water bodies (Beck et al., 2020). Indeed, these surfaces mirror the transmitted waveforms that have a pulse width of ~ 15 ns which corresponds to a ~ 2.25 m wide waveform (Dubayah et al., 2020). In total, 526,449 footprints from the GEDIv002 L2A product (Dubayah et al., 2021) were downloaded from NASA’s EarthDataSearch website (https://search.earthdata.nasa.gov/search) for this study, covering the entire area of interest for 2020. Due to atmospheric perturbations, some waveforms could not be used to give information on the vertical forest structure. Therefore, several filtering criteria were applied to remove unusable waveforms: (1) When the quality_flag provided in the GEDI data was set to
Open resource ↗GEDIv002 L2A · pdf-raw-page:6 lines:1-45Unverified paper record
High-resolution canopy height map in the Landes forest (France) based on GEDI, Sentinel-1, and Sentinel-2 data with a deep learning approach
arXiv · 20 Dec 2022 · 10.48550/arxiv.2212.10265
Abstract
In intensively managed forests in Europe, where forests are divided into stands of small size and may show heterogeneity within stands, a high spatial resolution (10 - 20 meters) is arguably needed to capture the differences in canopy height. In this work, we developed a deep learning model based on multi-stream remote sensing measurements to create a high-resolution canopy height map over the "Landes de Gascogne" forest in France, a large maritime pine plantation of 13,000 km$^2$ with flat terrain and intensive management. This area is characterized by even-aged and mono-specific stands, of a typical length of a few hundred meters, harvested every 35 to 50 years. Our deep learning U-Net model uses multi-band images from Sentinel-1 and Sentinel-2 with composite time averages as input to predict tree height derived from GEDI waveforms. The evaluation is performed with external validation data from forest inventory plots and a stereo 3D reconstruction model based on Skysat imagery available at specific locations. We trained seven different U-net models based on a combination of Sentinel-1 and Sentinel-2 bands to evaluate the importance of each instrument in the dominant height retrieval. The model outputs allow us to generate a 10 m resolution canopy height map of the whole "Landes de Gascogne" forest area for 2020 with a mean absolute error of 2.02 m on the Test dataset. The best predictions were obtained using all available satellite layers from Sentinel-1 and Sentinel-2 but using only one satellite source also provided good predictions. For all validation datasets in coniferous forests, our model showed better metrics than previous canopy height models available in the same region.
Plant phenotyping relevance
Sentinel/GEDI等のリモートセンシング画像から樹冠高を推定する深層学習手法を開発し、外部データで検証しているため、植物形質取得法が中心である。
abstractwe developed a deep learning model based on multi-stream remote sensing measurements to create a high-resolution canopy height map
abstractThe evaluation is performed with external validation data from forest inventory plots and a stereo 3D reconstruction model based on Skysat imagery
abstractThe model outputs allow us to generate a 10 m resolution canopy height map
Code and data availability
The paper's primary phenotyping-relevant input is the GEDI L2A canopy height dataset (526,449 footprints over the Landes forest, 2020), explicitly downloaded from NASA's EarthDataSearch. This is a public, paper-specific sensor dataset directly used for the study's canopy height measurements and model training. No code,
This is an automatically classified, unverified record. Curator approval is required before any resource enters the Catalog.