This work presents the potential for high-resolution remote sensing data (LiDAR digital terrain models) to determine the spatial heterogeneity of terraced landscapes. The study objective is achieved through the identification of a new parameter that distinguishes this unique landscape form from more natural land formations. The morphological indicator proposed is called the Slope Local Length of Auto-Correlation (SLLAC), and it is derived from the local analysis of slope self-similarity. The SLACC is obtained over two steps: (i) calculating the correlation between a slope patch and a defined surrounding area and (ii) identifying the characteristic length of correlation for each neighbourhood. The SLLAC map texture can be measured using a surface metrology metric called the second derivative of peaks, or Spc. For the present study, we tested the algorithm for two types of landscapes: a Mediterranean and an Alpine one. The research method involved an examination of both real LiDAR DTMs and simulated ones, in which it was possible to control terrace shapes and the percentage of area covered by terraces. The results indicate that SLLAC maps exhibit a random aspect for natural surfaces. In contrast, terraced landscapes demonstrate a higher degree of order, and this behaviour is independent of the morphological context and terracing system. The outcomes of this work also prove that Spc values decrease as the area of terraced surfaces increases within the investigated region: the Spc for terraced areas is significantly different from the Spc of a natural landscape. In areas of smooth natural morphology, the Spc identifies terraced areas with a 20% minimum height range covered in terraces. In contrast, in areas of steep morphologies and vertical cliffs, the algorithm performs well when terraces cover at least 50% of the investigated surface. Given the increasing importance of terraced landscapes, the proposed procedure offers a significant and promising tool for the exploration of spatial heterogeneity in terraced sites. (C) 2014 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS). Published by Elsevier B.V. All rights reserved.
