A full installation of the MCD can now take up to 1.9Gb of disk space. This can be reduced to less than 0.9Gb if a fully functional database, with all dust scenarios, is required, but only data from either the ``standard dust'' or ``bright dust'' experiments are needed on disk (equivalent to a full installation of the version 1.0 MCD). If absolutely necessary the amount of disk space needed could be reduced further by only retaining a limited range of dust scenarios or seasons of interest within the data subdirectory.
To extract everything at once, set the working directory to the one which will contain the database subdirectory, check that about 1.9Gb of disk space is available, place the cassette in the DAT drive and type
tar -xvf /dev/rmt0h
where /dev/rmt0h is the name of the DAT device on your machine (which may vary). This will create a directory mcd in which all the MCD data and software is stored.
To chose which set of dust properties is used, create a symbolic link data within the mcd directory to point to either data_s for ``standard dust'' or data_b for ``bright dust,'' e.g.
cd mcd
rm data # if a symbolic link already exists, remove it
ln -s data_s data # to select the ``standard dust'' dataset for use
# or
ln -s data_b data # to select the ``bright dust'' dataset for use
or, alternatively, just rename either data_s or data_b to
data as required, having first removed any symbolic link called data.
If there is not room to extract everything at once then first extract the smaller directories which contain the documentation, GrADS scripts and MCDGM interface routines
tar -xvf /dev/rmt0h mcd/docs
tar -xvf /dev/rmt0h mcd/grads
tar -xvf /dev/rmt0h mcd/mcdgm
and then extract either of the larger data directories as required
tar -xvf /dev/rmt0h mcd/data_s
or
tar -xvf /dev/rmt0h mcd/data_b
again renaming to data once extracted.