Volume No. 8 Issue No.: 3A Page No.: 777-785 Jan-Mar 2014




Yali Glen and Samanta Sailesh *

Department of Surveying and Land Studies, PNG University of Technology, Morobe, Lae, Papua (NEW GUNICA)


Received on : November 25, 2013




Remote Sensing (RS) and Geographic Information Systems (GIS) are the powerful tools in land use/ land cover mapping, environmental and climatological modeling. This study proposes an empirical methodology to prepare digital data set of land use/land cover, biomass and carbon stock estimation using RS and GIS techniques. The study area is the PNG University of Technology campus (UNITECH) in the Lae city under Morobe province of Papua New Guinea. The common methods of satellite image classification are supervised and unsupervised classification algorithms. These classification methods are faster than traditional image interpretation method. Object based classification algorithms are more powerful than the conventional classification algorithms because they involve classification of the entire object (shape and texture) rather than pixel-by-pixel basis. Quickbird satellite image (2.4 m multispectral and 0.6 m panchromatic) is used for this purpose. Erdas Imagine 11, eCognition Developer 8.64 and ArcGIS 10.1 software are used to extract land use/land cover information from the high resolution data. Better and more accurate result is achieved in the object based approach compared to supervised classification. Land use/land cover data is used as an input along with average height, diameter and wood specific gravity for each class to calculate biomass of the area. Finally carbon stock values were assigned to the different land cover classes based on calculated of biomass or carbon stocks in different land use/ land cover types.


Keywords : Remote sensing, Land use/land cover, Biomass, Carbon stock, Geographic Information Systems (GIS)