Examining Urban Tree Health using High Resolution Worldview-2 Satellite Image in Hong Kong
Refereed conference paper presented and published in conference proceedings
Officially Accepted for Publication

CUHK Authors
Author(s) no longer affiliated with CUHK

Full Text

Other information
AbstractIncreasing cases of diseased tree fall in urban area causes serious casualties in Hong Kong. There are more than a hundred diseased trees which are listed under the category of old and valuable tree in Hong Kong. On one hand, they are old trees with historical and cultural value; however, on the other hand, they threaten public safety at all time. High spatial resolution satellite image can identify individual tree crowns and allows examination of tree health in dense urban area. This study utilized 8-band WorldView-2 image acquired in 2012 and 2014. Innovative vegetation indices (VI) were derived from the spectral data, which provides information about vegetation vigor of a common urban tree species, Ficus microcarpa. With reference to the government database, more than 20 crown samples were collected for both healthy and diseased trees. The diseased group was further divided according to density and the types of disease they suffered from including Brown Root Rot Infection (BRR), Dead branches, Bodies, Obvious wound, Dieback twigs, Cavity, Decay. Four VIs emphasising different aspects of vegetation characteristics and minimizing the external effects were computed. The vegetation spectra and 24 VIs were analysed for the diseased and healthy groups statistically. The result showed that canopy water content index yielded significant difference between dense and sparse diseased group. Broadband greenness index such as Enhanced vegetation index (EVI); narrowband greenness index involving pigment content and red edge slope index were able to distinguish the types of disease especially for BRR and when the trees suffered from serious cavity. Results from two years’ image showed similar trend. It is suggested that satellite images with higher spatial resolution and with red-edge band have potential in examining individual tree health condition.
Acceptance Date16/05/2016
All Author(s) ListWong. F.K.K., Fung, T
Name of ConferenceThe 33rd International Geographical Congress
Start Date of Conference21/08/2016
End Date of Conference25/08/2016
Place of ConferenceBeijing
Country/Region of ConferenceChina
LanguagesEnglish-United States
KeywordsWorldView-2, Spectrum Analysis, Vegetation Index, Individual Tree, Tree Health Examination

Last updated on 2018-21-01 at 21:37