A Note on internal gravity wave spectra

Several measurements of the slope of internal wave spectra have given wave number k dependencies of the form k−5/3 or k−2. The first of these is sometimes discarded because it is thought that Kolmogoroff's ideas of small-scale isotropic turbulence should not apply to internal gravity waves. However, it is suggested in this note that an empirically determined k−5/3 dependence is quite divorced from Kolmogoroff physics. Simple dimensional arguments show that k−5/3 spectra are possible for long internal waves, but that k−3 spectra might be more appropriate for short waves even though microstructure effects may obscure the results.