Canopy structure from space using <scp>GEDI</scp> lidar

Large-scale, globally consistent characterizations of the Earth's terrestrial ecosystems are a critical component international conservation and ecological monitoring initiatives. To help coordinate operationalize these efforts, Convention on Biological Diversity (CBD) Group Earth Observations Biodiversity Observation Network (GEO BON) have devised list essential biodiversity variables (EBVs) that can be remotely sensed (Pereira et al. 2013). Among them, lidar (light detection ranging)-derived EBVs stand out for their ability to accurately consistently characterize three-dimensional (3D) properties ecosystems, such as biomass, vertical stratification, topography (Coops 2021). Lidar data been instrumental across range applications, including models tree species richness (Fagua 2021), community composition (Hakkenberg 2018), wildlife habitat (Burns 2020), wildfire behavior (Botequim 2019), microclimate (Davis 2019). Recent advances in airborne systems offer ecologists quantify ecosystem structure with unprecedented precision accuracy – albeit at single point time, over limited spatial extents, considerable logistical hurdles, relatively high costs. Spaceborne lidar, other hand, from missions like National Aeronautics Space Administration's (NASA's) Ice, Cloud, land Elevation Satellites (ICESat-1 ICESat-2) Global Ecosystem Dynamics Investigation (GEDI) largely resolve or mitigate issues providing open, consistent, multi-temporal forest near-global extents (Markus 2017; Dubayah 2020). In this letter, we focus GEDI mission, first spaceborne investigation specifically designed measure canopy science (Dubayah Perched International Station (ISS) since April 2019, has operationally collecting data, specifically: waveform profiles equatorial mid-latitudes (± 52°). Raw waveforms processed derive higher level products plant area index, volume density, foliage height diversity, aboveground biomass 2022; Duncanson 2022). Although GEDI's coverage temporal repeat frequency potential transform global capabilities, some important considerations should noted hoping employ novel datasets. estimates possess approximately 2–3 m (Beck This degree is due primarily uncertainty horizontal geolocation footprints sensor pointing precision, where small differences may accentuated long distances each laser pulse travels to/from ISS. As such, slight lateral shifts an individual footprint result discrepancies estimated heights, especially heterogeneous canopies abundant gaps. Despite challenges, version 2 geolocational (mean ± standard deviation) 10 already well within many current satellite-derived products, expected improve upcoming releases Additional factors affecting include signal attenuation dense difficulties ground finding (a estimation) steep terrain. Where paramount concern, studies demonstrated further accuracies few meters using return correlation matching (or “bullseye”) methods address systematic error by co-registering (Hancock At present, does not itself provide spatially continuous maps moderately fine resolutions (less than ~100 m). Unlike gridded “wall-to-wall”) imagery optical satellites Landsat, sampling instrument records measurements discrete ~25 diameter footprints, interstitial areas remain unsampled between (60 along-track) transects (600 across-track) (Figure 1a). Thus, although design ensures coverage, it comes expense gaps undersampled regions owing limitations ISS's orbital geometry (eg sparse overpass regions) inability near-infrared penetrate clouds. against discontinuous scheme analysis-ready research community, team released series structural features biomass. These produced statistically aggregating coarser resolution 0.1–1 km) raster grids (Tang 2019; 2022) interpolating ancillary remote-sensing datasets radar 1c) (Healey 2020; Potapov benefit multiple affect consistency derived maps, quality quantity observations natural variability conditions biogeographic gradients (Lang Wang Data users therefore carefully consider trade-offs modeled versus samples. By end its extended currently set March 2023, systematically collected 20 billion 3D structure. Moreover, several more years, which would greatly increase density allow comprehensive change detection. With measurements, high-frequency revisit intervals, instruments GEDI, those planning NASA's Surface Topography Vegetation mission), revolutionize characterization large-scale dynamics, far-reaching implications management. Financial support was provided NSF DEB award 1924942. used preparation letter publicly available.