高光谱技术提取植被生化参数机理与方法研究进展

概述了目前利用高光谱技术估测地表植被生化参数理论与技术的最新研究进展,着重介绍了前人为提高遥感精度不断改进从光谱数据中提取植被生化参数的一些方法和理论,重点论述了提高遥感信息的信噪比（SNR）、改进遥感数据的分析方法、植被物理参数的细化和逐步确定,是目前植被生化参数遥感估测研究的前沿领域和科学问题,为人们尽快全面了解高光谱技术在植被生化参数方面应用进展和方法拓展,提供了条件、概貌和综论。

HYPERSPECTRAL REMOTE SENSING FOR ESTIMATING BIOCHEMICAL VARIABLES OF CANOPY

Hyperspectral remote sensing (narrow bands) provides the possible estimates of biochemical characteristics of plant such as moisture, leaf pigment (for example, chlorophyll, carotenoids and anthocyanins), lignin, cellulose, protein, amino acids, sugar, starch, macro-element (nitrogen, phosphorus and potassium as the example), micro-element (such as Fe, Mn, Co, Zn, Cu). It is special significant to obtain biochemical information of canopy at territorial or even at global levels which could be widely used to study the ecosystem, accelerate the application of precision agriculture technology, and so on. At leaf or small area levels, not only it is significant for the researching on the forming of plant reflectance spectrum for remote sensing, but also, it may provide a rapid, nondestructive method for the detection of these characteristics in lab or in field. But from ground leaf or dry leaf level to fresh leaf level and to canopy level, more and more factors influence the spectrum. There are interferences among the factors, which may add the difficulty to estimate a certain biochemical variable or influence the accuracy. Some new progress in obtaining information from ground-level and airborne or satellitic level hyperspectral reflectance data were summarized and emphasis was especially focused on introducing the methods for improving accuracy of estimating biochemical variables and the mechanism of hyperspectral remote. The problems might be occurred because of the methods self that were being widely used to estimate biochemical variables of canopy also were pointed out. Future work should be focused both on improving the ratio of signal to noise, reforming the method of spectrum analysis and accelerating application of the existing research findings, but also on itemizing the biophysical variables and making clear their spectral characteristics (contribution to the reflectance spectrum). On the other hand, it is necessary to do more research on establishing new canopy spectrum reflectance models or improving existing models by mechanism in the future. The paper may supply the convenience, general picture and discuss to understand the new progress in the technologies and theories of hyperspectral remote sensing for estimating biochemical variables of plant.