Abstract
To estimate forest biomass in a large area, it is necessary to clarify the relationship between attributes of stand structure obtained by remote sensing and woodland biomass. We examined stand structure and estimated woodland biomass in an arid region of Western Australia. The research site was near Leonora located 600 km from Perth in Western Australia. The annual rainfall is approximately 200 mm. The dominant woody vegetation species is Acacia aneura. The spatial characteristics of the woodland in this region are that woodland canopies are not closed and the tree crown silhouette is relatively clear-cut. We established 35 plots (main size, 50 m × 50 m) and the diameters at 1.3 and 0.3 m heights, tree height and canopy silhouette area of all trees in the plots were determined. Each tree biomass was calculated by allometric equations using a destructive sampling method. Results of regression analysis indicated that the appropriate stand structural attributes for estimation of woodland biomass were stand basal area (SBA), canopy coverage (CC) and leaf area index (LAI). SBA had the highest estimation accuracy of woodland biomass but SBA was not suitable for estimation by satellite imagery. The woodland biomass estimation accuracy by CC ( R 2 > 0.94, P < 0.0001) or LAI ( R 2 > 0.92, P < 0.0001) was lower than that by SBA ( R 2 > 0.99, P < 0.0001) but CC and LAI are considered to be powerful indicators for woodland biomass estimation by satellite imagery, where open forest is distributed. Particularly, in the case of arid land, the correlation between CC and woodland biomass was reported in the Salhel region of Africa and in Queensland, Australia; therefore, the forest biomass estimation method by CC is considered to be applicable to other arid- and semiarid-area open forests and woodlands.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call