基于遥感技术的不同施氮水平下小麦条锈病病情反演研究

为构建不同施氮条件下,小麦条锈病病情光谱反演模型,设置了在不同氮素水平条件下接种小麦条锈病,将菌情指数与植被指数、一阶微分参数进行回归分析,构建抽穗期、开花期、灌浆期、乳熟期共5个模型。为了评估施氮量对病情反演模型的影响,在模型中加入氮素因子,模型病情反演预测效果表明,抽穗期模型加入氮素因子后预测效果有所提高,抽穗期的模型1-1(R2=0.392 8,P=0.005 4)、1-2(R2=0.449 8,P=0.011 3)、2-2(R2=0.573 3,P=0.001 7)预测效果较好且较稳定,开花期、灌浆期、乳熟期模型预测效果不理想。本研究结果表明,可以利用植被指数、一阶微分参数较好反演抽穗期小麦条锈病病情,加入氮素因子后预测效果有所提高,说明氮素因子对病情反演有影响。

Wheat stripe rust disease index inversion by remote sensing technology under different nitrogen application

In order to construct wheat stripe rust disease remote sensing estimation model under different nitrogen supply level, five estimation models at heading, flowering, filling and milk stages were built by regression analysis between disease index (DI) and vegetation index and the first order differential parameter.To estimate the impact of nitrogen to the disease inversion model, nitrogen factor was added into the model.Predicting efficiency of the disease inversion model showed that:The predicting efficiency of the heading stage models was improved when the nitrogen factor was added;Heading stage models 1-1 (R²=0.3928, P=0.0054), 1-2(R²=0.4498, P=0.0113) and 2-2(R²=0.5733, P=0.0017)displayed better and stable predicting result;Models of flowering, filling and milk stage showed poor prediction results.In conclusion, vegetation index and differential parameter can be used in estimating wheat stripe rust disease index at heading stage.Adding nitrogen factor can improve estimation efficiency.It is also proved that wheat stripe rust disease inversion is influenced by nitrogen supply level.