Short Term Forecasting of Cloud and Precipitation along Communication Paths.
Abstract
This report presents candidate techniques for forecasting precipitation and clouds along arbitrary paths. Mapping the troposphere into a three-layer model representing the lower, middle, and upper regions is suggested. The data to be used are radar reflectivity factor and visible and infrared satellite imagery. In combination, they provide for an effective description of the storm environment. Features to be tracked and forecast are discrete contours of these data which would be generated through threshold and field gradient methods. Efficient storage and analysis requires an efficient method for description of these features. Use of field attributes is suggested. Examples of candidate methods for generating the various attribute lists, such as simple icon and positional chain coding up through complex polynomial and vortex fitting are presented. Contour translation and rotation will be determined through use of cross correlation. Methods for detecting changes in individual contours such as in size and shape, are directly linked to the attribute set chosen, and examples are provided. Due to the rapid data update rate of 5 to 10 min and the large volume of data to be processed, the use of simple and efficient forecast tools is suggested. Various candidate techniques, such as simple linear exponential filtering, are discussed. Finally, the general philosophy for solution of the problem and a general outline of the entire analysis process is presented. Keywords: Atmosphere models; Weather forecasting; Storm structure; Atmospheric precipitation.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 31, 1985
- Accession Number
- ADA169744
Entities
People
- Alan R. Bohne
- F. I. Harris
Organizations
- Air Force Research Laboratory