Title of article :
Digital terron mapping
Author/Authors :
Florence Carré، نويسنده , , Alex B. McBratney، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 2005
Abstract :
The aim of this study is to propose a new concept, the terron, with an associated model, for helping environmental managers make decisions. Terron is a whole entity integrating soil properties (pedon) with topographic soil-forming factors (represented here by landform attributes)—which can be assimilated to terroir. For terron mapping, landform is not only an interpolation support as it is in existing models, but also a taxonomic variable. Data from French databases are used to illustrate this concept. Terron mapping consists of four stages. First is the classification of n area-representative soil types with landform attributes by fuzzy k-means with extragrades into n terron classes. Second is the establishment of terron-landform rules using a multi-linear regression. Third is the interpolation of the terron-landform rules using regression-kriging. Fourth is validation with a second sample dataset. The results are compared with a soil map created with a regression-kriging model.
Keywords :
Fuzzy k-means , Terron , FuzMe , Soil-landform , Decision-making support