Generating climate input files for LPJ-LMfire

1. Collect GCM output for the following variables (monthly mean unless specified):

   1. 2m air temperature (degC)
   2. diurnal temperature range (degC)
   3. monthly total precipitation (mm)
   4. days with precip. > 0.1 mm (optional) (days)
   5. total cloud cover (percent)
   6. 10m windspeed (m s-1)
   7. lightning stroke density (optional as long as you have a precursor variable) (km-2 d-1)

2. If the GCM output is delivered as a timeseries, calculate the climatological mean (multi-year monthly mean) of the above variables (12 months) cdo ymonmean

3. Where necessary, calculate the derived variables: (all steps calcwetdays.sh)

   • total monthly precipitation (cdo muldpm)
   • wet days (using calcwetdays.sh and calcwetdays.f90)
   • lightning on the basis of convective mass flux (Magi parameterization - magi_lightning.py; or other formula)

4. Generate de-biased GCM anomalies by subtracting GCM paleoclimate climatology from GCM baseline (2nd half of 20th century) climatology (12 months) make_LPJ_climate.sh

5. Interpolate GCM anomalies to 0.5 degree, this is the paleoclimate anomaly (paleo-anomalies, 12 months) (interpolate.sh) # This happens in make_LPJ_climate.sh

6. Add the paleo-anomalies to the hi-res present-day baseline climatology climate_wwna_wglc_shelves.nc, this becomes the "basefile" (12 months) (NB check units between GCM and baseline) (addanom.f90 in addanom.sh)

7. Create a timeseries of interannually variable climate by adding the basefile to detrended interannual variability from 20CR (makeclimate.f90 use ./makeclimate $jobfile $output)