Tutorial for ED2 at Ugent HPC

Step 1: Compile ED2

1. download the model

   git clone [email protected]:EDmodel/ED2.git
   

2. change the makefile in ED2/ED/build/make

   cd ED2/ED/build/make
   cp include.mk.intel include.mk.intel_hpc

3. Edit include.mk.intel_hpc: input and output lib

   mainly for input and output lib

   HDF5_INCS=-I/apps/gent/CO7/skylake-ib/software/HDF5/1.12.1-iimpi-2021b/include
   HDF5_LIBS=-lm -lz -L/apps/gent/CO7/skylake-ib/software/HDF5/1.12.1-iimpi-2021b/bin -lhdf5 -lhdf5_fortran -lhdf5_hl

4. load the module to complie

   ml purge; ml intel-compilers/2021.4.0 HDF5/1.12.1-iimpi-2021b UDUNITS/2.2.28-GCCcore-11.2.0; ui

5. cd to the build dir and compile

   #-k E fast and efficiently run/ -k A will give some clue when it broken.

   cd .. 
   ./install.sh -k E -p intel_hpc
   

6. see what you get

Step 2: Preparing the driving data

1. download data prepare tool

   git clone [email protected]:qpantc/ED2support.git
   

2. load R module, cd to script dir and run R

   ml av R see module suitable for R

   ml purge; ml R-bundle-Bioconductor/3.15-foss-2021b-R-4.2.0 
   
   cd ED2support/R
   R

3. run the data download and convert script

   source("download.and.convert.input.ED2.R")

Step 3: Running model

ED2IN file

the ED2IN file may can just generated by PEcAn

1. ED2IN for site with

   cordination of the simulated sites

   NL%N_POI = 1
   NL%POI_LAT = 9.25
   NL%POI_LON = -79.75
   NL%POI_RES = 1

2. start and end date for the simulation

   start date of the simulation

   NL%IMONTHA = 1
   NL%IDATEA = 1
   NL%IYEARA = 1901
   NL%ITIMEA = 0

   end date of the simulation

   NL%IMONTHZ = 1
   NL%IDATEZ = 1
   NL%IYEARZ = 1903
   NL%ITIMEZ = 0

3. output dir

   NL%FFILOUT = '/scratch/gent/vo/000/gvo00074/vsc44253/ED_results/BCI/analy/analysis'
   NL%SFILOUT = '/scratch/gent/vo/000/gvo00074/vsc44253/ED_results/BCI/hist/history'

4. input database

   NL%VEG_DATABASE = '/user/data/gent/gvo000/gvo00074/ED_common_data/veg_oge/OGE2_'
   NL%SOIL_DATABASE = '/user/data/gent/gvo000/gvo00074/ED_common_data/soils/FAO/FAO_'
   NL%LU_DATABASE = ''
   NL%PLANTATION_FILE = ''
   NL%THSUMS_DATABASE = '/user/data/gent/gvo000/gvo00074/ED_common_data/ed_inputs/'
   NL%ED_MET_DRIVER_DB = '/data/gent/vo/000/gvo00074/vsc44253/R/climate.site/site.lat9.25N.lon79.75W/ED2/ED_MET_DRIVER_HEADER'
   NL%SOILSTATE_DB = ''
   NL%SOILDEPTH_DB = ''

   • VEG_DATABASE: land/water mask

     vegetation database, used only to determine the land/water mask. Fill with the path and the prefix.

   • SOIL_DATABASE: soil type data

     soil database, used to determine the soil type. Fill with the path and the prefix.

   • LU_DATABASE: land-use change disturbance rates

     land-use change disturbance rates database, used only when IANTH_DISTURB is set to 1. Fill with the path and the prefix.

   • PLANTATION_FILE: forest plantation fraction

     Character string for the path to the forest plantation fraction file. This is used only when IANTH_DISTURB is set to 1 and the user wants to simulate forest plantations. Otherwise, leave it empty (PLANTATION_FILE='')

   • THSUMS_DATABASE: phenology about chilling-degree and growing-degree days

     input directory with dataset to initialise chilling-degree and growing-degree days, which is used to drive the cold-deciduous phenology (you must always provide this, even when your PFTs are not cold deciduous).

   • ED_MET_DRIVER_DB: meteorological driver instructions

     File containing information for meteorological driver instructions (the "header" file).

   • OBSTIME_DB:

     File containing times of desired IOOUTPUT Reference file: /ED/run/obstime_template.time

   • SOILSTATE_DB: soil moisture and temperature infromation if ISOILSTATEINIT=1

     If ISOILSTATEINIT=1, this variable specifies the full path of the file that contains soil moisture and temperature information.

   • SOILDEPTH_DB: soil moisture and temperature information if ISOILDEPTHFLG=1

     If ISOILDEPTHFLG=1, this variable specifies the full path of the file that contains soil moisture and temperature information.

5. Empty the IEDCNFGF config

   config and events

   NL%IEDCNFGF   = '/mypath/config.xml'
   NL%EVENT_FILE = '/mypath/event.xml'

   SFILIN

   NL%SFILIN = ''

6. set included PFTs

   NL%INCLUDE_THESE_PFT = 1,2,3,4,16

Step 4: Run and postprocess model

1. run the model

   Job.sh file Or inside an interactive job After load the needed modules

   ../build/ed_2.2-opt-master-8d4c3aff -f ED2IN
   

2. postprocess model on hpc

   source("post.process.ED2.outputs.R")

Step 5: Plot model outputs

load("./outputs/analysis.RData")
matplot(datum$szpft$gpp[,12,c(2,3,4,18)],type = "l")