サクラの満開日を広域で予測するモデル

書誌事項

タイトル別名
  • A universal model for predicting the full bloom date of Japanese flowering cherry
  • サクラ ノ マンカイビ オ コウイキ デ ヨソク スル モデル

この論文をさがす

抄録

Full bloom dates of the ‘Someiyoshino’ flowering cherry (Prunus×yedoensis Matsum.) and hourly temperature data at 46 observatories across Japan were used to develop a universal model for predicting the phenology without site specific calibration. The prediction model consisted of three sub models for endodormancy and ecodormancy stages respectively, as well as their transient dormancy stage. The chill unit model of the endodormancy stage was expressed as a concave-low parabolic equation with a maximum chill unit value of 1.0 optimal temperature. In ecodormancy, the effective temperature curve for bud growth to accumulate the growing degree hours (GDH) was expressed as a sigmoid function at temperatures exceeding 0℃. Usually temperatures below 0℃ are excluded for GDH accumulation; however, in this model a positive constant was included assuming the promotive effects on bud growth under successively higher temperatures. Since the transition from endodormancy to ecodormancy should be a gradual response, the GDH was weighted by another sigmoid curve of growth competence which determined the growth ability of buds to forcing temperatures. The growth competence was zero during the initial stage of endodormacny and started gradually climbing before the end of endodormancy. The model parameters were evaluated by examining the deviation of accumulated GDH and RMSE between the observed and predicted dates. First, the model was fitted to different climate sites, Aomori and Fukuoka, and then tested at other sites nationwide. The selected model required an accumulated chill unit of 1450 for endodormancy completion with an effective temperature range from -6 to 14℃, and then the accumulated development of 6000 GDH to reach full bloom. The model, based on hourly temperature data only, succeeded in expressing the trends of full bloom dates for almost all sites except for southern Kyushu region, with the RMSE between observed and predicted dates of 2.33 days.

収録刊行物

  • 農業気象

    農業気象 66 (4), 269-277, 2010

    日本農業気象学会

被引用文献 (2)*注記

もっと見る

参考文献 (22)*注記

もっと見る

詳細情報 詳細情報について

問題の指摘

ページトップへ