| dc.description.abstract | The increase of Green House Gases (GHG), particularly carbon dioxide in the atmospheric is one factor that affect the global warming. One cause of the increase was due to emission of carbon that came from deforestation and land degradation. In the mitigation process for reducing the impact of global warming, one crusial activity is estimation of carbon sequestered in the vegetation. Commonly, carbon content sequestered at vegetation is derived from biomass estimation either from above ground biomass or below ground biomass. In this study, the focus was on estimation of the above ground biomass using remote sensing technology. Nowdays, the development of remote sensing techniques had been widely used for providing various information related to natural resources. Indonesia which is located in the tropical region, have several handicaps in using optical remote sensing due to cloud covers during wet season and smoke and haze during dry season. For the above reason, this study emphasizes on the use of radar image technology for estimating the above ground biomass. In this study the used radar image is ALOS PALSAR (Advance Land Observing Sattelite – Phase Array L-band Synthetic Aperture Radar), that had been launched in Januari 2006 by the Japanese government. This sensor is an active microwave sensor which has capability to penetrate thick cloud layer. An active microwave sensor be accepted dan reflectanced by several diresction that usually called polarization. Polarization is the vector direction from electromagnetic microwave on vertical line (V) and horizontal line (H). The main objective of this study is to develop the biomass estimation model using 50 meter resolution of ALOS PALSAR image in concession of PT. Trisetia Intiga, Central Kalimantan. Biomass model was developed by analyzing the relationship between backscatter value and field biomass. Backscatter value from tree polarization images, namely HH, HV, and HH/HV were analyzed simultaneously with field biomassa that derived using Ketterings allometric equation. In this study the obtained stand structure equation, Y=2482 ℮-0,04X with negative exponential or J-reverse form having realtively steep slope. From the stand structure equation, then it can be concluded that stand condition within the study area is belonged to logged over area secondary forest, where tree density with diameter less or equal to 60 cm is relatively low in comparison with the smaller tree size. From 30 sample plots measured, the above ground biomass is ranged from 68,03 ton/ha to 599,43 ton/ha with coefficient of variation (CV) equal to 50,29%. Proportionaly, the highest biomass is come from tree biomass (64,97%), then followed by biomass of necromass (20,13%), pole biomass (12,16%), sapling biomass (2,75%), litter biomass (3,3x10-5), and undergrowth biomass (5x10-5). The correlation and regretion analyses between backscatter value and field biomass show that there are 38 statistically significant equations having R² ≥ 50%. However, the verification analysis found that only two models identified as the best models having no-significant X²-test, low RMSE (Root Mean Square Error), low bias, and mean deviation < 10%. The best models are: 1) AGB = 3880.40613930209exp (0.250184794335192 HV) with coefficient determination (R²) of 56% and RMSE of 0.47%. 2) AGB = 1022.27050600692 exp (-0.0114464608949151 HV²) with coefficient determination (R²) of 59% and RMSE of 0.55%. Those models was than use for estimating biomass in PT. Trisetia Intiga consession area. | en |
