Species type and forest health assessment via hyperspectral remote sensing in the part of Himalayan range, India

Abstract

Ground spectra using analytical spectral device were collected for dominant species in Rajpur and Mussoorie hills of Lesser Himalayas of India for mapping dominant plant species. Hyperspectral remote sensing data (EO?1) was used for classification of temperate species such as Quercus leucotrichophora, Cedrus deodara, Thuja orientalis, subtropical species like Pinus roxburghii, and tropical species such as Shorea robusta, Lantana camara, etc. using Spectral Angle Mapper. A total of 14 species were mapped along with the 4 other land use/cover classes (Agriculture, settlement, main river course, barren land). Map accuracy was 77% assessed on the basis of 66 ground truth ground control point. The exotic species, Lantana camara was mapped in the area which has been found to be distributed from tropical to lower temperate regions and was showing impact on the health of the neighboring species which was derived from the Hyperion data. Large impact has been observed in the Shorea robusta species and its health distribution map showed 48% healthy and 52% less healthy part.

DOI: http://dx.doi.org/10.3329/dujbs.v23i2.20093

Dhaka Univ. J. Biol. Sci. 23(2): 135-146, 2014