Systematic bias in baroclinic energy estimates in shelf seas

A simple model of an internal wave advected by oscillating barotropic flow
suggests flaws in standard approaches to estimating properties of the internal
tide. When the M2 barotropic tidal current amplitude is of similar size to
the phase speed of the M2 baroclinic tide, spectral and harmonic analysis
techniques lead to erroneous estimates of the amplitude, phase, and energy in
the M2 internal tide. In general, harmonic fits and bandpass or low-pass filters
that attempt to isolate the lowest M2 harmonic significantly underestimate
the strength of M2 baroclinic energy fluxes in shelf seas. Baroclinic energy
flux estimates may show artificial spatial variability, giving the illusion of
sources and sinks of energy where none are present. Analysis of previously
published estimates of baroclinic energy fluxes in the Celtic Sea suggests this
mechanism may lead some values to be 25 to 60% too low