Another dimension from LiDAR – Obtaining foliage density from full waveform data

LiDAR tells the user where surfaces are, not what they are. In this study we investigate the potential for waveform LiDAR to provide more information on the nature of the returns over forestry. Waveform LiDAR was acquired for ten Pinus radiata plots in a New Zealand plantation, along with comprehensive leaf area sampling in 2m vertical bands. The decay rate of each waveform peak was shown to be a useful tool for estimating foliage density, and has potential for identifying regions containing ground and understorey. Leaf Area Density (LAD) is an expression of foliage density per unit height, and a relationship between waveform decay rate and LAD was developed with an R2 of 56%. Incorporating the proportion of discrete LiDAR that fell in that band (which itself has an R2 of 50%) improves this model to explain 69% of the variation in LAD. This is a good result, especially given the costs and difficulties in measuring leaf area directly. As foliage density varies dramatically on a fine scale it was not possible to differentiate the nature of every single LiDAR return – but by averaging over a small area local variation in LAD could be easily mapped. Ground returns could be distinguished as having short decays, and broad leafed understorey typically had values between those of the canopy and ground, although surface roughness and slope make it impossible to robustly identify single returns. This study produced a useful model for estimating LAD in Pinus radiata which could easily be extended to other coniferous species