INITIALIZATION OF PRIMITIVE-EQUATION MODELS FOR NUMERICAL WEATHER PREDICTION.
Abstract
Numerical weather prediction models of the primitive equation type generate spurious gravity-inertial waves, detrimental to the forecast, if the initial wind and height fields are not in proper balance. In this thesis a new method of obtaining balanced initial fields is developed and demonstrated. The method uses calculus of variations to blend the observed wind and height information to yield an analyzed wind stream function through the use of wind vorticity and two forms of the balance equation. The contribution of the observed winds and heights to the analyzed stream function varies with wave length in a prescribed manner. This analyzed stream function is inserted into a complete form of the balance equation which is then solved for a new height field. With the analyzed stream function and new height fields some form of the omega-equation can be solved. The particular form of the omega-equation used depends upon the type of forecasting model. From omega and/or divergence the velocity potential can be solved. Finally, from the stream function and velocity potential fields a new wind field is computed. The new wind and height fields constitute balanced initial data for primitive equation numerical weather prediction models. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1964
- Accession Number
- AD0625348
Entities
People
- Warren M. Washington
Organizations
- Pennsylvania State University