Terrain slope effect on forest height and wood volume estimation from gedi data

Abstract

The Global Ecosystem Dynamics Investigation LiDAR (GEDI) is a new full waveform (FW) based LiDAR system that presents a new opportunity for the observation of forest structures globally. The backscattered GEDI signals, as all FW systems, are distorted by topographic conditions within their footprint, leading to uncertainties on the measured forest variables. In this study, we explore how well several approaches based on waveform metrics and ancillary digital elevation model (DEM) data perform on the estimation of stand dominant heights (HHdddddd) and wood volume (V) across different sites of Eucalyptus plantations with varying terrain slopes. In total, five models were assessed on their ability to estimate HHdddddd and four models for V. Results showed that the mod-els using the GEDI metrics, such as the height at different energy quantiles with terrain data from the shuttle radar topography mission’s (SRTM) digital elevation model (DEM) were still dependent on the topographic slope. For HHdddddd, an RMSE increase of 14% was observed for data acquired over slopes higher than 20% in comparison to slopes between 10 and 20%. For V, a 74% increase in RMSE was reported between GEDI data acquired over slopes between 0–10% and those acquired over slopes higher than 10%. Next, a model relying on the height at different energy quantiles of the entire waveform (HHTTnn) and the height at different energy quartiles of the bare ground waveform (HHGGnn) was assessed. Two sets of the HHGGnn metrics were generated, the first one was obtained using a simulated waveform representing the echo from a bare ground, while the second one relied on the actual ground return from the waveform by means of Gaussian fitting. Results showed that both the simulated and fitted models provide the most accurate estimates of HHdddddd and V for all slope ranges. The simulation-based model showed an RMSE that ranged between 1.39 and 1.66 m (be-tween 26.76 and 39.26 m3·ha-1 for V) while the fitting-based method showed an RMSE that ranged between 1.26 and 1.34 m (between 26.78 and 36.29 m3·ha-1 for V). Moreover, the dependency of the GEDI metrics on slopes was greatly reduced using the two sets of metrics. As a conclusion, the effect of slopes on the 25-m GEDI footprints is rather low as the estimation on canopy heights from un-corrected waveforms degraded by a maximum of 1 m for slopes between 20 and 45%. Concerning the wood volume estimation, the effect of slopes was more pronounced, and a degradation on the accuracy (increased RMSE) of a maximum of 20 m3·ha-1 was observed for slopes between 20 and 45%.