If we can come up with a fast-enough representation of the image tracing, as well as a flexible enough parameterization of the surface brightness/intensity distribution, we could actually take draws of the best-fit model and actually see what types of model are constrained by the data.
That would be great--no one's done this type of error visualization before.
If we can come up with a fast-enough representation of the image tracing, as well as a flexible enough parameterization of the surface brightness/intensity distribution, we could actually take draws of the best-fit model and actually see what types of model are constrained by the data.
That would be great--no one's done this type of error visualization before.