Extended Abstract (2.3M)10A.2
Local Predictability of the Performance of an Ensemble Forecast System
Elizabeth A. Satterfield, University of Maryland, College Park, MD; and I. Szunyogh
In this paper, the spatio-temporally changing nature of atmospheric predictability is assessed in a reduced resolution version of the model component of the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS). The Local Ensemble Transform Kalman Filter (LETKF) data assimilation system of the University of Maryland is used to assimilate observations in three steps, gradually adding more realistic features to the observing network. In the first experiment, randomly placed, noisy, simulated vertical soundings, which provide 10% coverage of horizontal model grid points, are assimilated. Next, the impact of an inhomogeneous observing system is introduced by assimilating simulated observations in the locations of conventional observations. Finally, observations of the real atmosphere are assimilated.
In addition to providing imperfect state estimates (analyses), the LETKF also provides analysis perturbations, which serve as initial conditions for an ensemble of forecasts. Ensembles of forecasts are used to provide both a mean forecast and an estimate of forecast uncertainty. The ability of the ensemble spread to provide an accurate estimation of forecast uncertainty is itself flow dependent. We show that, independent of experiment, large forecast error in the extra-tropics leads to an increased likelihood that the ensemble can fully capture the space in which forecast errors evolve. We explain this behavior using the E-dimension and explained variance diagnostics. Additionally, we explore the role that observation density plays in the ability of the ensemble to measure the importance of different state space directions.
Session 10A, Advanced Methods for Data Assimilation—II
Wednesday, 14 January 2009, 1:30 PM-2:30 PM, Room 130
Previous paper Next paper
Recorded presentation