GEDI meets the Sentinels: Exploration of the Mutual benefits from Joint Use of GEDI and Sentinel-1 / Sentinel-2 data

We investigated, how well ALS data, which is costly and often unavailable, can be substituted by GEDI data, e.g. for automated training data sampling to be used in classification of forest properties and disturbances in European and Tropical forests. Further, we evaluated the potential of GEDI together with Sentinel-1 and Sentinel-2 data to generate new better, cheaper or so-far not available forest attribute products in the frame of Copernicus and national forest monitoring initiatives.

The GEDI space-borne Lidar sensor, mounted on the International Space Station, has been delivering 3D measurements for the past three years. The plot wise data gives insights into the world's forests similar to airborne laser scanning (ALS), but at a lower resolution and free of charge. We found out, that GEDI is suitable to subsitute ALS data in areas with moderate terrain, while errors significantly increase for steeper slopes. Further analysis revealed limitations in canopy cover accuracy from GEDI. On the other hand, we found, that GEDI full waveform data is well explaining the vertical structure of forests. Related structural indicators have significant explanatory power for vertical structure and Sentinel data can be used to roll this information out to a wall-to-wall product. Such information can be used in future to evaluate the biodiversity potential of forests but may also be useful for mapping susceptibility to wind throw.

For tropical areas with limited ALS coverage, GEDI is a very good alternative data source, e.g. to estimate biomass for specific forest or land use types. One example is agroforestry in Uganda, for which this product delivers biomass values. We found estimated biomass levels in the region of Kampala to be on average 25.4 Mg/ha. Such information is important for national carbon accounting and REDD+ (Reducing emissions from deforestation and forest degradation in developing countries0.