New directory with tools to prepare restarts from any reanalysis by interpolating/extrapolating/filling empty seas with a climatology

Changed CFU_DTF so that removing the forecast GMT signature in the forecast field is removed by using one regression model only, not for each member.

Changed CFU_MODE making sure NAs are produced if NAs are projected, not 0s.

common_diagnostics.txt removed. Now it is part of s2dverification project. See: http://ic3.cat/wikicfu/index.php/Tools/s2dverification

Now it is using the land-sea mask of any forecast system.

Now the selection of forecast system via fcsys works.

In CFU_PB quantile computation type set to 8. Before type was set to 7.

Now lat and lon objects should be correct.

wasted) + times failed.

Usage:
- Upload exp_job_info.sh to the platform where yo are running the
experiment.
- Run the script ./exp_info.sh <ARCH> <EXPID> <ESTIMATED RUNTIME>
e.g.: ./exp_info.sh mn m00v 1.3 > m00v.txt
- Download the output to your local machine
- Run python plotting script

CFU_NAO() computes the North Atlantic Oscillation (NAO) index based
on the projection onto the pattern of the leading observed EOF mode.
See the extensive description in the header of the function.

CFU_plotbox() produces a time series of the obs vs. box-and-whisker plots of
the re-forcast for a specific target period.
See the extensive description in the header of the function.

Now working fine when forecast lead dimension equals 1.

- deploy has already been shifted under separate project (https://dev.cfu.local/deploy_models.git), in case if someone interested to look into something feel free to grep the stuff from new project
- while transfer related scripts has got minor update...