Hyperspectral Prediction Model for Maize Leaf SPAD in the Whole Growth Period

In the present article, a 2-year maize field experiment for the whole growth period was conducted. Hyperspectral data and chlorophyll contents (SPAD) of maize leaves were measured respectively using portable spectrometer and SPAD-502 chlorophyll meter. Various spectral parameters were selected to build construct prediction models for maize leaf SPAD in annual growth periods based on correlation analysis of the two factors, and detailed validation and evaluation were made. The results showed that there was little difference between the mean curves of spectral reflectance or its first derivative in the two years; the sensitive areas were basically the same about the correlation coefficient curves of leaf SPAD and spectral reflectance or its first derivative; prediction models built with spectral parameters based on first derivative were unstable; prediction models built with LCI, DSI(R-550around, R-680around) and DSI(R-680arouna, R-710around) were all in good conditions, and maize leaf SPAD could be effectively predicted.