Abstract: Based on hyperspectral remote sensing data,  correlations between absorbed photosynthetic active radiation (APAR) and spectral reflectance, first derivative spectra and second derivative spectra of  APAR were analyzed. And spectral reflectance,  derivative spectra and vegetative indices were used to establish APAR estimating models. The results showed  that  the correlation between the optimal bands of hyperspectral reflectance, the first derivative spectra, the second derivative  spectra and  APAR were highly significant. The   first derivative spectra could better estimate APAR with the determinant coefficient  of  0.5218 at 769 nm; The model based on the five vegetation indices  combining best band could also achived better accuracies since the determine coefficients were all above 0.67, among which  renormalized difference vegetation index RDVI (758,781)  was the best to estimate APAR with the coefficient of 0.7585.

Key words:  rice, hyperspectral remote sensing, absorbed photosynthetic active radiation, vegetation index

摘要： 利用高光谱遥感数据，对光谱反射率、一阶微分光谱及二阶微分光谱与吸收光合有效辐射（APAR）进行相关分析，并利用反射率、微分光谱、植被指数法研究了水稻冠层APAR的估算效果。结果表明，高光谱反射率、一阶微分光谱及二阶微分光谱的最优波段与APAR的相关性均极显著，一阶微分光谱能够更好的估算APAR，它与APAR在769 nm处的拟合决定系数为05218；5种植被指数所选的最优波段组合拟合方程的决定系数在0.67以上，其中以复归一化差值植被指数（758, 781）估算效果最好，决定系数达0.7585。

关键词: 水稻, 高光谱遥感, 吸收光合有效辐射, 植被指数