In Jun-Aug 2007, I participated in an internship with Andreas Münchow at the University of Delaware's program in Physical Ocean Science and Engineering.
The internship consisted of refurbishing some of my advisor's old FORTRAN code to analyze
some recently collected data from the R/V Hugh Sharp. The code parsed the raw
binary stream and performed its own statistical analysis on it. It was combined
with a series of shell scripts to further automate the process. Unfortunately,
the code had been left disused for many years, and was tailored to a specific
cruise. As a result, I was able to learn FORTRAN and increase my skills with
shell scripting by going through the code and making it generic. This
further involved reading documentation and papers
about the workings of
the ADCP, so that I could understand, and in some cases, improve the
analysis of the raw data.
The resulting documentation can be found here (410K *.pdf)
Analysis of Tides
Orientation of the Main Tidal Component
Recalculating the velocity profiles from the raw binary stream allowed us to use more or less stringent statistical techniques than the proprietary analysis software would have otherwise applied.
Phase of the Tidal Ellipse
For example, default values for data being thrown out by the program could be
increased in some
cases to salvage data that would have otherwise been unavailable. Similarly,
default thresholds for an ADCP used in say, an icebreaker, could be relaxed for
the much calmer Delaware Bay.
Semi-major Axis of the Tidal Ellipse
After that, I was amazed to find out that a combination of Fourier analysis and
knowledge of astronomy could be used to extract the tides from the velocity
profiles, and furthermore, that what was left over, the mean outflow, could
provide an important boundary condition on the dynamics of the Bay itself.
Subtidal Outflow Perpendicular to Transect
I have since learned that it is a visualization
to use the rainbow color map.
However, at the time I felt very clever in finding a way to register the 3-D
data onto a 2-D plane, and then to display the data in a simple way. The result is the
Outward- and Across channel flow images shown here. They are shown next to a
lat/long ship-track plotted with the CTD locations of the cruise. It is
easy to see the two lobes of the tidal ellipses in the first two images, and the floating,
fresh, and fleeing mass of water flowing out of the bay in the second.