First Near-Continental Leaf Area Index (LAI) and Vertical Foliage Profile (VFP) Product from the Geoscience Laser Altimeter System (GLAS)

Abstract:

Spaceborne lidar missions, such as ICESat, provide unique observations of forest vertical structure globally. To date several global forest height maps have been produced using GLAS data on board of ICESat. However, 3D information from the full waveform lidar has not been fully utilized, which can further be used to derive highly accurate LAI and VFP products. Here we present the first near-continental LAI and VFP product from GLAS data based on a geometric and radiative transfer model. We first retrived LAI and VFP from ~1.1 million waveforms using a developed recursive method and analyzed its distribution at ecoregion level over Contiguous United States (CONUS). We then assessed its accuracy with different existing remote sensing LAI products. Finally, we explored its relationship with factors from different environmental gradients. Our product exhibits a very good agreement with LAI data sets from airborne lidar data across major forest types (r² = 0.67, bias = -0.13, and RMSD = 0.75). It also shows a fair correspondence with a 30 m Landsat LAI map produced over CONUS (r² = 0.18, bias = 0.33, and RMSD = 1.99), reflecting fundamental differences in their retrieval methodologies. In particular, Landsat appears to saturate relative to ICESat at a value of about LAI = 4 or over high forest cover (> 70%). Derived LAI and VFP products also exhibit strong correspondence with environmental factors (e.g. elevation and annual precipitation). Results from this study highlight the capability of spaceborne waveform lidar in measuring vertical foliage structural parameters at continental to global scales and should help improve our understanding of the role vertical canopy structure plays in terrestrial ecosystem dynamics.