基于地面实测的水稻冠层反射光谱,计算了常用的8个植被指数,并在产量形成生理特征的基础上,系统分析了水稻籽粒产量及其构成因素与各植被指数之间的关系。结果表明,通过单一生育时期或某个生育阶段的光谱植被指数来直接估测产量精度较低。发现叶面积氮指数(叶片氮百分含量与叶面积指数的乘积)的变化趋势很好地反映了产量的形成过程,且与光谱植被指数极显著正相关,基于此建立了水稻的光谱植被指数-累积叶面积氮指数-产量估测模型(VICLANIYieldModel)。并将其与LAD-产量模型、多生育期复合估产模型进行了比较,表明本模型预测精度最高。

Spectral reflectance of rice canopies with different nitrogen treatment was measured over an entire growing season and eight spectral indices such as RVI, NDVI, PVI etc were calculated. Based on the biological mechanism of yield formation, relationships of these vegetation indices to yield and its components were analyzed. The results showed that it was limited to predict yield with vegetation index from single or multiple developing stages. However, the dynamic curve of Leaf Area Nitrogen Index (product of LAI by leaf nitrogen content on dry weight basis) can well track the process of yield formation. Due to the close relationship with the vegetation index, Cumulative Leaf Area Nitrogen Index (CLANI, the area below the curve) was used to derive a model named VI-CLANI-Yield model for rice yield estimation. The comparison of the present model with the LAD-Yield model and complex VI-Yield model indicated that the yield estimation accuracy was best for VI-CLANI-Yield model with average relative error of 0.075. This suggests that VI-CLANI-Yield model would be a practical and effective approach for rice yield forecasting.