Untangling the effect of urban vegetation type and structure on spectrally unmixed greenness. Issue 12 (2nd December 2021)

• Record Type: Journal Article
• Authors: Czekajlo, Agatha; Coops, Nicholas C.; Goodbody, Tristan R. H.
• Is Part Of: Remote sensing letters. Volume 12:Issue 12(2021)
• Page Start: 1216
• Page End: 1226
• Publication Date: 2021-12-02
• Subjects: Remote sensing -- Periodicals; Remote sensing; Periodicals; 621.3678
• DOI: 10.1080/2150704X.2021.1978580
• Languages: English
• ISSNs: 2150-704X

Abstract

ABSTRACT: Measurements of vegetation greenness, such as those derived from spectral unmixing of satellite imagery using a greenness–darkness–brightness model, are crucial for improved mapping and analysis of urban greenspaces and their socio-ecological benefits. Differences in vegetation types and structures can, however, influence how spectral wavelengths are reflected and thus represented through unmixed greenness and darkness fractions. This study untangles the relationship and distribution of sub-pixel greenness and darkness across key structural metrics for coniferous, deciduous, and low vegetation in select Vancouver, Canada greenspaces. We utilized airborne laser scanning (ALS) data to describe structural components of canopy cover, canopy height, rumple index, and vertical variability, which were then analysed with spectrally unmixed greenness and darkness fractions from Landsat satellite imagery and available land cover data. Greenness was linked to vegetation type, whereas darkness was also influenced by structure, especially for conifers. Specifically, tall conifers observed twice as much darkness as greenness, and darkness increased by ~30% to match moderate greenness levels of high-coverage deciduous stands. Results correspond to previous work on biophysical differences and shadowing effects from tree crowns. Future greenspace research utilizing spectrally unmixed greenness fractions should be mindful of potential shadowing impacts caused by tall and/or dense vegetation.