AOSC 652

AOSC 652: Analysis Methods in Atmospheric and Oceanic Science

Instructors: Ross Salawitch, Jeff Henrikson, Tim Canty, and Walt Tribett

MWF, 12:45 to 1:50 pm, CSS 3426

Fall 2016: 3 units

Primary textbooks to be used throughout the course:

FORTRAN 77 Programming (2nd edition, 1990) by T. M. R. Ellis

Numerical Recipes in FORTRAN 77: The Art of Scientific Computing (2nd edition, 1992) by William H. Press et al.

(available on-line at http://www.nrbook.com/a/bookfpdf.php )

Numerical Analysis by Richard L. Burden and J. Douglas Faires

Specialty textbooks to be used for various portions of the course:

An Introduction to Programming with IDL: Interactive Data Language by Kenneth P. Bowman

Numerical Computing with MATLAB by Cleve Moler

Python Programming and Visualization for Scientists by Alex J. DeCaria

Syllabus

ELMS Page

The following webpage, prepared by AOSC students for AOSC students prior to this course being taught, provides a wealth of information. Please explore the contents as this course evolves:

http://www.atmos.umd.edu/~gcm/useful.html

Please do not be intimidated by the large number of books. The instructors will guide our movement from "book to book" as the course proceeds. Most readings are available on-line (we will provide direct links as the course proceeds). For readings not available on-line, we will provide copies and/or place books "on reserve", so that students may copy the material of interest. Each student will be able to "rent" a copy of the Ellis book, the primary source material for the FORTRAN section of the course, from the instructors (we have acquired a substantial number of copies of this excellent book, which is out of print, enabling it to be used for the course). Finally, to become a proficient practitioner of the "tools of the trade", one must learn how to acquire and implement numerical algorithms from a variety of sources. This course will provide an ample number of examples of such algorithm acquisitions.

1. Course Description

2. Course Schedule

3. Grade Policy

4. Enrollment Requirements

5. Collaboration Policy

6. Office Hours

1. Course Description

        This course will be comprised of “hands on” lessons in the development of code to solve a variety of numerical problems and to visualize data, taught in a Computer Instructional Laboratory in a Linux environment. Will cover topics such as Least Squares Fitting, Spline Fitting, Numerical Integration, Newton-Raphson Minimization, Statistical Methods, Fourier Analysis, Solutions of Ordinary and Partial Differential Equations, and Analysis and Visualization of a variety of satellite observations.
       Typically, the first meeting of each week will be devoted to white board lecture of the weekly topic. The second meeting of each week will be conducted using projection of an active computer monitor, showing code development related to the weekly topic. An assignment will be given at the end of the second meeting. The third meeting of each week will be devoted to reviewing previous homework assignments, clarifying questions about prior lecture notes, and in class work on the assigned problem, which will be due (code plus numerical and/or graphical results) at the start of the first lecture of the following week. Assigned problems will be drawn from real world examples of the various topics in atmospheric and oceanic sciences: for example, an assignment on statistical methods may be based on time series of temperature in Washington, DC; an assignment on Fourier analysis may relate to the Vostok climate record.
        The objective of the course is to provide students experience in the development of code to solve a variety of numerical problems they are likely to encounter during their dissertation research and subsequent careers. The students will be exposed to FORTRAN, IDL, and MATLAB during the various exercises. We will also work with modern data formats such as HDF5 and NetCDF as well as modern graphical tools such as Google Earth. Introduction to these languages will be a natural component of the curriculum.

Back to Contents

2. Schedule

Week	Topic
1) 29, 31 Aug	Intro to Linux & Text Editing
2) 2, 7 & 9 Sept	Intro to FORTRAN: Simple Computations
3) 12, 14, 16 Sept	Intro to Graphics & Analysis of Satellite Measurements of Atmos. Comp.
4) 19, 21, 23 Sept	Getting to know FORTRAN: Data Sorting,Input/Output, and Simple Statistics
5) 26, 28, 30 Sept	Least Squares Analysis, Statistical Regression, and Spline Fitting
6) 3, 5, 7 Oct	Numerical Integration
7) 10, 12, 14 Oct	Root Finding & Newton Raphson Minimization
8) 17, 19, 21 Oct	Introduction to IDL and PYTHON (Students choose one track)
9) 24, 26, 28 Oct	Fourier Analysis
10) 31 Oct & 2, 4 Nov	Multiple Linear Regression; Data Access for Projects
11) 7, 9, 11 Nov	Satellite Data Visualization; HDF & NetCDF
12) 14, 16,18 Nov	Ordinary Differential Equations; File Compression
13) 21 Nov	Partial Differential Equations
14) 28, 30 Nov & 2 Dec	Work on Computationally Based Project
15) 5, 7, 9 Dec	Continue to Work on Computationally Based Project
16) 19 Dec	Project Presentations aka "The Show" or perhaps "The Marathon"

Back to Contents

3. Grade Policy

Grades will be determined based on attendance and participation (10%), weekly assigned projects (70%), and a final project (20%) that will be chosen either by each student or from a list of suggested final projects. Assignments are meant to be turned in on time. There will be a penalty (points per days late) for various assignments, announced as class progresses, that will be imposed unless an arrangement has been made for extenuating circumstances prior to the due date. The last two weeks of class will be devoted to the final project. The final project will emphasize application of the tools of the class to a problem of current interest in the geophysical sciences. We are not planning on having either a mid-term or final exam. Grades will be based solely on the weekly programming assignments, the final project, and class attendance and participation.

Back to Contents

4. Enrollment Requirements

The course is open to students with a basic scientific background. Successful completion of freshman calculus and physics is required. No past programming experience, or experience in Linux or Unix, is required.

Back to Contents

5. Collaboration Policy

For assigned programming projects, you may consult any texts or websites you desire. However, the work you turn in must reflect only your own code development. General course concepts, syntax issues, questions about an assignment, etc. may be discussed with other students. For certain programming assignments, clearly blocks of code are intended to be taken (in some cases copied) from either a text book or some other reference source. Again, we require that the code you turn in reflect your own code development using these resources and not be copied from another student. Details of the University of Maryland Honor Code can be found at the Student Honor Council website as well as the Testudo website.