Extended Abstract (36K)P10.1
Utilizing site-based data mining in national ceiling and visibility forecasting
Gerry Wiener, NCAR, Boulder, CO; and P. Herzegh, R. E. Bateman, and B. Chorbajian
Creating more accurate predictions of ceiling and visibility over the 0-12 hour time frame is of significant importance to aviation safety and efficiency. In this paper we will examine how current data mining techniques can be applied to the ceiling and visibility prediction problem. We will present results comparing data mining prediction against persistence in the 0-12 hour time frame for different metar sites in the continental United States. We will then discuss how data mining techniques can be incorporated in a national ceiling and visibility detection and prediction system.
Poster Session 10, Ceiling and Visibility, Poster Session
Thursday, 7 October 2004, 3:00 PM-4:30 PM
Next paper