ON THE HOTSPOT EFFECT OF LEAF CANOPIES - MODELING STUDY AND INFLUENCEOF LEAF SHAPE

Authors
Citation
Wh. Qin et Yq. Xiang, ON THE HOTSPOT EFFECT OF LEAF CANOPIES - MODELING STUDY AND INFLUENCEOF LEAF SHAPE, Remote sensing of environment, 50(2), 1994, pp. 95-106
Citations number
22
Categorie Soggetti
Environmental Sciences","Photographic Tecnology","Remote Sensing
ISSN journal
00344257
Volume
50
Issue
2
Year of publication
1994
Pages
95 - 106
Database
ISI
SICI code
0034-4257(1994)50:2<95:OTHEOL>2.0.ZU;2-4
Abstract
Hotspot is a prominent feature of the reflectance distribution of a ve getation canopy consisting of finite size foliage. It depends on the g eometric dimension and spatial organization of vegetation elements, an d therefore has a potential for diagnosing canopy geometric structure. In the context of this study, we first reconcile different notations used by previous workers. This leads to geometrically quantifying the cross-correlation function, which is essential for the hotspot modelin g. A comprehensive formulation for the hotspot effect at both leaf and canopy levels is then developed, by generally parameterizing some bas ic parameters such as mean area of shadows and overlap between shadows cast by scatterers. A rectangle model is proposed to account for the influence of noncircular shape of scatterers on the hotspot effect, an d explicit expressions for both the cross-correlation function and the hotspot width are obtained. It is shown that for a leaflike object, t he angular hotspot width progressively broadens with an increase of m, the ratio of mean leaf width to the length. For the whole canopy, the relative distribution of the hotspot intensity mainly depends on this ratio. That is, the cross-correlation function decreases more rapidly for smaller m as the viewing direction diverges from the illumination direction. As a result, canopy reflectance increases with m, particul arly in the region around the hotspot point where the reflectance dist ribution strongly relies on the ratio m. For m = pi/4, the rectangle m odel produces nearly same result as the disk model. This indicates thi s rectangle model is more realistic and flexible than those based on c ircle-shaped scatterers (leaves) or their shadows, which is the common assumption underlying in most existing hotspot models.