以不同施肥模式为基础,分析了晚稻群体冠层光谱反射率、一阶微分光谱和归一化光谱特征,并对叶片氮含量、氮积累量、产量、叶面积指数和叶干物质积累进行了相关性分析,构建了以高光谱特征参数为自变量的水稻氮素营养诊断模型.结果表明,叶片氮素含量与665 nm处冠层光谱反射率呈极显著相关性(p<0.001),与554 nm和672 nm处的一阶微分光谱也呈极显著相关性(p<0.001);以λr构建的指数函数y=684.91e0.028x,决定系数(R2)为0.90、(SDr-SDb)/(SDr+SDb)构建的指数函数y=0.66e0.11x,决定系数(R2)为0.88,均能很好地诊断在有机肥和无机肥配施模式下的水稻氮素营养.%Based on different fertilization patterns, the canopy spectral reflectance, the first derivative spectra and normalized spectral characteristics of late rice, and a correlation among leaf nitrogen content, nitrogen accumulation, yield, leaf area index and leaf dry matter accumulation were analyzed. Then the rice nitrogen nutrition diagnosis model with high spectral characteristic parameters as in-dependent variables was constructed. The results showed that: there has a great relevance between leaf nitrogen content with canopy spectral reflectance at 665 nm (p<0.001), the same with the first derivative spectra at 554 nm and 672 nm (p<0.001);There has an exponential function byλr:y=684.91e0.028x, the coefficient of determination (R2) was 0.90, and an exponential function by (SDr-SDb)/(SDr+SDb):y=0.66e0.11x, the coefficient of determination (R2)was 0.88, these two model could diagnose the late rice nitrogen nutrition well under the conditions of organic manure and chemical fertilizers.
展开▼
