Abstract
We analized data related to daily cedar pollen counts and meteorological factors for 53 days in the pollinosis season in 1991 by type IT quantification, a concept of multivariant analysis, and established criteria for the daily forecast of airborne cedar pollen counts.
This analysis makes it possible to introduce objectively factors such as weather or wind into the forecasts, and these criteria are useful throughout the pollinosis season.
These criteria were tested for 60 days in 1990 and 47 days in 1992, and good results were obtained (73.3% accuracy in 1990, 78.7% in 1992). Moreover, in the 1993 season forecasts of daily airborne pollen counts with the use of these criteria proved to be 73.8% accurate.
