Department of Atmospheric, Oceanic and Earth Sciences

The students will gain a basic understanding of the weather modeling process and how errors in the numerical weather prediction models affect the solution. The students will become familiar with the history of the numerical weather prediction and its evolution before providing step-by-step descriptions of the various model components. The well-know sources of errors will be described: errors in the initial conditions, in the numerical approximations, in the physical-processes parameterizations, and in the boundary conditions. The discussion on the ensemble methods will enable the students to understand the uncertainty in the numerical weather forecast. The students will become familiar with programing using FORTRAN language. The course will give the students a comprehensive of the techniques used in the numerical weather prediction and the skill that we can expect from these models.

1. Introduction
   1.1  The early days of numerical weather prediction
   1.2  The beginning of modern numerical weather prediction
   1.3  Numerical weather prediction today
   1.4  Developments in computing
2. What is a numerical weather prediction model?
   2.1  Weather prediction equations
   2.2  The concept of physical parameterizations
   2.3  Data assimilation
3. Grid point models
   3.1  The finite difference methods
   3.2  The common used grids and their properties
4. Spectral models
   4.1  Spectral methods
   4.2  Spectral methods on the sphere
5. Vertical Discretization
   5.1  Hydrostatic/non-hydrostatic models
   5.2  Sigma vertical coordinate
   5.3  Hybrid vertical coordinate
6. Time integration schemes
   6.1  Non-iterative schemes
   6.2  Iterative schemes
7. Operational Forecasting
   7.1  The forecasting process
   7.2  Ensemble forecasting
   7.3  Forecast verification
8. Examples of NWP models
   8.1  The North American Mesoscale (NAM) forecast System
   8.2  The Global Forecast System (GFS)

Coiffier, Jean 2012: Fundamentals of Numerical Weather Prediction, Cambridge Press, ISBN 978-1-107-00103-9

The class meets twice a week. Every week, the first class time is dedicated to NWP topics and the second class time is dedicated to programming.

Periodic homework is assigned and is due at the start of the class indicated. There will be one exam during the semester and a final. Exams are designed to test basic concepts and are closed books and closed notes. A class project will be assigned after the midterm exam. The project is designed to test programming concepts. The project is due on the day of final. The project will be evaluated based on an oral presentation.