REAL

Synoptic environmental indicators as image analogs for landscape analysis

Myers, W.L. and Patil, G.P. and Taillie, C. and Walrath, D.C. (2003) Synoptic environmental indicators as image analogs for landscape analysis. Community Ecology, 4 (2). pp. 205-217. ISSN 1585-8553

[img]
Preview
Text
comec.4.2003.2.7.pdf

Download (4MB) | Preview

Abstract

Spatially synoptic multivariate image data implicitly embody information on landscape pattern, for which analytical techniques of explicit pattern extraction are evolving. In parallel, a multiplicity of 'environmental indicators' is being generated in the arena of geographic information systems. Landscape ecological analysis offers substantial opportunity for configuring these indicators synoptically as cells over spatial extents and for stacking them into complementary sets of image-structured multiple environmental indicators whereby the values of the indicators become intensity analogs of brightness for spectral bands. As environmental signal analogs of multiband images, these data become available to image portrayal in both graytone and quasi-color renditions to reveal joint properties of pattern for visual interpretation. Likewise, many of the conventional image analysis operations can be conceived more broadly to allow their application in the indicator context. This includes combinatorial approaches such as calculation of an NDVI equivalent from indicator intensities. Similarly, supervised and unsupervised analyses can have meaningful application in the context of multiple environmental indicators. Furthermore, newer techniques of pattern-based image segmentation can also be applied. Application to habitat modeling for vertebrates from Gap Analysis shows the effectiveness of the approach.

Item Type: Article
Subjects: Q Science / természettudomány > QH Natural history / természetrajz > QH540 Ecology / ökológia
Depositing User: xBarbara xBodnár
Date Deposited: 11 Nov 2017 12:08
Last Modified: 31 Dec 2023 00:16
URI: http://real.mtak.hu/id/eprint/67442

Actions (login required)

Edit Item Edit Item