Soils can sequester carbon and contribute to climate change mitigation, but credible soil carbon offset markets require robust Monitoring, Reporting, and Verification (MRV) programs. Current approaches rely on models with soil sampling or on repeated probability sampling. We propose a digital soil mapping (DSM) framework to estimate soil organic carbon (SOC) stock change and uncertainty. Using physical samples from a hypothetical project across seven fields in two U.S. states, we calibrate, predict, validate, and spatially aggregate SOC estimates. We assess the effect of calibration sample size on field- and project-level means and variances and extend the framework to quantify change. Estimated uncertainty of averaged SOC stocks was <1 Mg ha(-1), yielding a 12% uncertainty deduction for a 5-year, 430-acre project when using a probability of exceedance method. Assuming the sequestration rate is 0.6% of the stock size, simulations suggest deductions <10% for most projects >5 years and <5% for large ( >50,000 acres), long-term projects. This DSM approach offers a cost-effective and scalable alternative to other existing methods.

Voluntary carbon market (VCM) programs in agriculture depend on accurate measurements of soil organic carbon (SOC) that can be deployed at scale efficiently, but barriers are preventing widespread adoption. To overcome these challenges, we developed a digital soil mapping (DSM) framework driven by machine-learning and numerous spatial covariates, including long-term climate proxies, short-term climate and weather-related variables, topographic and edaphic measurements, and remote sensing time-series summaries. We show that the model can predict SOC content in the top 30 cm of soil using 5,230 measurements of SOC in agricultural land within 47 states in the contiguous United States (CONUS). Model predictions closely matched independent measured values. The intercept and slope of the cross-validated relationship at the agricultural field level were -0.179 and 1.095. The coefficient of determination was R2 = 0.811, and the RMSE was 0.041. In contrast, comparison of independent field measurements to four publicly available SOC data products using 165 fields that contained 3,285 in-situ soil samples showed poor ability of existing public SOC maps to reproduce measured values, underscoring the importance of quantification technologies developed specifically for agricultural land and with recent soil measurements. Three prior SOC data products underestimated SOC content at small values and overestimated it at large ones, while one underestimated SOC content at all values examined. Analysis of feature importance showed that time series summaries from Sentinel-2 are the strongest predictors, followed by temperature variables and features related to surface hydrology. These findings underscore the value of geographically representative training and validation data for quantifying SOC content in agricultural land and demonstrate that feature engineering can increase the sensitivity of SOC quantification to optical remote sensing summaries. Data-driven algorithms can generate accurate estimates of field-level SOC content in agricultural land in CONUS that overcome barriers to scale in the VCM.

