Teaching numerical weather prediction and data initialization using Richardson's model: An internet-based utility

As technology grows and more weather data become available, the use of numerical weather prediction and data initialization is becoming an increasingly integral aspect of the weather forecasting process. A critical question is, how do we bring these cutting-edge uses of weather data to the classroom?

The process of data initialization removes the high frequency components in initial weather data that would otherwise disrupt the numerical weather forecast product. However, in atmospheric science oriented departments across the country, the instruction of data initialization is rarely a focus due to limiting factors such as a lack of visual examples and interactive models for instructional use.

The internet-based utility discussed in this poster solves both of these limiting factors by providing a student-friendly interface and choices from a variety of examples. The examples in the utility are developed from the first numerical weather forecast calculated by Lewis Fry Richardson in 1922 and a revised process created by Peter Lynch in the early 1990s which incorporates various initialization techniques with Richardson's methods.

By using various combinations of initialization techniques, the student user is able to see the effects of initialized data on a forecast produced by a model against non-initialized data. Instructors of atmospheric science classes will be able to use the product in a lab environment as an aid to the discussion of data initialization from class lecture. As a result, students will increase their understanding of data initialization and numerical weather prediction. In this way, the early history of weather forecasting can be used to bring 21st-century applications of weather data into our classrooms.