Compute innovation and call PDAFomi_omit obs
This routine is used by some of the global filters (EnKF, LEnKF, PF, NETF) with OMI to omit observations if the innovation is too large. These filters do not compute the innovation for the ensemble mean, but this is needed to omit observations. Thus, this step is done here and then the omission routine of OMI is called.
Type | Intent | Optional | Attributes | Name | ||
---|---|---|---|---|---|---|
integer, | intent(in) | :: | dim_p | |||
integer, | intent(in) | :: | dim_obs_p | |||
integer, | intent(in) | :: | dim_ens | |||
real, | intent(inout) | :: | state_p(dim_p) | |||
real, | intent(in) | :: | ens_p(dim_p,dim_ens) | |||
real, | intent(inout) | :: | obs_p(dim_obs_p) | |||
real | :: | U_init_obs | ||||
real | :: | U_obs_op | ||||
integer, | intent(in) | :: | compute_mean | |||
integer, | intent(in) | :: | screen |