On the Optimal Specification of the Initial State for Deterministic Forecasting

Abstract

In the design of a rational global observing system, one of the most important questions is ?What are the best variables (wind, temperatures, etc.) to observe for the optimal specification of the initial conditions for numerical forecast models?? The slow manifold theory of Leith (1980), together with geostrophic adjustment theory, can be used to analyze the information content of different variables under different conditions. This type of analysis was applied to the shallow-water equations to demonstrate that under most conditions the rotational wind components optimally specify the initial state.