Extended Abstract (96K)J5.5
Automated forecasting of ceiling and visibility for aviation: exploration of a data mining-based forecast method
Paul H. Herzegh, NCAR, Boulder, CO; and G. Wiener, R. Bateman, and B. Weekley
Recent development of an exploratory automated ceiling and visibility (C&V) analysis and forecast system for aviation applications in the continental U.S. has utilized expert system methodology to merge numerical and observational inputs in the synthesis of forecast products out to twelve hours. In this paper we outline the use of current forecast techniques used in this system (RUC, ETA, and persistence), and we explore the potential roles of other techniques being explored for future use. In particular, we discuss the new development of a short-term observations-based forecast method that utilizes forecast rulesets derived from data mining of long-term observational records at selected sites. We outline encouraging early results of the method, current challenges associated with performance improvement, and the potential role of this data mining approach in the automated forecast system under development.
Joint Session 5, AI Applications with a Nowcasting Flavor (Joint between the Fourth Conference on Artificial Intelligence and the 21st International Conference on Interactive Information and Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology)
Tuesday, 11 January 2005, 9:00 AM-11:30 AM
Previous paper Next paper
Recorded presentation