Abstract
The currently available model-based global data sets of atmospheric circulation are a by-product of the daily requirement of producing initial conditions for numerical weather prediction (NWP) models. However, due to limitations in the early data assimilation systems and inconsistencies caused by numerous model changes, the available model-based global data sets may not be suitable for studying global climate change. A comprehensive analysis of gobal observations based on a four-dimensional data assimilation system with a realistic physical model should be undertaken to integrate space and in situ observations to produce internally consistent, homogeneous, multivariate data sets for the earth's climate system. The concept is equally applicable for the atmosphere, the oceans, and the biosphere. -from Authors
