I've got a bunch of simulations, some of which get pretty violent, where w and zeta don't seem to match. In a rectangular channel model I get a big standing wave across the channel which looks realistic except the surface is going up when w in the top layer is negative. Sometimes it seems to be exactly reversed and in others its just different. In simple cases where there is just a wave of normal magnitude it looks correct so I don't think anything is flipped in the displays. I looked at the raw history files and see the same thing. I also see non-zero omega at the top layer (although it is small). Is that possible? I guess the real question is what can cause w to not be the same as dzeta/dt as reported in the history files.
Thanks
Wayne
reconciling w and dzeta/dt
I'll answer my own question, which mightl help someone else someday. I should have followed the basic rule: " when things look funny, increase the sampling rate." The wave motions which I thought I was seeing (and seemed to match what I expected) where aliased snapshots of an unexpected, much higher frequency disturbance. So my history sample time, nhis, was way too slow for what was actually occuring. As a result I got correct snapshots of w but was fooled by bogus (low freq) motion of the surface. It's amazing how real an aliased signal can look when you like what you see. Sorry for this disturbance.
Wayne
Wayne