Date of Award
Master of Science
Department of Electrical and Computer Engineering
Michael L. Talbert, PhD
Air Force pilots require dependable weather reports so they may avoid unsafe flying conditions. In order to better gauge the accuracy of its weather products, Air Force Weather has established the requirement for an Air Force-wide automated weather metrics program. Under the guidelines for this program, forecasts will automatically be compared to observed weather to determine their accuracy. Statistics will be collected in the hopes of determining forecast error trends that can be corrected through education and training. In order for the statistical data produced by such a program to draw reliable conclusions about forecast accuracy, however, the correct format of the raw forecasts and observations must be ensured before the reports are disseminated. Beyond a simple check for typographical errors, however, the system must also have weather domain knowledge to understand when the input data content does not fit the context of the report, even though it has been formatted properly. This thesis proposes the application of an intelligent user-interface "critic" advice system, to ensure not only correct product format and provide content quality control, but to collaborate with and advise the forecaster or observer during product generation, with the ultimate goal of producing more accurate weather products.
DTIC Accession Number
Leon, Darryl N., "An Intelligent User Interface to Support Air Force Weather Product Generation and Automated Metrics" (2000). Theses and Dissertations. 4821.