Author: Sara Minoli Affiliation: Potsdam Institute for Climate Impact Research, Germany Contact: sara.minoli@pik-potsdam.de
This folder includes R-coded scripts for computing sowing and harvest dates following the approaches from Waha et al. (2012) and Minoli et al. (2019).
Running the code:
- Create a configuration file:
./compute_sdate_hdate/configuration/configuration_myproject.R
- Define I/O paths (make sure automatically created directories are correct)
- Define crops to simulate
- Grid coordinates (by default LPJmL grid.bin is used, which is converted into a dataframe)
- Prepare monthly climate files:
./preprocessing/prepare_montly_climate.Rand./preprocessing/prepare_montly_climate.sh
- Compute long-term monthly average of temperature (degC), precipitation (mm), potential evapotranspiration (mm), P/PET ratio (frac), stored in a datatable with header: lon, lat, month, tas, pr, pet, ppet
- pass arguments and submit job via bash script
- Set up main script:
./compute_sdate_hdate/main_myproject.R
- Choose if On Cluster TRUE/FALSE
- Define Working directory
./compute_sdate_hdate/main_myproject.sh
- Define loops for running scenarios and crops in batch
- To run it type:
bash main_myproject.sh(note: module purge might be needed beforehand, if need to open old Rdata files)
Postprocessing:
- Winter-wheat suitable area can be derived by removing pixels with sowing_month==0 which indicates that the deafault sowing month was chosen
- A script for merging spring and winter wheat is provided: ./postprocessing/merge_winter_spring_wheat.R
Parameters:
- Adjust crop parameters in crop_parameters.txt (default are taken from Waha 2012 & Minoli 2019)
Debugging:
- ./compute_sdate_hdate/src/debugging.R can help troubleshooting