Evaluation of Remotely Sensed Vegetation and Elevation Products in a Tropical Forest

Abstract:

Space-based methods of remote sensing have enabled the generation of global or near-global elevation and vegetation products that have been applied to a wide spectrum of earth sciences, while airborne remote sensing platforms have enabled the collection of high resolution data products with regional coverage. In order to use small-footprint airborne data together with space-based products in future applications, it is necessary to understand how the data products are related, particularly in regards to accuracy and systematic biases. This work addresses the need for a better understanding of the quality of existing global elevation and vegetation products by validating a radar-derived, void-filled digital elevation model and a lidar-derived global canopy height model against ICESat waveforms and a small-footprint airborne lidar survey over tropical forests in the country of Gabon. The bare-earth digital terrain model and canopy height model obtained from the airborne lidar survey are validated against ICESat waveforms, and are then used to evaluate errors and biases in SRTMv3 and the 1km global canopy height model developed by Simard, et al. across the extent of the survey region near the city of Libreville, Gabon. ICESat waveforms are subsequently used to evaluate SRTMv3 and Simard, et al.’s canopy height model for the entire country of Gabon.