Mapping an invasive plant species of Solidago altissima in urban landscape using hyperspectral sensing

M.R. MAHMUD, S. NUMATA, T. HOSAKA, N. HASSAN, WIDIYATNO, M.M. AZMY

The feasibility of the remote sensing technique in detecting an invasive plant species, Solidago Altissima, in urban environment is investigated. This study examines the potential to upscale the in-situ hyperspectral information at crown level for landscape scale detection. Sensitive and peak reflectance of the S. altissima flower was identified at 480-590 nm. The hyperspectral indices named S. altissima flower index (SAFI) was introduced. Crown scale observation shows that SAFI able to discriminate S. altissima from the surrounding features with average probability of 54%. In the community level within a 1-m grid plot, the SAFI sensitivity to discriminate plot with different S. altissima flower distribution was effective from 45% dominancy. At landscape scale, effective detection was found at SAFI value over 0.3. Direct upscaling of crown scale hyperspectral information to landscape scale was influenced by spectral bands, phenology status, community size and carotenoid contents.