耳鼻咽喉科臨床
Online ISSN : 1884-4545
Print ISSN : 0032-6313
ISSN-L : 0032-6313
臨床
福島市におけるスギ花粉飛散予測モデルの構築
室野 重之佐藤 廣仁尾股 千里野本 美香
著者情報
ジャーナル 認証あり

2021 年 114 巻 1 号 p. 27-33

詳細
抄録

Prediction of pollen dispersion, especially the first day of pollen release and the pollen counts during season, is important for medical care and self-care of patients with Japanese cedar pollinosis. Based on meteorological data collected from 1998 to 2017, we attempted to establish a prediction model for the first day of pollen dispersion. We found that the cumulative temperature over 0 degrees Celsius beginning January 21 seemed to be the most effective model, with the average of 253.9 during this period is predictive of the first day of pollen dispersal. Application of this model to 2018 and 2019 showed an error between the predicted and actual first days of pollen release was 1 day for both years. A prospective observation in 2020 revealed an error of 2 days. We also attempted to establish a prediction model of the total pollen counts during the season. Multiple regression analyses demonstrated that four factors, namely, the average temperature, the sunshine duration and precipitation in previous July, and the total pollen counts of the previous year, as well as three of the above factors, without precipitation, allowed the construction of similarly effective prediction models. Retrospective investigation of the correlation between the predicted and actual total pollen counts demonstrated a high correlation coefficient. Furthermore, a prospective observation in 2020 showed good concordance with the predicted total pollen counts. The proposed prediction models for Japanese cedar pollen release and counts are acceptable, although further improvements would be needed.

著者関連情報
© 2021 耳鼻咽喉科臨床学会
前の記事 次の記事
feedback
Top