KalmanSmoother
Apply the Rauch-Tung-Striebel smoother to a gravity field time series computed by KalmanFilter. This is the implementation of the approach presented in
Kurtenbach, E., Eicker, A., Mayer-Gürr, T., Holschneider, M., Hayn, M., Fuhrmann, M., and Kusche, J. (2012). Improved daily GRACE gravity field solutions using a Kalman smoother. Journal of Geodynamics, 59–60, 39–48. https://doi.org/10.1016/j.jog.2012.02.006.
The result has zero phase and the squared magnitude response of inputfileAutoregressiveModel (see autoregressiveModel for details). inputfileUpdatedState and inputfileUpdatedStateCovariance are the output of a KalmanFilter forward sweep. The matrix files foroutputfileUpdatedState, inputfileUpdatedState and inputfileUpdatedStateCovariance can also be specified using loops.
See also KalmanBuildNormals, KalmanFilter and KalmanSmootherLeastSquares.
| Name | Type | Annotation |
|---|---|---|
outputfileState | filename | estimated parameters (nx1-matrix) |
outputfileStateCovarianceMatrix | filename | estimated parameters' covariance matrix |
inputfileUpdatedState | filename | |
inputfileUpdatedStateCovarianceMatrix | filename | |
inputfileAutoregressiveModel | filename | file name of autoregressive model |