Species Distribution Model using Hyperspectral Data Application in Peatland, Central Kalimantan

摘要

Tumih/Parepat (Combretocarpus-rotundatus Mig. Dencer) of the family Anisophylleaceae and Meranti (Shorea Belangerang, ShoreaTeysmanniana Dyer ex. Brandis) of Dipterocarpaceae are groups of species of vegetation that is modeled for the designation of its spread. Pioneer species are predicted as an indicator of ecosystem restoration succession of wet tropical peatland characteristic (unique) and very susceptible (fragile) in an endemic Sundaland hotspot. Projected climate change and conservation planning a topic of heated discussion, the analysis of alternative approaches and the development of a combination of search algorithm projection models within the fabric of the diffusion of the species to attain this through a geospatial information system technology. Model approach used to work out problems in research on this vegetation species level based machine learning is a method (hybrid), which is a combination of wavelet and neural network. Often the approach to field data into general use in the management of natural resources and biodiversity assessment. Expression of the hybrid models provides encouraging results, to decide the classification of the species Tumih/Parepat and Meranti through mapping of vegetation species that are backed up by the pattern of the spectral curve of the field spectrometer.