NYCCS/ITPA-MSRC Seminar

Istvan Szunyogh

Department of Atmospheric Science
Institute for Physical Science and Technology and Department of Atmospheric and Oceanic Science,
University of Maryland

“Applications of an Ensembl-based Kalman Filter to the Assimilation of Observations and the Assessment of Predictability”

Thursday, February 15, 2007
Endeavour Hall, Room 120
3:00-4:14 PM

The accuracy and computational efficiency of a four-dimensional implementation of the Local Ensemble Transform Kalman Filter (LETKF) data assimilation scheme is investigated on a state-of-the-art operational numerical weather prediction model. The model selected for this purpose is a reduced resolution (T62 and 28 vertical levels) implementation of the 2004 version of the Global Forecast System (GFS) of the National Centers for Environmental Prediction (NCEP). The parallel computer implementation of the LETKF described here provides good scaling of the of the wall-clock computational time, with the number of processors, up to about 60 processors. This parallelization strategy is expected to scale well too many more processors as the model resolution and the number of observations are
increased.

All atmospheric observations that were operationally assimilated by NCEP in 2004, except for satellite radiances, are assimilated with the LETKF. The performance of the LETKF is evaluated by comparing its performance to that of the Spectral Statistical Interpolation (SSI) system, which is the operational global data assimilation system of NCEP. For the selected set of observations, the LETKF analyses are more accurate than the SSI analyses in the SH extratropics and are comparably accurate in the NH extratropics and Tropics.

The spatio-temporally changing nature of predictability is also studied in the NCEP GFS. Atmospheric predictability is assessed in the perfect-model scenario for which forecast uncertainties are entirely due to uncertainties in the estimates of the initial states. Uncertain initial conditions (analyses) are obtained by assimilating simulated noisy vertical soundings of the ``true"
atmospheric states with the LETKF.