Estimation of chengal trees relative abundance using coarse spatial resolution hyperspectral systems

Abstract

Tree species relative abundance estimation of a tropical rainforest is quite a challenge especially when coarse spatial resolution data is utilized. In this chapter, modified canopy fractional cover (mCFC) was enhanced from the canopy fractional cover (CFC) model in estimation of chengal trees relative abundance in Hyperion EO-1 data, which is coarse spatial resolution hyperspectral data. Besides, mixture tuned matched filtering (MTMF) was employed to Hyperion EO-1 data to test the ability of mCFC in estimating the chengal trees relative abundance. Thus we hypothesized that mCFC has better capability to estimate the abundance of chengal trees more accurately than MTMF, while MTMF has better capability in estimating undisturbed forest. The accuracy of mCFC model (r2 = 0.667, P< 0.05) shows that mCFC has capability to estimate relative abundance of chengal trees better that MTMF. Therefore it can be concluded that the relative abundance of certain tree species extracted from Hyperion EO-1 satellite data using modified CFC is an obtrusive approach for identifying tree species composition