This CD contains objective map "cubes" of the final version
of the Scanfish data data in matlab (.mat) format, loacted in the
subdirectory "cube_data".

These cubes were generated by objetively mapping the Scanfish
data at 60 depth horizons between 2m and 120m. As input
we used the "average profiles", also contained on this CD.


The data are contained in several .mat files, according to
the following naming scheme:
the file "sw06_grid5_sal_map.mat" contains the "cube" for salinity
data from the 5th (out of 12) grid, or Scanfish deployment.

The file contains the following variables (as shown in the format
of the matlab command "whos"):

latitude	31x1	248	double array	latitude of mapped grid
longitude	36x1	288	double array	longitude of mapped grid
sal_m	60x31x36	535680	double array	the mapped salinity
sal_e	60x31x36	535680	double array	the error map
lat_prof	121x1	968	double array	latitude of the input average profiles
lon_prof	121x1	968	double array	longitude of the input average profiles
z	120x1	960	double array	depth levels of the maps

For the 5th grid, there happened to be 121 input profiles,
hence the dimension of lon_prof and lat_prof. Other grids
had different number of profiles, depending on the length of
the deployment.

The error map is essentially a measure of the distance of a
grid point to the nearest input data. Its values depend on the
prescribed correlation length (we used 10km), and the prescribed
error level (here: 0.1). For plotting, we used a filter that
excluded all gridded data for which the error map exceeded 0.4.