Establishment and differences of high spectral estimation models for leaf SPAD values of different maize species
[Objective] This study explored the differences between leaf SPAD values of maize species and the hyper spectral feature,built the estimation models for SPAD values,and verified the application range.[Method] Through field experiment,this study measured the leaf SPAD values of multiple maize varieties and their hyper spectral data.By using the correlation analysis,stepwise regression analysis and other methods,regression models of SPAD values and related spectral parameters were constructed and screened.The accuracies of models were also tested using deviation rate.[Result] Leaf SPAD values of different maize varieties were different from high spectral reflectance and the first derivative of relevant band gap,while the differences among varieties were small.The closest relationships located near 560 and 700 nm.When comparing the first derivative of spectral reflectance between varieties,"red shift" phenomenon appeared,which was consistent in size with SPAD values.Leaf SPAD values and first derivative of spectral reflectance were significantly correlated near 510,615,690,and 740 nm.The model constructed by single band spectrum parameter was better than the model by combinational band,and the models were multiple and exponential equations.The built model with leaf spectral parameters of single maize can be used for other varieties with different accuracies.[Conclusion] The model established with hyper spectral band reflectance near 560 nm had the highest precision,and the deviation rate for predicting leaf SPAD values was generally less than 5.00 ％.
NETL
NSTL
万方数据
