Professor Dennis L. Hartmann
710 Atmospheric Sciences/Geophysics Building
543-7460 dhartm@uw.edu
Class Meets MWF 2:30-3:20 in 310c ATG
You must consult the Class Web Page at: www.atmos.washington.edu/~dennis/552.index.html
Purpose of the Course:
Objective analysis is here defined to be the extraction of information from data using objective, as opposed to subjective, methods, most often via a computer program. The goals of this course are to provide a working knowledge of the basic methods of objective analysis of meteorological, oceanographic, and related data, and to provide a critical facility for evaluating published studies utilizing these techniques. The topics concentrate on techniques for extracting information from data directly, such as compositing, time series analysis, singular value decomposition, principal component analysis, and filtering. If time allows, we may touch on some new topics, such as wavelet analysis or more general statistical modeling procedures. We will try to tailor the course to the needs and interests of the students in the class, while encouraging a broadening of outlook and experience.
Method of the Course:
¥ Class Lectures: About three classes out of four will consist of lecture/discussion led by me. I will provide notes in advance for most of this material.
¥ Reading assignments: In addition to class notes, students may wish to consult a standard statistics textbook, when appropriate. Mathematical and computer books about the techniques are also available. A few examples are:
Statistical Analysis in Climate Research, von Storch and Zwiers, 1999, U.Cambridge Press
Statistical Methods in the Atmospheric Sciences, Daniel S. Wilks, 1995/2005, Academic Press.
If you need more elemental background reading, I recommend
the SchaumÕs outline series books on Linear Algebra, Probability and Statistics and Probability, Random Variables and Random Processes. ThatÕs
three separate books. You can get
them at the bookstore, or at Amazon for about $12 each.
I will not require that you buy or read these books, but merely indicate that they are available if you need them. Additional references to textbooks and published papers are given in the class notes.
¥ Class Presentations by Students: We will read and critique papers from the literature that use the techniques we are studying. Discussions will be led by the graduate students, who will take about 20 minutes to introduce the main points of the paper, followed by general discussion. Students will sometimes present and discuss homework.
¥ Homeworks: There will be approximately weekly homeworks. Many of these will require you to use a workstation running Matlab.
¥ Examinations: A mid-term and a final will be given
¥ Grading:
Homeworks 35%
Class Presentations 15%
Mid-Term 20%
Final 30%