利用新型光谱指数改善冬小麦估产精度

    Improving winter wheat yield prediction by novel spectral index

    • 摘要: 基于冬小麦返青期至乳熟期8次采样的地面光谱数据和收割时的产量数据,首先,利用光谱反射率与产量进行了统计分析,可见光波段的光谱反射率与产量在起身后期才达到稳定的显著负相关水平;近红外波段的光谱反射率与产量在所有生育期都表现出稳定的显著正相关;短波红外波段的光谱反射率与产量在进入灌浆期后才达到稳定的显著负相关水平。其次,根据冬小麦冠层光谱的波形特征,利用近红外波段890 nm反射峰、980 nm和1 200 nm两个弱水汽吸收谷、短波红外1 650 nm和2 200 nm反射峰,设计归一化差值光谱指数,并与冬小麦产量进行相关分析,结果表明:利用上述波段组合定义的归一化差值光谱指数与产量在各个生育期都达到了显著或极显著相关水平,而归一化差值植被指数(NDVI)与产量间的相关在营养生长阶段不显著。最后,以(890 nm,1200 nm)弱水汽吸收光谱指数为例,建立了各个生育期的产量预报模型,为实现冬小麦营养生长期长势监测与更早、更可靠的产量预报提供了依据。

       

      Abstract: The coefficients of correlation between yields and spectral reflectances in eight different growth stages were first calculated. The statistical results show that yield is positively correlative with spectral reflectance in NIR bands for all the growth stages, negatively correlative in visible bands from late jointing stage, and also negative correlative in shortwave-infrared bands from seeding stage. Second, the normalized difference spectral indices combined by [890 nm, 1200 nm], [890 nm, 980 nm], [890 nm, 1650 nm], [890 nm, 1650 nm] and [890 nm, 2200 nm] are significantly correlative with yields in all the 8 growth stages, which is better than NDVI. Finnally, the remote sensing models in different growth stages for winter wheat yield were built by [890 nm, 1650 nm] weak water absorption index.

       

    /

    返回文章
    返回