pyesmda.ESMDA.cov_mm

property ESMDA.cov_mm: ndarray[tuple[Any, ...], dtype[float64]]

Get the estimated parameters autocovariance matrix. It is a read-only attribute.

The covariance matrice \(C^{l}_{MM}\) is approximated from the ensemble in the standard way of EnKF [Evensen, 2007, Aanonsen et al., 2009]:

\[C^{l}_{MM} = \frac{1}{N_{e} - 1} \sum_{j=1}^{N_{e}}\left(m^{l}_{j} - \overline{m^{l}}\right)\left(m^{l}_{j} - \overline{m^{l}} \right)^{T}\]

with \(\overline{m^{l}}\), the parameters ensemble means, at iteration \(l\).