Abstract
Abstract. This paper compares the performance of the Local Ensemble Transform Kalman Filter (LETKF) with the Physical-Space Statistical Analysis System (PSAS) under a perfect model scenario. PSAS is a 3D-Var assimilation system used operationally in the Goddard Earth Observing System Data Assimilation System (GEOS-4 DAS). The comparison is carried out using simulated winds and geopotential height observations and the finite volume Global Circulation Model with 72 grid points zonally, 46 grid points meridionally and 55 vertical levels. With forty ensemble members, the LETKF obtains analyses and forecasts with significantly lower RMS errors than those from PSAS, especially over the Southern Hemisphere and oceans. This observed advantage of the LETKF over PSAS is due to the ability of the 40-member ensemble LETKF to capture flow-dependent errors and thus create a good estimate of the evolving background uncertainty. An initial decrease of the forecast errors in the Northern Hemisphere observed in the PSAS but not in the LETKF suggests that the LETKF analysis is more balanced.
Highlights
Three-dimensional variational data assimilation (3D-Var) was adopted for the first time in operational data assimilation at the National Centers for Environmental Prediction (NCEP) with the Spectral Statistical Interpolation (SSI) scheme in 1991 (Parrish and Derber, 1992), and has been proven to be considerably more accurate than the scheme it replaced (Optimal Interpolation, OI)
We evaluate the performance of both Physical-Space Statistical Analysis System (PSAS) and the Local Ensemble Transform Kalman Filter (LETKF) by computing the Root Mean Square (RMS) errors for both the analyses and the forecasts
The difference is especially apparent on 12 February, when PSAS has a large spike in the RMS error
Summary
Three-dimensional variational data assimilation (3D-Var) was adopted for the first time in operational data assimilation at the National Centers for Environmental Prediction (NCEP) with the Spectral Statistical Interpolation (SSI) scheme in 1991 (Parrish and Derber, 1992), and has been proven to be considerably more accurate than the scheme it replaced (Optimal Interpolation, OI). System (PSAS), a 3D-Var scheme developed at NASA1, differs from other 3D-Var schemes, such as the NCEP SSI and the 3D-Var scheme of European Centre for Medium-Range Weather Forecasts (ECMWF) (Courtier et al, 1998), mainly in that it is formulated directly in physical space, rather than in spectral space (Cohn et al, 1998) It was the operational data assimilation system in the Goddard Earth Observing System Data Assimilation System (GEOS-4 DAS) (Bloom et al, 2005). Though adapted from the LEKF, the computational cost of the LETKF is significantly lower because it solves the analysis equations in the subspace spanned by the ensemble members without using singular value decomposition This computational efficiency, simplicity of implementation (e.g., it does not require the adjoint of the observational operator and the adjoint of the model dynamics) and its accuracy make the LETKF a appealing ensemble Kalman filter scheme.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
