关键词: 小麦, 株型, 冠层光谱, 覆盖度, 植被指数, 叶面积指数

Abstract:

With the application of remote sensing in estimating crop biochemical composition, the estimation precision has became a research focus. Plant-type is the main influential factor to canopy spectral due to its influence on canopy structure. It is very important to identify plant type of wheat cultivars for improving estimation accuracy. 10 winter wheat cultivars with two plant types were used to discuss their canopy spectral reflectance and the method to identify plant type using spectral characteristics and vegetation coverage (COV). The results were as follows: (1) Their canopy spectral reflectance had obvious difference in near infrared bands at earlier stages, especially at jointing stage (Fig.1). It was proved that Jointing was the best stage to identify plant-type, and in this stage there was a big differences for canopy spectral reflectance between 2 plant types in wavebands of 680 nm and 760－900 nm (Table 2). These bands lied in the position of red edge and near infrared, which were the plant canopy sensitive bands. So it was very useful to improve the measurement accuracy of leaf biochemical composition. (2) Leaf area index (LAI) was positively correlated with normal difference vegetable index [NDVI(680, 890)]. The correlation coefficient was different at different stage(Fig.3,4). The correlation between LAI and NDVI was the foundation to identify plant type by using NDVI and COV(vegetable coverage). (3）With COV increasing, NDVI was increased, but the difference between two plant types became diminishing. It reflected their plant-type characteristics, that is to say, leaves are erect and upwards in erect cultivars but horizontal and downwards in horizontal cultivars. So COV of horizontal cultivars was higher than that of erect ones under the same NDVI. On the other hand, when COV of the two cultivars was equal, NDVI of the erect leaf type varieties was higher than that of horizontal leaf type cultivars (Fig.5). As for wheat, jointing stage with low and different COV in two pant-type cultivars is the best stage to identify plant-type. There was significant difference in NDVI of erect and horizontal cultivars (P＜0.05) at jointing stage. It is feasible to know plant type information using COV and NDVI. This method is fast and non-destructive. It provides the foundation to assess plant biochemical composition quantitatively by remote sensing.

Key words: Wheat, Plant type, Spectral, Coverage density, NDVI, LAI