基于高光谱植被指数的冬小麦产量监测

Monitoring of Winter Wheat Yield Based on Hyperspectral Vegetation Index

为了研究利用不同生育时期的植被指数监测冬小麦产量,以2 a不同肥料处理的冬小麦为研究对象,分析不同生育时期植被指数与冬小麦产量的关系,构建冬小麦产量的光谱植被指数监测模型。结果表明,植被指数能有效监测冬小麦产量,其中,孕穗期和抽穗期植被指数的监测效果较好,孕穗期的校正均方根误差(RMSEC)和决定系数(R_C^2)分别为1 131.42和0.78,抽穗期的RMSEC和R_C^2分别为1 015.59和0.77,拔节期的监测效果次之,灌浆期和成熟期的监测效果较差;但从作物生产角度考虑,拔节期监测冬小麦产量具有更重要的现实意义。研究表明,利用拔节期植被指数能够实现冬小麦产量的早期估测。

To predict the yield of winter wheat by using the vegetation indexes derived from the hyperspectrum at different growth stages,the two-years experiments of winter wheat treated with different fertilizers were implemented.The canopy spectral data of winter wheat were collected at different growth stages and the relationship between the vegetation indexes derived from the canopy spectral data and winter wheat yield data at different growth stages was analyzed to construct the predictive models of winter wheat yield.The results showed that the vegetation indexes at different growth stages had a good relationship with the winter wheat yield,of which most of the vegetation indexes at booting and heading stages performed best to estimate the yield with RC 2=0.78,RMSEC=1 131.42 at the booting stage,and RC 2=0.77,RMSEC=1 015.59 at heading stage.Then,the jointing stage was the next,and the filling and maturity stage were the worst.However,from the perspective of crop production,it had more important practical significance to monitor winter wheat yield in jointing stage.The results show that the early estimation of winter wheat yield can be achieved by using the vegetation index in jointing stage.