which is, apart from a constant 3 / (32 π2), the ratio of the amplitudes of the second-order to the first-order term in the free surface elevation.[2]
The used parameters are:
H : the wave height, i.e. the difference between the elevations of the wave crest and trough,
h : the mean water depth, and
λ : the wavelength, which has to be large compared to the depth, λ ≫ h.
So the Ursell parameter U is the relative wave height H / h times the relative wavelength λ / h squared.
For long waves (λ ≫ h) with small Ursell number, U ≪ 32 π2 / 3 ≈ 100,[3] linear wave theory is applicable. Otherwise (and most often) a non-linear theory for fairly long waves (λ > 7 h)[4] – like the Korteweg–de Vries equation or Boussinesq equations – has to be used.
The parameter, with different normalisation, was already introduced by George Gabriel Stokes in his historical paper on surface gravity waves of 1847.[5]
^This factor is due to the neglected constant in the amplitude ratio of the second-order to first-order terms in the Stokes' wave expansion. See Dingemans (1997), p. 179 & 182.
^Stokes, G. G. (1847). "On the theory of oscillatory waves". Transactions of the Cambridge Philosophical Society. 8: 441–455. Reprinted in: Stokes, G. G. (1880). Mathematical and Physical Papers, Volume I. Cambridge University Press. pp. 197–229.