Remote Sensing–Derived Forest Regrowth Canopy Height, Biomass, and Age Across the Americas

OverviewThis dataset provides spatially explicit observations of secondary forest regrowth derived from the integration of remote sensing–based canopy height, aboveground biomass (AGB), and regrowth age. The dataset is designed to support analyses of forest structural and biomass recovery chronosequences, including comparisons between natural regeneration and commercial plantations. Each record represents a spatially joined point where canopy height (or AGB) observations are matched with regrowth age and forest type. Spatial CoverageThe dataset covers three broad regions: the Contiguous United States, Central America, and South America (southward to approximately 50°S). Forest Type StratificationEach data point is classified as either natural regeneration or commercial plantation. Forest type classification is based on spatial overlays with remote sensing–derived plantation maps (Fagan et al., 2022; Richter et al., 2024). Canopy Height Data SourcesCanopy height is derived from multiple sources: 1) GEDI Level 2A (GEDI_L2A v002; April 2019–March 2023; 2) global canopy height map from Potapov et al. 2021 (GLAD_HT); 3) global canopy height map from Lang et al. 2023 (ETH_HT). For GEDI L2A, only shots meeting conservative quality thresholds were retained, including quality_flag = 1, beam sensitivity greater than 0.95, and canopy height between 5 and 60 m. Aboveground Biomass (AGB) Data SourcesAGB are derived from GEDI Level 4A products (GEDI_L4A, version 1; April 2019–March 2023) and the ESA Climate Change Initiative (ESA_CCI) AGB map for 2021. For GEDI L4A, only shots meeting conservative quality thresholds were retained, including quality_flag = 1, l4a_quality_flag = 1, beam sensitivity greater than 0.95, and AGB values between 20 and 1000 Mg ha⁻¹. Regrowth Age EstimationForest regrowth age is calculated as the number of years since the most recent transition from non-forest to forest, inferred from Landsat-based land-cover change products spanning 1985–2021. Region-specific data sources include the National Land Cover Database (NLCD) for the United States (1985–2021; Sohl et al., 2025), the Tropical Moist Forests (TMF) dataset for Central America (1990–2021; Vancutsem et al., 2021), and MapBiomas national datasets for South America, including Bolivia, Brazil, Colombia, Ecuador, Paraguay, Uruguay, and Venezuela (1985–2021), Argentina (1998–2021), and Chile (2000–2021) (Souza et al., 2020). Spatial Overlay and Data IntegrationFor each region, GEDI footprints or wall-to-wall raster grids were spatially overlaid with regrowth-age maps and remote sensing–derived plantation maps. This integration assigns each observation a regrowth age and a forest-type classification, distinguishing between natural regeneration and commercial plantations. File Organization and FormatSix compressed (.zip) files are provided, containing canopy height and aboveground biomass (AGB) data for each of the three regions. The dataset is organized to facilitate regional and source-specific analyses while remaining compatible with common geospatial and statistical workflows. File StructureWithin each archive, data are stored as comma-separated values (CSV) files and spatially tiled into 10° × 10° geographic tiles. File names follow the convention {region}_{age_source}_{variable_source}_{tile}.csv, which encodes the geographic region, the source of regrowth age information, and the source of the canopy height or AGB data. CSV ContentsEach CSV file contains point-level observations with geographic coordinates recorded as latitude and longitude in decimal degrees. Associated attributes include regrowth age (in years) and indicator fields identifying whether each observation corresponds to natural regeneration or a commercial plantation. Intended UseThis dataset is intended to support analyses of forest structural and biomass recovery, including the derivation of canopy height and AGB chronosequences, comparisons between natural regeneration and plantations, and benchmarking of ecosystem and carbon-cycle models across broad spatial scales. Contact For questions or additional information about this dataset, please contact Lei Ma at [e-mail].