Tagging and Mapping of Mixed Dipterocarp Mountain Forest at Species Level using an Airborne Hyperspectral Imager

Abstract

Forest inventories describe the quantity and quality of trees and other organisms of the forest and the characteristics of the land on which the forest grows. Thus, to manage the resource of the forest systematically, the forest has to be precisely identified and classified before implementing decisions. A study on the spatial distribution of mountain mixed hill dipterocarp forest tree species in Gunung Stong Forest Reserve, Kelantan, Malaysia was conducted using an airborne hyperspectral imaging technique to identify, tag and map tree distribution at species level for future sustainable harvest and management planning. The general objective of this study is to assess the capability and usefulness UPM-AISA airborne hyperspectral sensor in Pre-Felling forest inventory while the specific objectives are to identify, map and tag individual trees in Gunung Stong Forest Reserve, Kelantan. A Sobel filter was used to enhance the image followed by a Spectral Angle Mapper (SAM) analysis to classify the individual tree species within the study plot. A digital map of tree tagging was produced with the tree species identified were Kelat (Syzgium spp), Keledang (Artrocarpos spp), Mengkulang (Heritiera spp), Tempinis (Streblus elongatus), Keranji (Dialium spp), Tulang Daing (Callerya atropurea), Meranti Sarang Punai (Shorea parvifolia), Kembang Semangkuk Jantung (Scaphium macropodum), Bintangor (Calophyllum spp), Nyatoh Minyak (Sapotaceae spp), Mersawa (Anisoptera spp), Resak (Cotylelobium spp), Sepetir (Sindora spp), Temponek (Artocarpus rigidus) and Petaling. The mapping accuracy of 89.66% was attained in the 1 ha study plot. Tree tagging using airborne remote sensing has a great potential in the Pre-F inventory and should be integrated with a GIS database management for future decision in management, development and utilization of sustainable forest resource.