Formal, brief syllabus
More detailed, evolving syllabus
What is a "good" (or "bad") weather forecast?
What's the relationship between the quality and value of a forecast?
Evaluating forecasts is essential in order to improve their quality or value. Anyone involved in making or using weather forecasts can benefit from an understanding of the basic principles of forecast evaluation. From routine day-to-day weather to rare events, getting information about the relationship of forecasts to events can be a challenging problem. This course is designed to present an overview of the primary methods by which forecasts are evaluated in order to allow students to evaluate forecast systems properly. Issues such as stratification of forecasts and appropriate upper and lower bounds on skill will be addressed in the context of dichotomous, polychotomous, and continuous forecast elements. In addition, the basics of decision analysis will be presented. This is a critical topic for anyone who will ever interact with forecast users or who will forecast in the commercial sector, where relationships with clients may depend upon the ability to convey useful information about the forecast. Examples will be drawn from weather forecasting, as well as other disciplines where forecasts are made. It is hoped that everyone leaving this class will have a basic understanding of the power and complexity of forecast evaluation and its application.
Required texts: None.
Questions? E-mail to Harold Brooks