Further discussion related to 'A formal derivation and numerical modeling of the Improved Boussinesq equations for varying depth' by Beji and Nadaoka

To finalise a discussion, it is pinpointed exactly where Beji and Nadaoka ( 1996) make an implicit and arbitrary choice in their Boussinesq equation en hancement procedure. Without this choice their claimed superior formalism v anishes. More importantly, it is shown that their new conclusions (put forw ard in the reply by Beji;md:Nadaoka, 1998) are not correct: it was claimed that a) Boussinesq equations formulated in either the depth-averaged veloci ty or the surface velocity contain superior accuracy and b) that the presen ce of truncation errors in both the mass and the momentum equation is the r eason for Nwogu's (1993) equations not exhibiting "exact" energy conservati on.