Strong instabilities of the background tend to have simple shapes (perturbations lie in a low-dimensional subspace)
2.5 day forecast verifying
on 95/10/21.
Note that the bred vectors
(difference between the
forecasts) lie on a 1-D space
It makes sense to assume that the errors in the analysis
(initial conditions) have the same shape as well:
the errors lie in the subspace of the bred vectors