摘要
本文研究了华南地区主要土壤的光谱反射特性,进行了土壤光谱的主组元分析、模糊聚类及与土壤理化性状作相关性分析,土壤光谱特征分类与现行中国土壤分类系统及中国土壤系统分类进行了比较对照。其主要结果表明,华南地区主要的光谱曲线的形态特征,主要表现为平直型(玄武岩发育砖红壤类)、缓斜型(水稻土类)和陡坎型(红、黄壤类)三种类型;土壤光谱分类与按母质母岩划分高度一致;应用主组元分析对土壤光谱进行分类,结果与现行土壤分类有一定的吻合性,而不吻合部分(如水稻土、火山灰土)由于受地带性因素的深刻影响,被归属于所处地带的地带性土类;综合应用土壤光谱特征指标及其它一些分类指标(如B层粘粒硅铝率、阳离子交换量、砂粘比等能更有效区分土壤类型,分类结果与按中国土壤系统分类相当吻合。土壤光谱特征指标尤其适用于土属一级的划分,符合率在83%以上。
This paper deals with the spectral characteristics of main types of soils in
sou-thern China. Meanwhile,using methods of principal component analysis(PCA)andfuzzy
category, 33 samples are classified according to 10 spectral character indexs.The results are
compared with traditional soil classification and Chinese Soil Taxo-nomic Classification.The
relationships between the spectral characters and physicaland chemical proparties of the soils
have been also investigated.The results showthat different soil groups give different spectral
reflectance curves,including flattype of latosels developed from basalts.slope type of the
paddy soils and steep typeof red soils and yellow soils sequences;the suitability between soil
spectral characterclassofication and parent material types is very high;and by the method of
PCA,itcan be found soil spectral classification is identical to traditional soil classificationin
some senses.Then,as paddy soils and andosols are deeply affected by natural zon-al
factors,they can be sorted into different groups of zonal soils.Using soilspectral characteristic
indexes and other classification indexes(e.g. clay SiO_2/Al_2O_3of subsoil,CEC, coarse
silt/clay,etc.), the soils can be classified effectively, theresults are consistent with those
obtained based on Chinese Soil Taxonomic Classifi-cation.The soil spectral characteristic
indexes is especially suitable to be used forthe division of soil genus, and the suitability
between soil spectral characteristicclassification and soil genus classification is more than 83
percent.
出处
《土壤学报》
CAS
CSCD
北大核心
1995年第1期58-68,共11页
Acta Pedologica Sinica
关键词
土壤
光谱反射特性
数理分析
土壤分类
Soil spectral characteristics, Mathematical analysis, Soil classi-fication