基于PROSAIL模型和无人机高光谱数据的冬小麦LAI反演

为及时准确高效监测小麦叶面积指数（leaf area index，LAI），获取了冬小麦挑旗期和开花期地面实测光谱与无人机高光谱遥感影像数据，并基于查找表建立PROSAIL辐射传输模型得到冬小麦冠层模拟光谱数据，利用数学统计回归模型与偏最小二乘回归法分别构建冬小麦LAI单变量、多变量预测模型，以实测LAI数据对预测结果进行精度评价，将最佳预测模型应用于无人机高光谱影像以分析LAI空间分布情况。结果表明，冬小麦各生育时期的预测模型均具有较高的预测精度，单变量预测模型和多变量预测模型的决定系数分别为0.598~0.717和0.577~0.755，其中以基于植被指数的多变量预测模型表现最优，其在开花期的验证精度最高，RMSE和MAPE分别为0.405和12.90%。在LAI空间分布图中，开花期预测效果优于挑旗期，各试验小区的LAI分布较为均匀。

Retrieval of Winter Wheat Leaf Area Index by PROSAIL Model and Hyperspectral Data

In the cause of monitoring leaf area index(LAI) timely,accurately and efficiently,the ground measured spectrum and UAV hyperspectral remote sensing image data of winter wheat at flagging stage and flowering stage were acquired,and the PROSAIL radiation transfer model was established based on the look-up table to get the simulated spectral data of winter wheat canopy.The univariable prediction models and multivariable prediction models of LAI were constructed by mathematical statistical regression models and partial least square regression method.The accuracy of the prediction results was evaluated by the measured LAI data,and the best prediction model was applied to the hyperspectral image of UAV to analyze the spatial distribution of LAI.The results showed that the prediction models of each growth period had high prediction accuracy.The determination coefficient (r²) of the univariable prediction model was in the range of 0.598-0.717,and the r² of the multivariable prediction model was in the range of 0.577-0.755.The optimal prediction model was the multivariable prediction model based on vegetation index,and the verification accuracy of the model was highest at flowering stage,with RMSE and MAPE of 0.405 and 12.90%,respectively.In the LAI spatial distribution map,the prediction effect at flowering stage was better than that at flagging stage,and the LAI distribution of each experimental area was relatively well-distributed.