DYNAMIC DATA ASSIMILATION: A LEAST SQUARES APPROACH

DYNAMIC DATA ASSIMILATION: A LEAST SQUARES APPROACH John H. Lewis, S. Lakshmivarahan, and Sudarshan Dhall, 2006, 654 pp., $150.00, hardbound, Cambridge University Press, ISBN 978-0-521-85155-8

I believe this is the first book about data assimilation with so many theories and algorithms put together systematically.

Audience: Data assimilation researchers and developers will find this work to be a very useful handbook on their desks.

This is a perfect textbook for postgraduate/ postdoctorate-level courses. However, I would advise students to first attempt an introductory-level data assimilation course.

Strengths: The strengths of the book are the theoretical overviews of the data assimilation issues and the well-organized chapters from simple examples to the most advanced topics.

Weaknesses: For atmospheric and oceanic data assimilation systems, the most important issues are background covariance modeling, dynamic balancing, nonlinear observation operators, computation cost, and nonlinearity (non-Gaussianity). I think these topics could be emphasized more in this book, in particular in the later chapters.

Illustrations: The diagrams for different reviews and comparisons are very useful.

Bottom line: This book provides readers with a good mathematical framework for data assimilation, with all important proofs and deviations.

I recommend the book for data assimilation system developers and colleagues who work with data assimilation research and applications.

- XIANG-YU HUANG

[Author Affiliation]

Xiang-Yu Huang is the lead scientist of the WRF variational data assimilation system and the director of the data assimilation testbed center (DATC) at NCAR in Boulder, Colorado.