The current observations at Site B made with the bottom-mounted upward-looking Acoustic Doppler Current Profiler (ADCP) differ significantly from currents measured by the point current meters
during periods when the short-wavelength internal waves are present, except within about 20
m from the instrument (Scotti and others, 2005). These periods are flagged with
high error velocities,
defined as the difference between the vertical velocities estimated from each pair of beams, by the
standard ADCP algorithm. The algorithm assumes
that the current field is spatially uniform over the distance separating the beams, and combines
data from opposite beams to determine the horizontal and vertical velocities. The
ADCP fails to correctly determine the
components of the velocity in earth coordinates during passage of the short-wavelength internal
waves because the standard algorithm combines beam components that, at a given time, are sampling
different phases of the wave, violating the assumption of spatial homogeneity. However, the
individual velocity measurements in beam coordinates are uncorrupted. Scotti and others (2005)
show that measurements from the four spatially-diverging beams and the backscatter intensity signal
can be used to calculate the propagation direction and celerity of the internal waves. Once this
information is known, a modified beam-to-earth transformation that combines appropriately lagged
beam measurements can be used to obtain current estimates in earth coordinates that compare well
with point current measurements made by the FSI
and VMCM current meters.
The following MATLAB .mat files contain corrected ADCP
current observations for 2-hour intervals following the start of each internal wave event.
The following .nc files contain corrected ADCP
current observations for 2-hour intervals following the start of each internal wave event, merged into the original
file with data from periods with no internal waves.