Recovery of Mesoscale Covariance Using Time-phased Ensembles

Abstract

One of the most important factors that determines meteorological analysis fields in a data assimilation system is the background covariance matrix. Not only do the correlations in this background covariance matrix perform the spatial spreading of information at the observation points throughout the model domain including data-sparse areas, but they also play a decisive role in how to smooth the analysis increments in datadense areas. One of the key issues for data assimilation is the specification of the background error covariance. There are classic methods to estimate the background error covariance mainly based on the study of forecasts started from the analyses, like the NMC method (Parrish and Derber 1992). Their theoretical foundation is now considered rather unclear. In this research, we proposed a new method to estimate background error covariances, namely a time-phased ensemble forecast system, which was recently developed at NOAA’s Global Systems Division (GSD) of the Earth System Research Laboratory (ESRL). In this research, we will examine how well this newly structured background error covariance can capture the flow dependent mesoscale features. We will also compare the covariance structure obtained from the time-phased ensemble method with that from NMC method.