 8-day data from 2003 to 2015, combined with the current day PFT distributions, and CRU monthly climatology surface temperature data. The new PFT LAI and SAI values are generated following updated methods from Lawrence and Chase (2007). The new PFT LAI and SAI data are combined with the ICESAT derived PFT Canopy Height, to provide CLM5 prescribed monthly vegetation phenology and morphology. Hope that answers your question :-) Peter

 

[ykp990521]

Hello, Yes the CLM5 PFT LAI values are derived from MODIS MCD15A2 version 5 8-day data from 2003 to 2015, combined with the current day PFT distributions, and CRU monthly climatology surface temperature data. The new PFT LAI and SAI values are generated following updated methods from Lawrence and Chase (2007). The new PFT LAI and SAI data are combined with the ICESAT derived PFT Canopy Height, to provide CLM5 prescribed monthly vegetation phenology and morphology. Hope that answers your question :-) Peter

Thanks a lot for your reply! I wish to ask a little bit in detail, if I run a 1850 or mid-holocene case but with CLM under SP condition, is the satellite phenology still derived from 2003 to 2015? Since I wish to run two experiments, one is 1850's constant forcing and another is mid-holocene's forcing, I hope to know exactly when the SP data is derived from. Additionally, could you tell me where you get the information about where SP's data comes from? Many thanks!

 

[lawrencepj1]

Thanks a lot for your reply! I wish to ask a little bit in detail, if I run a 1850 or mid-holocene case but with CLM under SP condition, is the satellite phenology still derived from 2003 to 2015? Since I wish to run two experiments, one is 1850's constant forcing and another is mid-holocene's forcing, I hope to know exactly when the SP data is derived from. Additionally, could you tell me where you get the information about where SP's data comes from? Many thanks!

Yes the CLM5 Satellite Phenology PFT LAI is taken from 2003 to 2015 for SP simulations independent of climate period being simulated. To address the need to have different PFT distributions the PFT LAI for each month is extrapolated globally so there will always be an LAI value for any PFT for any month. This does not guarantee that they will be correct for the climate being simulated. I am currently writing up details of all of the CLM5 data in the CLM5 Land Data Technote. If you would like a copy of the still in progress technote please email me at lawrence@ucar.edu cheers Peter

 

[ykp990521]

Yes the CLM5 Satellite Phenology PFT LAI is taken from 2003 to 2015 for SP simulations independent of climate period being simulated. To address the need to have different PFT distributions the PFT LAI for each month is extrapolated globally so there will always be an LAI value for any PFT for any month. This does not guarantee that they will be correct for the climate being simulated. I am currently writing up details of all of the CLM5 data in the CLM5 Land Data Technote. If you would like a copy of the still in progress technote please email me at lawrence@ucar.edu cheers Peter

Many thanks for your so in time and quick answer!