Hi Marc
in the newest version for windows the poisson's ratio can not be fixed!!
when i choose the poisson ratio from .2 to .2 no run complets!
poisson's ratio
Currently the Poisson's ratio implementation does not work if it is completely fixed. It might be something to implement in the future: when Poisson's ratio is fixed, all Vp parameters must be removed for the inversion itself. In fact, if Poisson's ratio is fixed, Vp is directly computed from Vs. And conditions set on Vp profile can be directly translated into conditions on Vs profile.
In real world, generally we don't know all these parameters with a very good precision. If you suspect the Poisson's ratio to be around 0.45, set limits like 0.44 to 0.46 (depending on your prior knowledge). If you set such restrictive conditions on Poisson's ratio, maybe that it will difficult to find a good misfit, maybe not. If your inversion does not work properly, I would suggest to start with a large prior range (200 to 6000 for Vp, 0.2 tp 0.5 for Poisson's Ratio and 100 to 3500 for Vs) and a very simple model (a simple gradient over a half space). Then progressively increase the number of layers (3, 4, not more in general). Never keep just the results of a two-layer model, there might be some over estimation of your resolution power a greater depth.
Poisson's ratio full control as implemented in last snapshot is still in its early stage. I've still not a wide experience with it on real cases. Waiting for your feedback
In real world, generally we don't know all these parameters with a very good precision. If you suspect the Poisson's ratio to be around 0.45, set limits like 0.44 to 0.46 (depending on your prior knowledge). If you set such restrictive conditions on Poisson's ratio, maybe that it will difficult to find a good misfit, maybe not. If your inversion does not work properly, I would suggest to start with a large prior range (200 to 6000 for Vp, 0.2 tp 0.5 for Poisson's Ratio and 100 to 3500 for Vs) and a very simple model (a simple gradient over a half space). Then progressively increase the number of layers (3, 4, not more in general). Never keep just the results of a two-layer model, there might be some over estimation of your resolution power a greater depth.
Poisson's ratio full control as implemented in last snapshot is still in its early stage. I've still not a wide experience with it on real cases. Waiting for your feedback