Metr 4313-001                                                  Statistical Meteorology                                                                         Metr 5313-001

Class meets: MWF 11 – 11:50 AM  NWC 5720

Professor: Michael Richman;  Office: 5646;  Office hours: MWF 12:00 – 1:00 PM or by scheduled appointment

Contact points: Can be reached by phone: 5-1853 or by email: mrichman@ou.edu

Feel free to Email me to schedule a meeting or to Email after office hours (but before midnight, if you want a quick turnaround)

Philosophy behind the course:

This course is designed to teach the interplay between statistics and meteorology.   The relevance to the atmosphere will be examined    through use of meteorological data sets.   There is a large breadth of material in this class and you will learn a considerable number of        tools. It is my hope that you will know what method best fits the question, and the assumptions of each method. Consider that statistics is a field of study, just like meteorology.  The  average  meteorology  student  in  the  SoM  takes  a  minimum  of  14  meteorology classes and a single statistics class.   The average statistics student takes about 10 statistics classes and two science classes         to satisfy her or his general education requirements.  To  present a  small portion  of what a  statistics major would  take, this class  is  presented at a fairly brisk pace. If you don’t understand something in lecture, ask a question during class! I enjoy answering questions.   Otherwise, contact me immediately after class, so you do not fall behind.   I hang out after class for that reason.   Owing to        the amount of material, I don’t have the luxury of using homework as a practice for tests.   By computing with Splus or R, the homework     gives you a new tool and skill set, making you more employable or better preparing you for graduate school.  Homework is designed to  reinforce concepts in the lectures by giving you practice at coding up the concepts. If you don’t see the link between homework and the  lectures, see me.  Tests reflect concepts stressed in the lectures and there is no computer coding on tests.  If you want to get the   most out of the class, put in some time on the homework and study the lecture material for the tests.

Course work & Grading Policies

Books and Handouts:  Lecture notes covering the presentations will be distributed in several packets and posted to D2L.

Book Required: Wilks, Daniel, 2011: Statistical Methods in the Atmospheric Sciences. Third Edition. ISBN: 0123850223.

Optional: Spector, Phil: 1994: An Introduction to S and Splus. ISBN: 0-534-19866-X [also 1 copy available in Bizzell Library] Free Splus primer: Spoetry is on D2L under the Metr 4313 “content” tab [Do NOT print on any SOM printer as it is >400 pp]

Homework: Assigned at least a week prior to due date. Use a Word processor to do all homework and staple it. Late assignments penalty: 50% (if fewer than 7 days late).  I am always available to give advice on homework – see or Email me.

Tests: There are three semester tests and a final (Fri. Dec. 15 at 1:30 – 3:30 PM).  The lowest semester test grade is dropped. Test  date for semester tests is determined by class poll.  Unless you have ≥3 finals on the same day, OU final exam policy has no flexibility.

Computing: All students who do not have a School of Meteorology (SoM) computer account should obtain one from Shawn Riley (NWC 5640). Course work will be reinforced by application of real meteorological data sets using the Splus statistical/ mathematical package. This is available on any of the Metlab Workstations. A free version is available for Windows machines. Those with Macs have three options: (1) Purchase Parallels Desktop for $25 from the ItStore – under “Apple Software” (https://shop.itstore.ou.edu/catalogsearch/result/?q=parallels) and run Splus, (2) Run a Windows Virtual Box (free from Oracle https://www.virtualbox.org/) and run Splus or, (3) Use the “R” statistical package (free at http://www.r-project.org/) as there is an R binary for Macs.  Splus and R are very similar, but Splus is fully supported and more user friendly whereas R is open source.  The  optional text by Spector and the free web link primer (Spoetry — see D2L “content”) have good material on the use of Splus. Data sets will be available online for testing statistical methods and for homework.  I will make myself available for help with Splus/R.

Student feedback and participation: Students are expected to participate actively in a professional manner. In class, students are encouraged to ask questions. Note that there is a grade for participation that reflects your interaction with others in the class and for asking questions.

Graduate students: A research paper is due the day of the final exam.  See me within the first month about your topic.

Grades: Grade percentages will be constructed as shown below. There is almost always a curve in this class. Please don’t ask me “how much is the curve” because that is impossible to assess until the grades are in.

Undergraduate Students                                                                                                     Graduate Students

Homework:       25%                                                                                                          Homework:          20%

Tests:               45%                                                                                                          Tests:                 30%

Final:                25%                                                                                                          Final:                  25%

Participation:       5%                                                                                                          Participation:         5%

Research Paper:   20%

Syllabus: Note that the syllabus topic dates are meant as a guide to the progression of topics in the class and for the topics covered on each test, rather than a rigid list of exactly that date a topic will be presented.

10/18         Meteorological Applications (continuous distributions)

10/20         Bootstrapping to achieve uncertainty estimates on the mean 10/23    Chi-square distributions

10/25         Test 2 [Semester test dates are approximate.  Test is after the packet is finished at a mutually agreeable time]

10/27         Confidence intervals on means with known standard deviation                                         5

10/30         Confidence intervals on means with unknown standard deviations 11/01         Confidence intervals on medians

11/03         Unpaired vs. Paired measurements, t-tests (one and two sample) 11/06       Bootstrapping to achieve confidence intervals on the mean and median 11/08 Comparison of t-tests to permutation tests

11/10         Hypothesis tests

11/13         Hypothesis tests

11/15         Errors and power

11/17         Power curves and type I, II errors, p-values

11/20         Test 3 [Semester test dates are approximate.  Test is after the packet is finished at a mutually agreeable time]

11/22         No class – Thanksgiving 11/24         No class – Thanksgiving

11/27         Introduction to regression                                                                                          6

11/29         Regression

12/01         Univariate regression and diagnostics 12/04   Multiple regression

12/06         Stepwise regression 12/08           Bootstrapping regression

12/15         Scheduled final exam – Comprehensive – 1:30 am – 3:30 PM

Reasonable Accommodation Policy: “The University of Oklahoma will reasonably accommodate otherwise qualified individuals with a disability unless such accommodation would pose an undue hardship, would result in a fundamental alteration in the nature of the  service, program or activity, or would create undue financial or administrative burdens. The term “reasonable accommodation” is used in its general sense in this policy to apply to employees, students and visitors. Student requests for reasonable accommodation should be addressed to the Disability Resource Center, Goddard Health Center, 620 Elm Avenue, Suite 166, (405) 325-3852, TDD (405) 325-4173,

FAX (405) 325-4491, or ods@ou.edu”

Academic Misconduct Policy: Integrity in all aspects of scholarship is essential to the University’s mission. The Academic Misconduct Code sets forth the rights and responsibilities of all students on the Norman Campus regarding academic integrity, and provides the procedures to be followed in cases of suspected misconduct. Academic misconduct is defined as any act which improperly affects the evaluation of a student’s academic performance or achievement. It specifically includes cheating, plagiarism, fabrication, fraud, destruction of property, and bribery or intimidation, as well as assisting others or attempting to engage in such acts. It is the responsibility of each student to be familiar with the definitions, policies and procedures concerning academic misconduct; unfamiliarity with the code alters none of a student’s rights or responsibilities thereunder. The Academic Misconduct Code is printed with the Student Code and is also available at http://integrity.ou.edu/students.html