Title: Application of weak constraint dual formulation 4D-Var to the California Current System

Authors: Polly Smith (University of California, Santa Cruz CA)
Andrew Moore (University of California, Santa Cruz CA)

Model errors are a significant source of uncertainty in analysis and forecast systems. The character, structure, and properties of these errors are generally poorly known; it is common practice in both oceanographic and meteorological data assimilation applications to neglect these errors and assume that the numerical model is perfect.

Here we describe work to attempt to quantify the magnitude and structure of model errors in the Regional Ocean Modeling System (ROMS) configured for the California Current System (CCS). It is known that ROMS estimates of sea surface temperature along the California coast are in error, particularly during the peak upwelling season. Independent analyses of the surface forcing data used to drive ROMS suggest that this tendency is related to deficiencies in the model rather than errors in the data. Previous work with the ROMS strong constraint 4D-Var system found that to compensate for this model error and produce a better fit to the observations the data assimilation acts to reduce the rate of upwelling by adjusting the alongshore surface wind stress. Clearly, this effect is undesirable; an appropriate solution is to account for the model error during the assimilation process by employing a weak constraint approach.

Practical implementation of weak constraint 4D-Var requires prescription of the prior model error covariance and offers a considerable challenge. Due to the large dimension of the problem, weak constraint 4D-Var is only possible in ROMS when working dual form. Here we will discuss issues related to application of the ROMS weak constraint dual form 4D-Var system to the CCS. We will present results from a series of experiments that aim to identify the nature of the CCS model errors and thus inform the design of suitable prior model error covariance structures.

GMAO Head: Michele Rienecker Global Modeling and Assimilation Office NASA Goddard Space Flight Center

Curator: Nikki Privé Last Updated: May 27 2011