Diffusive/Dispersive behavior

[eusoubrasileiro]

Is there something am I missing why am i having dispersion?? Though propagation seams correct, we can see dispersion in space, wave width increasing with time and energy peak decreasing ( space differences??)

[eusoubrasileiro]

Even implicit has diffusion! Lax Method would be better or other high space approximation when we don't have wave front shocks or big gradients. More related in http://scicomp.stackexchange.com/questions/2360/how-can-i-derive-a-bound-on-the-spurious-oscillations-in-the-numerical-solution

For advection equation that is variable separation consequence of wave equation.

[eusoubrasileiro]

• numerical diffusion is related to lost of energy and broadening of the pulse with time. Problems related to approximate the solution by a taylor series.
• numerical dispersion is another whole thing....if related to spurious oscillations it's related to the convergence of the solution (non implicit methods) so Von Newman Analysis works for that and the post on stack-exchange is about it. Another thing also must be consider when dealing with waves that is niquest criteria and spatial resolution Alford 1979 Geophysics deals with that using its G parameters number of grid-points per-waveleght.

[eusoubrasileiro]

too much crazy thinking ... look here ****http://geophypython.blogspot.com.br/