DOI: 10.1016/j.jag.2017.06.012
Scopus记录号: 2-s2.0-85032196558
论文题名: Pattern-based, multi-scale segmentation and regionalization of EOSD land cover
作者: Niesterowicz J ; , Stepinski T ; F
刊名: International Journal of Applied Earth Observation and Geoinformation
ISSN: 15698432
出版年: 2017
卷: 62 起始页码: 192
结束页码: 200
语种: 英语
英文关键词: Clustering
; EOSD land cover
; Forest patterns
; Landscape metrics
; Segmentation
Scopus关键词: forest management
; GIS
; land cover
; landscape
; regionalization
; sustainable development
; thematic mapping
; Canada
; Ontario [Canada]
; Quebec [Canada]
英文摘要: The Earth Observation for Sustainable Development of Forests (EOSD) map is a 25 m resolution thematic map of Canadian forests. Because of its large spatial extent and relatively high resolution the EOSD is difficult to analyze using standard GIS methods. In this paper we propose multi-scale segmentation and regionalization of EOSD as new methods for analyzing EOSD on large spatial scales. Segments, which we refer to as forest land units (FLUs), are delineated as tracts of forest characterized by cohesive patterns of EOSD categories; we delineated from 727 to 91,885 FLUs within the spatial extent of EOSD depending on the selected scale of a pattern. Pattern of EOSD's categories within each FLU is described by 1037 landscape metrics. A shapefile containing boundaries of all FLUs together with an attribute table listing landscape metrics make up an SQL-searchable spatial database providing detailed information on composition and pattern of land cover types in Canadian forest. Shapefile format and extensive attribute table pertaining to the entire legend of EOSD are designed to facilitate broad range of investigations in which assessment of composition and pattern of forest over large areas is needed. We calculated four such databases using different spatial scales of pattern. We illustrate the use of FLU database for producing forest regionalization maps of two Canadian provinces, Quebec and Ontario. Such maps capture the broad scale variability of forest at the spatial scale of the entire province. We also demonstrate how FLU database can be used to map variability of landscape metrics, and thus the character of landscape, over the entire Canada. © 2017 Elsevier B.V. Citation statistics:
资源类型: 期刊论文
标识符: http://119.78.100.158/handle/2HF3EXSE/80001
Appears in Collections: 气候变化事实与影响
There are no files associated with this item.
作者单位: Space Informatics Lab, Department of Geography, University of Cincinnati, Cincinnati, United States
Recommended Citation:
Niesterowicz J,, Stepinski T,F. Pattern-based, multi-scale segmentation and regionalization of EOSD land cover[J]. International Journal of Applied Earth Observation and Geoinformation,2017-01-01,62