py__init_obs_covar_pdaf()

Provide observation error covariance matrix to PDAF.

This function is used in stochastic EnKF for generating observation perturbations.

Parameters

step: int

current time step

dim_obsint

dimension of global observation vector

dim_obs_p: int

dimension of process-local observation vector

covar: np.ndarray[np.float64, dim=2]

Observation error covariance matrix. shape: (dim_obs_p, dim_obs_p)

obs_p: np.ndarray[np.float64, dim=1]

Process-local observation vector. shape: dim_obs_p

isdiag: bool

Flag indicating if the covariance matrix is diagonal.

Returns

covar: np.ndarray[np.float64, dim=2]

Observation error covariance matrix. shape: (dim_obs_p, dim_obs_p)

is_diag: bool

Flag indicating if the covariance matrix is diagonal.