The Objective Analysis of Atmosphric Structure

Abstract

An Objective Analysis Model is presented which will handle the four-dimensional data assimilation problem where observations are randomly situated in both space and time. The structure functions of both the desired atmospheric signal and the system plus geophysical noise are modelled and used to produce the required weight functions. Modelling of the dynamics of the atmosphere is incorporated through the use of linear constraints in the hypothesis testing. The resulting analysis is a maximum likelihood, least squares (in the population sense) production and has associated fields of confidence limits.