Fuzzy land element classification from DTMs based on geometry and terrain position

Land elements have been used as basic landform descriptors in many science disciplines, including soil mapping, vegetation mapping, and landscape ecology. We describe an approach in terrain classification, with the objective of deriving a method for classifying land elements from DTMs based on their fundamental characteristics. The methodology for modelling land elements is implemented as a two-step process: first, form elements are classified based on local geometry, and second, land elements are derived by evaluating the form elements in their landscape context. Form elements are derived by fuzzy classification of slope and curvature at a specified global scale (window size), The form elements are reclassified according to their geomorphometric context using a higher scale terrain position index. The resulting land elements are evaluated with respect to their predictive value for modelling soil properties. It is shown that scaling geomorphometric properties is important for applying them to predict soil properties and to model landform units. The presented model, based on scaled geomorphometric properties and geomorphometric context, using a limited number of model parameters, is capable of modelling fundamental land elements that can be utilized in soil-landscape modelling and in other applications in land resource management. (C) 2003 Elsevier B.V. All rights reserved.