enkf_step: EnKF part of EnKPF

Description Usage Arguments Value

Description

Compute the non-random elements of the EnKPF update typically used for choosing gamma adaptively

Usage

1
enkf_step(y, xb, gam, P, H, R, y.resx, kloc = FALSE, ...)

Arguments

y

the observations

xb

the background ensemble

gam

gamma parameter

P

background covariance matrix

H

observation operator

R

observation covariance

y.resx

= y - Hxb the innovations

kloc

if TRUE use the localized Kalman gain formulation

...

additional arguments used in case kloc=TRUE (see kloc_gain)

Value

a list with Kalman gains, weights and other quantities of interest


robertsy/assimilr documentation built on May 27, 2019, 10:33 a.m.