Workshop on the

USE OF ENSEMBLES IN DATA ASSIMILATION

April 13-14, 1999, NCEP, Camp Springs, MD

TENTATIVE AGENDA

PARTICIPANTS

ABSTRACTS

LOCAL ARRANGEMENTS
 
 

John Derber:
    Data assimilation plans at NCEP (20 mins)

Jan Barkmeijer:
    Recent developments in the ECMWF Ensemble Prediction System   (30 mins)

Zoltan Toth:
    Estimating analysis uncertainty using the NCEP global ensemble  (30 mins)
 
 

Dale Barker:
    The Specification and Use of Synoptically-Dependent Background Errors in 3DVAR using information from an Error Breeding Cycle

Eugenia Kalnay:
    Preliminary experiments on applications of ensembles to data assimilation

Peter Houtekamer:
    Prospects for an operational ensemble Kalman filter (30 mins)

Martin Ehrendorfer (Presented by J. Barkmeijer, 10 mins):
    TBA

Craig Bishop:
    The Ensemble Transform Kalman Filter (30 mins)

Jeff Whitaker:
    Spread/Analysis Error relationships in a simple model (20 mins)

Chris Snyder:
    Dynamics and statistics of forecast errors in a quasi-geostrophic model (10-30 mins)

Tom Hamill:
    Analysis and Forecast Errors and the Ensemble Kalman Filter (30 mins)

Jim Purser:
    Estimates of flow dependent covariance structures (20 mins)

Lars Peter Riishojgaard:
    Anisotropic flow-dependent modeling of the forecast error correlations
 

Jeff Anderson, GFDL
Dale Barker, UK Met. Office, Bracknell, UK
Jan Barkmeijer, ECMWF, Reading, England
Craig Bishop, Pennsylvania State University, State College, PA
Kerry Emanuel, MIT, Boston, MA
Brian Etherton, Pennsylvania State University, State College, PA
Peter Houtekamer, AES, Dorval, Canada
Tom Hamill, NCAR, Boulder, CO
Eugenia Kalnay, University of Oklahoma, Norman, OK
Sharan Majumdar, Pennsylvania State University, State College, PA
Lars Peter Riishojgaard, NASA GSFC, Greenbelt, MD
Chris Snyder, NCAR, Boulder, CO
Jeff Whitaker, CDC, Boulder, CO

    From NCEP:
John Derber
Steve Lord
Dave Parrish
Jim Purser
Istvan Szunyogh
Zoltan Toth
Steve Tracton
Wan-Shu Wu
 

The Specification and Use of Synoptically-Dependent Background Errors
in 3DVAR using information from an Error Breeding Cycle

Dale Barker

A study is currently under way at the UKMO to use 3D synoptically-dependent
background error modes (SBEMs) within 3DVAR. Current background errors are
`static',  derived via the `NMC' method. An error-breeding cycle is used to
provide SBEMs which are used in 3DVAR via a new control variable and cost
function. Details of the methodology and early results will be presented.
 
 

Recent developments in the ECMWF Ensemble Prediction System\\

J. Barkmeijer

Perturbations used in ensemble forecasting ask for a  careful computation.
One of the conditions they should satisfy is that their statistics
resemble what is known of the analysis error. Prelimenary results on the
EPS performance will be presented of singular vectors computed
with a Reduced Rank Kalman Filter. Such singular vectors are constrained
by the analysis error covariance matrix at initial time. This is achieved
by using the Hessian of the full 4D-Var costfunction. Also the use of different
analyses in the EPS or the construction a so-called consensus analysis
will be discussed.
 
 

Prospects for an operational ensemble Kalman filter

Peter Houtekamer

An ensemble of short-range forecasts can provide the flow-dependent
covariances of the forecast error, needed by the Kalman filter.
The finite ensemble size causes the estimated correlations to be noisy.
To filter small forecast-error correlations associated with remote
observations, a Schur (termwise) product of the covariances of the
forecast error and a correlation function with local support is used.

To solve the Kalman filter equations, the observations are organized
into batches which are assimilated sequentially. The ensemble of
background fields is updated at each step, and thus provides a measure
of the improving quality of the background fields as more and more
batches of observations are assimilated. For each batch, a Cholesky
decomposition method is used to solve the linear system of equations.
Observations from several regions of the globe may be selected for a
single batch, such that information from different regions
has zero correlation due to the Schur product. The linear system
then becomes block diagonal.

A prototype sequential filter has been developed for atmospheric
data assimilation. Application in real time would appear to be
feasible.
 
 

Dynamics and statistics of forecast errors in a quasi-geostrophic model.

Chris Snyder

 Results from QG model.  Time scale for collapse
 of specified initial ensemble to "the attractor", characteristics
 of perturbations on the attractor.  Instantaneous statistics of
 analysis and forecast errors for 3DVAR, in particular influence of
 past dynamics through projection of errors onto leading Lyapnov
 subspace.  Singular vectors for approximate "analysis error covariance
 norm" and their differences from energy SV's.