Premium
Use of an OSSE to evaluate background‐error covariances estimated by the NMC method
Author(s) -
Errico Ronald M.,
Privé Nikki C.,
Gu Wei
Publication year - 2014
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1002/qj.2384
Subject(s) - data assimilation , reciprocal , scaling , covariance , mathematics , statistics , context (archaeology) , usable , meteorology , computer science , physics , philosophy , linguistics , geometry , world wide web , paleontology , biology
The NMC method has proven utility for prescribing approximate background‐error covariances required by variational data assimilation systems. Here, untuned NMC method estimates are compared with explicitly determined error covariances produced within an OSSE context by exploiting availability of the true simulated states. Such a comparison provides insights into what kind of rescaling is required to render the NMC method estimates usable. It is shown that rescaling of variances and directional correlation lengths depends greatly on both pressure and latitude. In particular, some scaling coefficients appropriate in the Tropics are the reciprocal of those in the Extratropics. Also, the degree of dynamic balance is grossly overestimated by the NMC method. These results agree with previous examinations of the NMC method which used ensembles as an alternative for estimating background‐error statistics.