Country specific_input_files - PIK-LPJmL/LPJmL GitHub Wiki
Country-specific input files
To establish a new grid for Mexico, log in to the cluster
- First you have to compile the utilities via make all
- then you need to set the PATH to the binaries of LPJmL. If lpjml is,
e.g. located in
/home/username/lpjml/trunk/bin
, then setexport PATH=$PATH:/home/username/lpjml/trunk/bin
- and go to the directory
/your/data/directory
which contains input data.
If you have a country file for managed lands here (e.g. cow_mg_2006.bin) and
the global grid file grid.bin, you can use these two files to “cut out” the grid for Mexico.
You can print the header of the binary country file
cow_mg_2006.bin with the command
printclm cow_mg_2006.bin
(if you are working with the new soil data
set, you may want to use cow_mg_2006_full.bin
in which all grid cells
have a valid country and region code (the cow_mg_2006.bin
) only has
valid values for the 59199 pixels with valid (old) soil data plus some
additional random ones.
It says that the file contains nbands = 2 which means the number of columns in this data file. In this case, the country file has two bands because for each grid cell ID, there are two entries for the countrycode and regioncode (sub-national) (see input). In the model, there are 67420 grid cells in total for the entire globe.
which printclm
shows the location of printclm
- The next step is to get the country ID and the coordinates for Mexico and create a new grid file specific for Mexico from the global files.
To do this, go to your LPJ root directory.
Enter the folder $LPJROOT/include
Within the file managepar.h
, you find the country IDs for each country
on the planet.
The country ID 106 is for Mexico.
Now enter the folder $LPJROOT/bin/
which contains utility functions for LPJmL. Among these files, you first
need to call the routine getcountry
on the global country and grid
files. The function takes 2 input files, the filename for the new grid,
and the country number as arguments. The command is
getcountry /p/projects/lpjml/input/historical/input_VERSION2/cow_mg_2006.bin /p/projects/lpjml/input/historical/input_VERSION2/grid.bin grid_mexico.clm 106
The newly created grid_mexico.clm
should contain a subset of cell IDs
from 0,…803 with entries for latitudes and longitudes (view with
printclm).
Well done! Now you have the grid for Mexico.
Apart from this grid, you also need climate and soil data to run your simulation. In the default setup, these are the files
tmp.clm //temperature
drainclass.bin //currently unused, but still required by LPJmL
wet.clm //number of wet days (important for weather generator)
pre.clm //precipitation
cld.clm //cloudiness
soil.bin //soil type
To create Mexico-specific files, you need the utility functions
regridclm and regridsoil. These take both the global grid file and
your new Mexico-specific grid and re-sample e.g. the global temperature
file to a Mexico-specific temperature file. To do this, run the command
regridclm -same /p/projects/lpjml/input/historical/input_VERSION2/grid.bin grid_mexico.clm /p/projects/lpjml/input/historical/input_VERSION2/tmp.clm tmp_mexico.clm
Note that you would have to include the path to your grid_mexico.clm
,
if you have not created it within the current directory
($LPJROOT/src/utils
).
Run the corresponding command for the other input files wet.clm
,
pre.clm
, and cld.clm
.
For soil.bin
and drainclass.bin
(scalar data), use the command
regridsoil -same /path/to/data/grid.bin grid_mexico.clm /path/to/data/soil.bin soil_mexico.bin
Last but not least, you need landuse data if you run a simulation including landuse and crop functional types in LPJmL. The default files are
/path/to/data/cow_mg_2006.bin
/path/to/data/cft1700_2005_16cfts_SR.bin
To “cut out” the Mexico-specific data, you need again the same utility functional as for the climate data. Run the commands
regridclm -same /path/to/data/grid.bin grid_mexico.clm /path/to/data/cow_mg_2006.bin cow_mg_2006_mexico.bin
regridclm -same /path/to/data/grid.bin grid_mexico.clm /path/to/data/cft1700_2005_16cfts_SR.bin cft1700_2005_16cfts_SR_mexico.bin
and you’re done.
You can now use all your newly established input files:
grid_mexico.clm
tmp_mexico.clm
drainclass_mexico.bin
wet_mexico.clm
pre_mexico.clm
cld_mexico.clm
soil_mexico.clm
cow_mg_2006_mexico.clm
cft1700_2005_16cfts_SR_mexico.clm
in the input.conf
for your LPJmL simulation.