本文選題：水稻　切入點：稻曲病　出處：《江蘇農(nóng)業(yè)科學》2017年17期 　論文類型：期刊論文

【摘要】：為了提前1個月預報出水稻稻曲病發(fā)生的氣象條件適宜程度,根據(jù)中長期預報原理,采用因子膨化滑動相關(guān)普查、空間拓撲和最優(yōu)相關(guān)技術(shù),篩選出對綜合稻曲病指數(shù)影響最顯著的預報因子,分別構(gòu)建基于氣象要素、海溫因子、大氣環(huán)流指數(shù)的預報模型,并對3種模型的預報結(jié)果采用算術(shù)平均、加權(quán)平均和多元回歸方法進行集成。結(jié)果表明,建立的3種模型均通過了顯著性檢驗,預報效果較為理想,經(jīng)過集成后提高了單個模型的擬合精度和獨立樣本試報的準確性,其中多元回歸集成的效果更好。因此,建立的稻曲病預報模型可投入業(yè)務(wù)使用,預報結(jié)果將為稻曲病防治工作提供較為充足的時間。
[Abstract]:In order to predict the suitable meteorological conditions of rice scab occurrence one month in advance, according to the principle of medium and long term forecast, the factor expansion sliding correlation survey, spatial topology and optimal correlation technique were adopted. The forecast models based on meteorological factors, sea surface temperature factors and atmospheric circulation index were constructed, and the results of the three models were calculated by arithmetic average. The results show that the three models have passed the significance test and the prediction effect is satisfactory. The fitting accuracy of the single model and the accuracy of the independent sample test are improved after the integration. Therefore, the forecast model can be put into operation, and the forecast result will provide sufficient time for the prevention and control of rice smut disease.
【作者單位】： 江蘇省氣象局;
【基金】：公益性行業(yè)(氣象)科研專項(編號:GYHY201306035) 江蘇省氣象局科研基金(編號:KM201504)
【分類號】：S165.28;S435.111.4
，

本文編號：1644202