Question for you or someone in the audience: when using an anisotropic variogram, is it convention to fit ellipsoidal parameters or fix them in advance prior estimating model parameters? Any thoughts are greatly appreciated.
Thanks Jef for this lecture. I have a question. at 21:42 min. when you are explaining the variogram modeling with anisotropy. We agree that total anisotropy has a geometric and zonal components. But Is there any chance that the labels of Y and X directions in the variogram on the right are exchanged? Because it seems to me that along the Y- direction the variance is larger and range is shorter than along the X-direction. Thanks and regards.
