wood harvest problem

[liliyao]

Hi, I have a question about the wood harvest in CLM5BGC mode. I was doing a simple test on one grid cell to see the impact of wood harvest rate on different output variables. I ran the offline CLM5 simulation with compset I2000Clm50BgcCropGs using CESM2.1.3 for 2 years, res=CLM_USRDAT. On the surface data, PCT_NATVEG = 99.996 % , and 99.2% of the natural veg land unit coved by PFT 1 (i.e., Needleleaf evergreen tree-temperate). I tried different values of constant harvest rates (including CONST_HARVEST_SH1, SH2, SH3, VH1, VH2). But the output variable “WOOD_HARVESTC” was always 0. I think this result means no wood harvest occurred in my simulation, but I have no idea why. I have attached my domain and surface files.

Any help you could provide would be much appreciated. Thanks.

 

[erik]

There's a couple things here I wonder about with your case. Wood harvest only occurs on the year boundary (so midnight on Dec/31st). So if you ran for two years there's only two time steps when you'd see harvest. There's also a namelist item that turns wood harvest off, and for that matter wood harvest is ONLY enabled when you are running with land-use timeseries files. So it really depends on how your namelist was setup. It also looks like WOOD_HARVESTC is setup as an average, so it's only going to be non-zero for January, all other months would be zero.

 

[liliyao]

Hi Erik, thank you very much for your reply. After got your reply, I modified my case. This time it runs with land-use timeseries for 5 years, and wood harvest is turned on in user_nl_clm. Yes, I see non-zero WOOD_HARVESTC from the output.

I have a question on the principle of wood harvest in CLM, i.e., mass-based:
Based on the CLM Technical Note, wood harvest in CLM is mass based rather than area based, however, the output shows that total Leave Area Index (LAISHA+LAISUN) changes with wood harvest rate as presented in the attached figure (though the difference is very small). Does it mean that wood harvest by mass will change the areal fraction of tree in CLM? Thanks!
I have attached the surface data, domain file, and land-use timeseries in case you are interested in them. Actually the small harvest case is ran with the original 1x1_brazil data from NCAR data repository, and the wood harvests were replaced with larger values for the large-harvest case.

 

[oleson]

You can see how the harvest rates are applied in this routine:

src/biogeochem/dynHarvestMod.F90

 

[liliyao]

Thank you very much, Dr. Oleson.
After reading dynHarvestMod.F90 and CNVegStructUpdateMod.F90, I think the simulation results make sense to me now. Wood harvest rate affects leave C. Leaf area index is calculated as a function of leave C and specific leaf area in CLM, therefore, affected by wood harvest rate.