The data are assumed to have already undergone quality control (I think).

Directories:
   Deep        -  Deep data.
   Shallow     -  Shelf data.
   ShallowTop  -  Shelf data, casts that stop w/in 8m of surface are extended
                  to the surface.

The first step is to select the density bins with which HydroBase will
average.  Samples that I chose can be found in ShallowTop/sigma*.dat.
Note that due to pre-set limitations a maximum of 500 bins can be
specified.

   ConvJob     - Converts extended shelf data to MODS format for examination
   ConvJobDeep - Converts deep data to MODS format for examination
      HB2Mods1File - called by both of the above.

   show_pden.m - MATLAB script to plot density for a guide in bin selection.

After examining the data, I decided to extend the shallow data to the
surface.  Roughly estimating the mixed layer depth to be 8m, I extended
any cast that stopped within 8m of the surface to the surface.

   ExtendShallow - Extend the shallow data
      Extend1File - Called by the above

The next step is to use HydroBase to average the data.  Due to the
size of the data sets, I break-up the task by Marsden Squares.

   GridEm_New     - Grid the extended shelf data.
   GridEmDeep_New - Grid the deep data.
      Grid1FileNew - Called by both of the above.

Then I convert the netCDF file produced by HydroBase to MODS format.
I also, at this step, reject values constructed with less than a desired
number of data points.  I've been using 5 as the cut-off.

   loc2mods2.m - MATLAB script to convert the HydroBase files.
      pjh_greg.m   - called by loc2mods2
      date_stamp.m - called by loc2mods2

The various Marsden square MODS files are concatenated using

   CatJob

The data have a very uneven distribution over the shelf.  Around MassBay
I could construct a 1/8 degree climatology.  Over some portions of the
Gulf of Maine, I'm luck to get 1 degree.  I've included an attempt to
combine climatologies with different resolutions, retaining the most
finely resolved sets where available.

   no_overlap_climos.m - MATLAB script to combine climatologies.
      find_overlap.m - called by no_overlap_climos
      rm_ovrlp.m     - called by no_overlap_climos

A final step I use is to restrict the resulting MODS files to a
region more closely approximating the domains (reduced wear and
tear on OA)

   SelPolyJob   - Limit climatologies to neighborhood of domains
   SelPolyShJob - Limit results of above to 250m isobath.
