Improvement in Estimation of Aboveground Biomass of Albizia lebbeck using Fraction Reflectance of Landsat TM Data

O. P. Tripathi; J. Deka; J. Y. Yumnam; P. Mahanta

doi:10.18811/ijpen.v1i1.7118

O. P. Tripathi Department of Forestry, North Eastern Regional Institute of Science and Technology (Deemed University),
J. Deka Department of Forestry, North Eastern Regional Institute of Science and Technology (Deemed University), Nirjuli-791109, Arunachal Pradesh, INDIA
J. Y. Yumnam Department of Forestry, North Eastern Regional Institute of Science and Technology (Deemed University), Nirjuli-791109, Arunachal Pradesh, INDIA
P. Mahanta Department of Forestry, North Eastern Regional Institute of Science and Technology (Deemed University), Nirjuli-791109, Arunachal Pradesh, INDIA

DOI: https://doi.org/10.18811/ijpen.v1i1.7118

Keywords: Above ground biomass (AGB), spectral unmixing, endmember, Landsat TM, Fraction reflectance value.

Abstract

The study emphasizes on the ability of spectral mixture analysis (SMA) to improve the estimate of above ground biomass from spectral reflectance of satellite data. Digital number (DN) values of satellite data were converted to surface reflectance and fraction reflectance data. Linear regression analysis was performed between above ground biomass data with both total surface reflectance and SMA produced fraction reflectance values separately. The analysis revealed that the best positive correlation was observed between AGB and the fraction reflectance values of FR_TM 5/3 with R2 = 0.865 whereas with the total reflectance data, atmospherically resistant vegetation index (ARVI) and TM5/3 resulted moderate correlation R2= 0.452 and - 0.248, respectively. Pixel level AGB ranges between 0.001t to 13.53t. The total AGB of the study area (275.85 ha) was estimated to be 14249t. Thus separation of endmembers from the mixed pixel from a course to medium resolution satellite data can play an important role in improving AGB estimation performance, especially for those sites with multiple cover features. The study demonstrates the potentiality of Landsat TM data in concatenate with SMA in improving the estimation of AGB which can be further improved by identifying all possible endmembers and highly configured methodology.