Object-based class modelling for multi-scale riparian forest habitat mapping |
| |
| Institution: | 1. Forest Ecology and Conservation Group, Department of Plant Sciences, University of Cambridge, Downing Street, Cambridge CB2 3EA, UK;2. Department of Sustainable Agro-ecosystems and Bioresources, Research and Innovation Centre, Fondazione E. Mach, Via E. Mach 1, 38010 San Michele all''Adige (TN), Italy;3. Department of Global Ecology, Carnegie Institution for Science, 260 Panama Street, Stanford, CA 94305 USA;4. Centre for Ecology and Hydrology, Bush Estate, Penicuik, Midlothian, EH26 0QB;5. School of Biological Sciences, University of Aberdeen, Cruickshank Building, St Machar Drive, Aberdeen AB24 3UU, UK;6. School of Geography, University of Leeds, Leeds LS2 9JT, UK;7. Department of Geography, University College London, London WC1E 6BT, UK;8. Forest Research Centre, Sabah Forestry Department, P.O. Box 1407, 90715 Sandakan, Sabah, Malaysia;9. Faculty of Science and Natural Resources, Universiti Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia |
| |
| Abstract: | Object-based class modelling allows for mapping complex, hierarchical habitat systems. The riparian zone, including forests, represents such a complex ecosystem. Forests within riparian zones are biologically high productive and characterized by a rich biodiversity; thus considered of high community interest with an imperative to be protected and regularly monitored. Satellite earth observation (EO) provides tools for capturing the current state of forest habitats such as forest composition including intermixture of non-native tree species. Here we present a semi-automated object based image analysis (OBIA) approach for the mapping of riparian forests by applying class modelling of habitats based on the European Nature Information System (EUNIS) habitat classifications and the European Habitats Directive (HabDir) Annex 1. A very high resolution (VHR) WorldView-2 satellite image provided the required spatial and spectral details for a multi-scale image segmentation and rule-base composition to generate a six-level hierarchical representation of riparian forest habitats. Thereby habitats were hierarchically represented within an image object hierarchy as forest stands, stands of homogenous tree species and single trees represented by sunlit tree crowns. 522 EUNIS level 3 (EUNIS-3) habitat patches with a mean patch size (MPS) of 12,349.64 m2 were modelled from 938 forest stand patches (MPS = 6868.20 m2) and 43,742 tree stand patches (MPS = 140.79 m2). The delineation quality of the modelled EUNIS-3 habitats (focal level) was quantitatively assessed to an expert-based visual interpretation showing a mean deviation of 11.71%. |
| |
| Keywords: | Semi-automated habitat mapping OBIA Multi-level class modelling Object hierarchy WorldView-2 Natura 2000 |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
