Tree diversity assessment and above ground forests biomass estimation using SAR remote sensing: A case study of higher altitude vegetation of North-East Himalayas, India

Abstract

Abstract Forest biomass is one of the significant components for assessing the productivity and sustainability of the forest s ecosystem. In the present study, the mapping and estimation of above ground biomass (AGB) was performed using field observations employing allometric equations and C band Sentinel 1A and L band Advanced Land Observation System (ALOS) Phased Array L-band Synthetic Aperture Radar (PALSAR) remote sensing satellite images for the parts of north eastern Himalayan forests, Arunachal Pradesh. Total 57 transects of 500\u202fm\u202f×\u202f10\u202fm size was laid in West Kameng and Tawang districts (altitude range: 1047–4161\u202fm) of Arunachal Pradesh, in which a total of 12203 individuals of 67 tree species (Diameter at Breast Height (DBH) range: 3.18–235.45\u202fcm) was recorded. The study exhibited higher tree density (428 individuals ha−1) and basal cover (7658.63 m2ha-1) in the studied forests. The highest above ground biomass was contributed by Castanopsis indica (103.82\u202ft\u202fha−1) followed by Pinus roxburghii (46.24\u202ft\u202fha−1) and Quercus semicarpifolia (23.14\u202ft\u202fha−1). Further, AGB was estimated using the Horizontal-Horizontal (HH), Horizontal-Vertical (HV) polarizations acquired by the ALOS PALSAR 50\u202fm mosaic (2008) dataset and Vertical-Vertical (VV), Vertical-Horizontal (VH) polarizations acquired by the Sentinel 1A dataset. The relationship between field based AGB and radar backscatter for selected sample plots was established using pair wise correlation. While comparing the two satellite-based SAR images, it is observed that high correlation between radar backscatter and field based AGB was observed in ALOS PALSAR as compared to Sentinel 1A SAR. This is due to higher sensitive to backscatter values and higher penetration level of L band ALOS PALSAR as compared to C band Sentinel 1A SAR. The correlation study indicates high correlation of HV polarizations in ALOS PALSAR (HV polarization R2\u202f=\u202f0.63; HH polarization R2\u202f=\u202f0.58) as compared to Sentinel 1A (VV polarization R2\u202f=\u202f0.054 and VH polarization R2\u202f=\u202f0.044). The most accurate and reliable method to estimate AGB is the field based method, but it is time consuming and tedious. Whereas, AGB estimation by L band SAR remote sensing is time effective and can be used for larger area in comparatively less time. The present study exhibited poor species diversity in higher altitudes northeast Himalayan forests as species diversity decreases with increase in altitudes. On the other hand, the higher AGB in the studied forests may be due to favorable environmental conditions including less temperature and more nutrient availability as well as less biotic interferences in terms of anthropogenic disturbances, herbivory and grazing. The study signifies the potential use of L and C band backscatter information in the reliable estimation of spatial AGB in sub-tropical to alpine forests of Arunachal Pradesh.