实验测定了黄土高原半干旱区63个土壤样本的14个土壤肥力指标和其可见光波段内的光谱反射率,然后对二者结果做多因子逐项单相关分析,以此探究用光谱表征土壤肥力的可行性.结果表明:(1)土壤肥力的多个重要因子与光谱反射率的相关性都在0.7以上;(2)土壤剖面各肥力因子与光谱的相关性比土壤表层差,并且二者在大部分肥力因子上都存在着显著的差别.(3)用光谱反射R值、反射率倒数的对数A值和反射率一阶导数R'值分别与土壤肥力因子做逐步线性相关性分析发现光谱R'值相关性明显比其它两种好.其中有机碳、阳离子交换量、全氮、速效氮、速效钾与光谱的相关性r的分别达到-0.9063、-0.9185、-0.8771、-0.8604、-0.744.这五个因子可作为土壤肥力在可见光波段的光谱表征指标.%This study has tested 14 soil fertility indexes of 63 soil samples in semiarid regions of the Loess Plateau,measured its spectral reflectance in visible bands, and then made multiple factor synthesis analysis so as to explore a method to measure soil fertility by spectral characterization. The results indicate that: (1) The spectral reflectance correlation of many soil fertility multiple core factors are above 0.7; (2) The spectrum correlation of soil profile fertility factor is poorer than surface soil, and the differences are significant in most fertility factors. (3) The linear correlation analysis among spectral reflection R value, reflectivity reciprocal logarithm A value, reflectivity first-order derivative R' value and soil fertility factors shows that spectrum R' value correlation is obviously better than other two kinds of Values. The spectrum correlation coefficients of organic carbon, cationic exchange capacity, total nitrogen, available nitrogen, available potassium are 0.9063, 0.8771, 0.9185, 0.8604, and 0.744 respectively. These five factors can be used as soil fertility characterization indexes in visual band spectrum.
展开▼
