SINGULAR VECTORS AND ESTIMATES OF THE ANALYSIS-ERROR COVARIANCE METRIC

An important ingredient of ensemble forecasting is the computation of initial perturbations. Various techniques exist to generate initial pe rturbations. All these aim to produce an ensemble that at initial time , reflects the uncertainty in the initial condition. In this paper a m ethod for computing singular Vectors consistent with current estimates of the analysis-error statistics is proposed and studied. The singula r-vector computation is constrained at initial time by the Hessian of the three-dimensional variational assimilation (3D-Var) cost function in a way which is consistent with the operational analysis procedure. The Hessian is affected by the approximations made in the implementati on of 3D-Var; however, it provides a more objective representation of the analysis-error covariances than other metrics previously used to c onstrain singular vectors. Experiments are performed with a T21L5 Prim itive-Equation model. To compute the singular vectors we solve a gener alized eigenvalue problem using a recently developed algorithm. it is shown that use of the Hessian of the cost function can significantly i nfluence such properties of singular vectors as horizontal location, v ertical structure and growth rate. The impact of using statistics of o bservational errors is clearly visible in that the amplitude of the si ngular vectors reduces in data-rich areas. Finally the use of an appro ximation to the Hessian is discussed.