EMPIRICAL ORTHOGONAL FUNCTIONS AND NORMAL MODES.
Academic Article
Overview
Identity
Additional Document Info
View All
Overview
abstract
An attempt to provide physical insight into the empirical orthogonal function (EOF) representation of data fields by the study of fields generated by linear stochastic models is presented. In a large class of these models, the EOFs at individual Fourier frequencies coincide with the orthogonal mechanical modes of the system - provided they exist. The precise mathematical criteria for this coincidence are derived and a physical interpretation is provided. A scheme possibly useful in forecasting is formally constructed for representing any stochastic field by a linear Hermitian model forced by noise.