Geography 820.01: Seminars in Physical Geography: Numerical Simulation in Microclimatology
Instructor: John Arnfield
The course is built around the use of numerical simulation techniques to investigate small-scale climatic systems, ranging from the very small (e.g., a leaf) up to the scale of the planetary boundary layer. Students will expand their substantive knowledge of microclimatology (beyond that encountered in 622.01) but will do so by setting up simulation models that can then be used to experiment with the modelled system and to answer “what if…?” type questions. Students will also learn some fundamental numerical methods and improve their computer programming skills.
In the early part of the course, simulations are based on concepts from 622.01. Students will also be constructing program modules that are later used to construct more complex models. Later in the course, the class will read some research literature in microclimatology and boundary-layer climatology and will seek to use the ideas presented in those papers to simulate the phenomena described.
This course will focus on micro-scale climatological phenomena but will also be of interest to those whose interests lie at larger scales since many ideas covered are directly related to the ways in which surface processes are handled in meso-scale and global scale models (e.g. GCMs and numerical forecasting models).
This class is run primarily on a “hands-on” basis. There is some formal classroom instruction but most of the time is spent in an informal “workshop” setting defining problems, discussing their solution using simulation strategies and creating simulation code.
The prerequisites for this course are successful completion of Geography 622.01 (Boundary Layer Climatology), and knowledge of an appropriate general computer programming language, like FORTRAN or C, under Windows, DOS, Unix or mainframe/MVS environments.
It will be assumed that students have a good recollection of and access to the concepts, theories and terminologies covered in 622.01.
This course is available either for 3 or 5 credit hours.
For the 3 credit hour option, the final grade depends solely on the work performed on individual and group programming and simulation projects throughout the quarter. In the case of group projects, a grade assigned for a piece of work will be allocated to every member of the group. Scoring will be done on a letter-grade basis (rather than using a numerical score) and the course grade will be evaluated using the “grade-point-average” for all components of the course. Reduction of grades below “A” will result if students (a) fail to complete the project as defined, (b) fail to meet the deadline for completion, (c) produce code that generates erroneous results, (d) produce code that does not meet the criteria for good programming discussed in the course, and/or (e) produce code that did not represent a valid simulation representation of the theory presented in a paper assigned as the basis for the project.
For the 5 credit hour option, students are, in addition, required to undertake an individual project relevant to the course (and, it is hoped, to their research interests) which will necessitate creation of a functioning simulation model, a brief “user manual” for the code and a paper that describes the nature of the question the model is intended to investigate, its theoretical basis, the numerical simulation strategies used and some simulation results exploring the question. Presentation of this project to the class in the final week is encouraged.
There are no exams in this course.
Reference Materials and Software Resources
There is no text for the course. Some sources that might be found useful, besides the 622.01 text, are:
- Etter, D.M. Fortran 77 with Numerical Methods for Engineers and Scientists. Benjamin/Cummings Publishing Company.
- Gerald, C.F. & P.O. Wheatley. Applied Numerical Analysis. Addison Wesley.
- Scheid, F. Theory and Problems of Numerical Analysis. Shaum’s Outline Series, McGraw-Hill.
There are many software packages available that could be used to complete some projects that are undertaken in this seminar (e.g., Matlab, Mathcad). Students should not make use of these programs. While such software uses many of the same techniques employed in this class, students are encouraged to achieve a more intimate interaction with the material than is possible in the “black box” environment they provide.
Topics To Be Covered
Given the informality of this course, no attempt is made to lay down a rigid outline of topics to be covered. In general, after the first week when students will be developing some useful all-purpose modules for later application, the course will be structured around substantive topics rather than methods.
Some types of problems covered in the seminar include:
- Radiation calculations in urban and alpine areas.
- Pollution dispersion models.
- Energy budgets of leaves, organisms etc.
- The surface energy budget (increasing complexity from airless planets to a realistic Earth case) and SBL wind, temperature and humidity structure.
- Temperature evolution in substrates.
- Local-scale advection.
- Frost prediction.
- Simulations of special surface types, like snow, water etc.
The course outline is open to modification based on the suggestions from and needs of class members, the availability of time during the quarter and ideas that occur to the instructor and students as the class progress.