为了更好地估算植被指数,必须计算不同土壤类型的土壤线;但由于土壤线的影响因素较多,因此土壤线参数获取困难.本文以室内土壤高光谱反射率为主要研究对象,分析并确定影响土壤线的主要因素;利用室内土壤光谱反射率计算土壤线,将所得土壤线参数用于与土壤线有关的植被指数的计算,比较这些植被指数与大豆生理参数(叶绿素a与叶面积)的相关关系是否强于归一化植被指数(NDVI),分析该土壤线参数计算方法的可行性.结果表明:土壤线的影响因素主要有土壤类型、有机质、矿物组成、秸秆覆盖等;而土壤水分、粗糙度等尽管对土壤光谱反射率大小也有很大影响,但由于对光谱曲线形状影响较小,因此对土壤线的影响也较小;与土壤线有关的植被指数部分消除了土壤背景的影响,其与大豆生理参数的相关系数显著高于NDVI,说明利用室内土壤光谱反射率计算土壤线的方法可行,所得参数适于计算基于土壤线的植被指数.

In order to estmi ate vegetation index precisely, the soil lines of different soil types must be identified.However, it is difficult to obtain the soil line parameters, because there aremany influencing factors. With the laboratory soilhyperspectral reflectance studied, the main factors influencing the soil line were determined. The laboratory soil hyperspectral reflectance was used to calculate soil line parameters, which were introduced to compute the vegetation indices related to soil line. The correlation between the indices and soybean chlorophyll a or leaf area index (LAI) was comparedwith thatbetween normalized difference vegetation index (NDVI) and chlorophyll a orLAI, then the feasibility of the soil line computingmethodwas analyzed.The resultswere as follows: themain factors affecting soil line parameterswere not the bandwidth but the soil class,the latitude, organicmatter, mineral composition, and straw residues. Soil classes, with differentphysical and chemical properties, show significant differences in spectral characteristics, especially the reflectance shape at red and infrared bands. The slope ofsoil linewith low soil albedowasnotalways smaller than thatwith higher soil albedo. What’smore,latitude also influenced the soil line parameters because of the regular distribution of soil parentmaterial and other soil properties. The effectoforganicmatter on soil reflectance in visible and infrared bands(620—810nm) ismuch stronger than other spectral regions, the reflectance curve at visible and infrared domain is concave, so the slope of soil line increaseswith organicmatter content. What’smore the effect of organicmatter on soil line parametersmay be stronger than soil class.Based on soil line, the soilvegetation indicespartly elmi inate the soilbackground noises, and the correlation between the indices and soybean chlorophyll a orLAI ismore significant than thatbetweenNDVI and chlorophyll a orLAI, which indicates that the soil line parameter derivingmethod is feasible and appropriate for the vegetation indices based on soil line.In order to enhance themonitoring precision ofvegetation indices, the soil line parameters should be calculated based on the spatial difference of the main factors influencing the parameters. However, the results of the study is based on ground crop spectral reflectance data and obtained in the laboratory, as a result, the influence of atmosphere is not completely considered. Therefore, when the result is used in remote sensing mi age data, the influence of atmosphere on\nsoil and crop reflectancemustbe analyzed.