I am implementing a method for modelling realistic ocean surfaces using
a model as described by Jerry Tessendorf (), describing the ocean
surface as a quasi-stationary gaussian process, using the Phillips
spectrum as described in the article.
I have implemented the model and get good results, however I am quite
confused by the role, the "numerical constant A" plays in the overall model.
Obviously it is affecting the global wave height, but it seems to be
highly dependent on the parameters chosen, like wind velocity, region covered,
etc. Since I am not able to retrieve the paper by Phillips et al, I would
appreciate any comments by people which are using the same technique, if
they are changing the variable at all when varying the parameters to the
model, and if so, which heuristic they use.
BTW: I am intending to use more sophisticated energy spectra like the
Pierson-Moskowitz, JONSWAP and others. Any references on the
implemention issues of such models would be highly appreciated !
 Jerry Tessendorf, "Simulating Ocean Water", SIGGRAPH2001 Course Notes