Correct the instructions and the run script

Trying out the instructions myself, I found two little bugs that are
corrected with this commit.  I also reduced the default number of CPU
cores from 16 to 4, so that the setup can be run locally on a personal
computer, without the need for a server or a larger desktop computer.

README.md

@@ -42,8 +42,9 @@ Hans Burchard (submitted to JAMES, 2023)
 ### Preparation
 
 1. Download this repository and go to the downloaded directory.
-2. Create a sub-directory `bin` and copy the GETM executable to `bin/getm`.
-3. Look at the script [run.sh](run.sh), check the two settings at the
+2. Create the sub-directories `bin` and `store`.
+3. Copy the GETM executable to `bin/getm`.
+4. Look at the script [run.sh](run.sh), check the two settings at the
    beginning of the file, and correct them if needed:
     1. The path `GETMDIR` must point to the code-directory of GETM.
     2. The number of CPU cores in `nCPU` must be available on your computer.

run.sh

@@ -11,13 +11,13 @@
 GETMDIR=$HOME/tools/getm/code/ 
 
 # Choose the number of CPU cores to use
-nCPU=16
+nCPU=4
 
 echo "Creating input files"
 python3 Make_fjord_setup_322.py
 
 echo "Creating namelist files"
-editscenario --schemadir $(GETMDIR)/schemas -e nml . fjord_322.xml
+editscenario --schemadir $GETMDIR/schemas -e nml . fjord_322.xml
 
 echo "Starting GETM ..."
 time mpirun -n $nCPU bin/getm