Chapter 10
OCEAN DATA ASSIMILATION
USING SEQUENTIAL METHODS
BASED ON THE KALMAN FILTER
From theory to practical implementations
Pierre Brasseur
CNRS/LEGI, Grenoble, France
Abstract
The main purpose of this chapter is to review the fundamentals of the
Kalman Filter for ocean data assimilation and to expose the basic ingredients of practical assimilation algorithms developed for applied ocean
research and operational forecasting, focusing mainly on high-resolution
applications. Important implementation issues such as the reduction
in dimensionality of the estimation problem, the simplification of the
schemes based on static error covariances, the formulation of low-rank
filters, the problem of consistency verification, and the concepts of adaptivity and incremental analysis updating will be addressed using scientific and operational examples. Finally, the discussion will conclude
with a number of key questions related to the assimilation challenges of
the next decade.
Keywords: Data assimilation, mesoscale ocean circulation, reduced-order Kalman
filters, statistical estimation, ocean forecasting.
1.
Introduction
Operational ocean prediction systems are being developed with a variety of objectives in mind, such as ocean current hindcasting and shortrange forecasting, estimation of the thermodynamic state of the ocean
for seasonal and climate predictions, and production of retrospective
analyses of the changing ocean through the merging of models and data.
In addition to simply combining model estimates with observations,
data assimilation provides a means to systematically compare theoretical
:
271
E. P. Chassignet and J. Verron (eds.), Ocean Weather Forecasting, 271-316.
© 2006 Springer. Printed in the Netherlands.
OCEAN DATA ASSIMILATION
USING SEQUENTIAL METHODS
BASED ON THE KALMAN FILTER
From theory to practical implementations
Pierre Brasseur
CNRS/LEGI, Grenoble, France
Abstract
The main purpose of this chapter is to review the fundamentals of the
Kalman Filter for ocean data assimilation and to expose the basic ingredients of practical assimilation algorithms developed for applied ocean
research and operational forecasting, focusing mainly on high-resolution
applications. Important implementation issues such as the reduction
in dimensionality of the estimation problem, the simplification of the
schemes based on static error covariances, the formulation of low-rank
filters, the problem of consistency verification, and the concepts of adaptivity and incremental analysis updating will be addressed using scientific and operational examples. Finally, the discussion will conclude
with a number of key questions related to the assimilation challenges of
the next decade.
Keywords: Data assimilation, mesoscale ocean circulation, reduced-order Kalman
filters, statistical estimation, ocean forecasting.
1.
Introduction
Operational ocean prediction systems are being developed with a variety of objectives in mind, such as ocean current hindcasting and shortrange forecasting, estimation of the thermodynamic state of the ocean
for seasonal and climate predictions, and production of retrospective
analyses of the changing ocean through the merging of models and data.
In addition to simply combining model estimates with observations,
data assimilation provides a means to systematically compare theoretical
:
271
E. P. Chassignet and J. Verron (eds.), Ocean Weather Forecasting, 271-316.
© 2006 Springer. Printed in the Netherlands.