PFTs area

[heyaqian1987@gmail_com]

Hi everyone,I have finished runing expriment with componnet set: IHistClm50BgcCrop from 1982 to 2013 with transient land use and land cover. Now, I want to calculate forest and grass areas during 1982-2013. I looked the input data: landuse.timeseries_0.9x1.25_hist_78pfts_CMIP6_simyr1850-2015_c170824.nc, however it only provides PCT_NAT_PFT, which is percent plant functional type on the natural veg landunit (% of landunit), but not % of grid cell. Does anyone know how to calculate the areas?Thanks in advance. 

 

[sacks]

Sorry for the late reply here. (We had some interruptions when moving to the new forum software.)

If you look at the CLM output (history) files, you should see PCT_LANDUNIT. The first element (index 0) in each grid cell is the % of the natural vegetated landunit on the grid cell. You can get the % of each PFT on the grid cell by multiplying PCT_NAT_PFT by PCT_LANDUNIT(0).

 

[heyaqian1987@gmail_com]

Thanks Sack, I got it.

 

[AnuNTNU]

But when area of each grid cell is given in the output of CLM5, is it not enough to multiply each PFT with area to get the % of each PFT on the grid cell ? And may i know what are the 9 itypes, one coordinate and dimension of PCT_LANDUNIT, ?

 

[oleson]

I don't understand your first question, sorry. But you can find the 9 landunits types in the global attributes of a CLM history file,

:ltype_vegetated_or_bare_soil = 1 ;
:ltype_crop = 2 ;
:ltype_UNUSED = 3 ;
:ltype_landice_multiple_elevation_classes = 4 ;
:ltype_deep_lake = 5 ;
:ltype_wetland = 6 ;
:ltype_urban_tbd = 7 ;
:ltype_urban_hd = 8 ;
:ltype_urban_md = 9 ;

 

[AnuNTNU]

Thanks a lot for the reply. I can make my first question more clear. I am looking at the variable PCT_NAT_PFT of CLM input data,
landuse.timeseries_1.9x2.5_SSP5-3.4_78pfts_CMIP6_simyr1850-2100_c190228.nc. So I divided it by 100 and multipolied it with AREA (given in the data) of the cell to get the area fraction. Is this correct? or can you tell me why do we need to multiply PCT_NAT_PFT with PCT_LANDUNIT(0) from the output ?

 

[sacks]

PCT_NAT_PFT gives the percentage of each PFT on the natural vegetated landunit. Imagine a grid cell that is 95% lake and 5% natural vegetation. In this situation, the sum of PCT_NAT_PFT will still be 100%, because it just tells you the breakdown of the natural vegetation landunit into each PFT. So to get the actual areas of each PFT, you also need to multiply by PCT_LANDUNIT(0), which gives the area of the natural vegetated landunit.

 

[AnuNTNU]

So I assumed that we use PCT_LANDUNIT from the first timestep of the output. But it seems PCT_LANDUNIT is a function of time and it slighlty changes with time. Dont we need to take that into account?

 

[sacks]

Yes, in a transient run, both PCT_NAT_PFT and PCT_LANDUNIT can change in time. They stay constant for a given year, then can change on the first time step of the new year.