Intercomparison of Real and Simulated GEDI Observations across Sclerophyll Forests

Forest structure is an important variable in ecology, fire behaviour, and carbon management. New spaceborne lidar sensors, such as the Global Ecosystem Dynamics Investigation (GEDI), enable forest structure to be mapped at a global scale. Virtual GEDI-like observations can be derived from airborne laser scanning (ALS) data for given locations using the GEDI simulator, which was a tool initially developed for GEDI's pre-launch calibration. This study compares the relative height (RH) and ground elevation metrics of real and simulated GEDI observations against ALS-derived benchmarks in southeast Australia. A total of 15,616 footprint locations were examined, covering a large range of forest types and topographic conditions. The impacts of canopy cover and height, terrain slope, and ALS point cloud density were assessed. The results indicate that the simulator produces more accurate canopy height (RH95) metrics (RMSE: 4.2 m, Bias: -1.3 m) than the actual GEDI sensor (RMSE: 9.6 m, Bias: -1.6 m). Similarly, the simulator outperforms GEDI in ground detection accuracy. In contrast to other studies, which favour the Gaussian algorithm for ground detection, we found that the Maximum algorithm performed better in most settings. Despite the determined differences between real and simulated GEDI observations, this study indicates the compatibility of both data sources, which may enable their combined use in multitemporal forest structure monitoring.

Automated summary

Forest structure is a crucial variable in ecology, fire behavior, and carbon management. New spaceborne lidar sensors, like GEDI, allow for global mapping of forest structure. This study compared real and simulated GEDI observations and found that the simulator produced more accurate metrics for canopy height and ground detection. Despite some differences between real and simulated GEDI observations, the study suggests that both data sources are compatible and can be used together in monitoring forest structure over time.