Airborne hyperspectral AVIRIS-NG data for vegetation carbon stock mapping based on red edge position parameter and narrowband vegetation indices in Sholayar reserve forest, Kerala

Abstract

The biophysical and biochemical spectral responses in the form of red-edge position (REP) and narrowband indices for carbon stock mapping have been evaluated in this study using AVIRIS-NG data in Sholayar reserve forest, Kerala. Maximum first-derivative spectrum, linear four-point interpolation technique, and adjusted linear-extrapolation techniques were employed for REP estimation. Results exhibited a maximum peak of REP at 717 nm and two extreme dips at 677 nm and 772 nm. Regression analysis between field-based AGB with REP (R2 = 0.6537) and narrowband indices; EVI, ARVI, VRI, and RNDVI was adopted exhibiting significant correlations (R2=0.5100, R2=0.5106, R2=0.5220, and R2=0.5507, respectively) (P-value < 0.001). Results show that dry-deciduous plantations (mainly Tectona grandis) have the lowest AGB values of 25-100tons/hectare, whereas, Wet-evergreen forest exhibited the highest values (300-350tons/hectare). Correlation between plot-level AGB and predicted AGB being 0.7065. Conclusively, the retrieval of structural measures like biomass can be achieved through these methods using AVIRIS-NG data.