[Prospective study of pollen dispersal prediction and identifying the usefulness of different parameters].

Background: Birch pollen is the major pollen allergen in Hokkaido, Northern Japan. We reported a Betula masting model based on the resource budget model hypothesis. In addition to weather conditions, cumulative hours of sunlight and mean temperature from May to July of the previous year, this model used the amount of annual pollen dispersed in previous and penultimate years as a parameter based on data from 1990 to 2000.
Objective: We compared the predicted and observed amount of pollen dispersed for 3 years from 2001 to 2003 and evaluated the usefulness of each parameter in this model.
Methods: Birch pollen was measured using the Durham sampler at the Hokkaido University Graduate School of Medicine Research Institute in Sapporo.
Results: The difference between predicted and observed amounts of pollen dispersal was about 200-500 grains cm(-2). The annual pollen dispersed in the previous year was found to be the most useful parameter.
Conclusion: This model is useful in predicting whether the amount of birch pollen will be less than average, about average, more than average, or much more than average.