Does the model fit the measured record and its uncertainty?
Check units, scale, boundary conditions, measurement uncertainty and sensitivity to changed assumptions. Compare plausible event scenarios rather than treating a visually convincing reconstruction as independent proof.