Woody canopies regulate exchanges of energy, water and carbon, and their three-dimensional (3D) structure supports much of terrestrial biodiversity. Remote sensing technologies such as airborne laser scanning (ALS) now enable the 3D mapping of entire landscapes. However, we lack the large, harmonized and geographically representative ALS collections needed to build a global picture of woody ecosystem structure. To address this challenge, we developed the Global Canopy Atlas (GCA): 3,458 ALS acquisitions transformed into standardized and analysis-ready maps of canopy height and elevation at 1 m2 resolution. The GCA covers 56,554 km2 across all major biomes. 19% of this area has been scanned multiple times, and 87% of all GCA products are openly available, covering 95% of the total area. To showcase its wide range of applications, we applied the GCA in three case studies. First, we validated three global satellite-derived canopy height maps, finding poor performance at native resolution (1-30 m, R2 < 0.38) and moderate performance at 250 m resolution (R2 < 0.65). Second, analyzing global patterns in canopy gap size frequency we discovered an unexpectedly large variation of power law exponents from branch to stand level (α = 1.52 to 2.38), pointing to a fundamental scale-dependence of forest structure. Third, we developed a framework to standardize forest turnover quantification from multi-source, multi-temporal ALS. In a temperate forest in North America it revealed that 21% of canopy gaps closed within 12 years of opening and would thus be missed by infrequent monitoring. As demonstrated by these case studies, the GCA provides a novel data source for ecologists, foresters, remote sensing scientists and the ecosystem modelling community that substantially advances our ability to understand the structure and dynamics of woody ecosystems at global scales. ### Competing Interest Statement A Burt and M Demol are employees and/or shareowners of Sylvera Ltd. No other competing interests declared. Leverhulme Trust, https://ror.org/012mzw131, RPG-2020-341, RPG-2024-342 European Research Council, 101001905, 757526, 101059548 Natural Environment Research Council, https://ror.org/02b5d8509, NE/S01537X/1, NE/T011084/1, NE/Z504191/1 UK Research and Innovation, EP/Y003810/1 European Space Agency, FRM4BIOMASS Agence Nationale de la Recherche, ANR-10-LABX-25-01, ANR-19-CE32-0005-01 Oak Ridge National Laboratory United States Department of Energy, https://ror.org/01bj3aw27, DE-AC05-00OR22725 Project LIFE+ ForBioSensing, LIFE13ENV/PL/000048 National Fund for Environmental Protection and Water Management (Poland), 485/2014/WN10/OPNMLF/D Forest Research Institute, 261509 Royal Society, https://ror.org/03wnrjx87, RG/R1/251370 World Bank Government of Mozambique Innovate UK, 10004871 Czech Republic–Bavaria Free State ETC goal 2014–2020, Interreg V project No. 99 Deutsche Forschungsgemeinschaft, 411263134/2019-2022 NASA Land Cover Land Use Change and Carbon Monitoring System Program. Conselho Nacional de Desenvolvimento Científico e Tecnológico, 403297/2016-8, 401053/2019-9, 306386-2022-4 Amazon Fund, 14.2.0929.1 USAID, AID-OAA-A-11–00012 DELTA Lidar National Natural Science Foundation of China, 32401574 Australian Agency for International Development Dutch government Norwegian Agency for Development Cooperation, https://ror.org/04zbn7k04 United Kingdom Climate Change Unit European Cooperation in Science and Technology, https://ror.org/01bstzn19, LUC23023 JSPS KAKENHI, #21H05314, #21H02564 Universiti Brunei Darussalam, UBD/RSCH/1.18/FICBF(b)/2023/006 The Nature Conservancy ForestGEO FAPESP, 2017/22269-2 French Ministry of Higher Education and Scientific Research, One Forest Vision initiative French Ministry of Europe and Foreign Affairs, One Forest Vision initiative Western Australian government, State NRM program, Ngadju Conservation Aboriginal Corporation, Landgate program Western Australian Department of Biodiversity, Conservation and Attractions Australia's Terrestrial Ecosystems Research Network Government of Finland Government of Nepal German Federal Ministry for the Environment, Nature Conservation, Building and Nuclear Safety, International Climate Initiative (IKI) German Development Bank KfW, International Climate Initiative (IKI) Natural Science Foundation of Shanghai, 23ZR1419200 National Key R&D Program of China, 2024YFF1308100 USAID/Wake Forest University, Cooperative Agreement No. 7205-2721-CA-00005 NASA, NISAR CAL/VAL French Government Ministry of Higher Education of Malaysia, FRGS/1/2020/WAB03/UKM/02/1 Canadian Forest Service, https://ror.org/0430zw506 Ontario Ministry of Natural Resources National Key Research and Development Program of China, 2024YFF1306501, 2024YFF0810500

Increases in organic carbon within agricultural soils are widely recognized as a “negative emission” that removes CO2 from the atmosphere. Accurate quantification of soil organic carbon (SOC) to a certain depth in the spatial domain is critical for the effective implementation of improved land management practices in croplands. Currently, there is a lack of understanding regarding what depth strategy should be used to estimate SOC at 0–30 cm when sample datasets come from multiple depths. Furthermore, few studies have examined depth strategies for mapping SOC at the agricultural management level (i.e., field level), opting instead for point-based analysis. Here, three types of approaches with different depth strategies were evaluated for their ability to quantify 0–30 cm SOC content based on soil samples from 0–5 (surface), 5–30 (subsurface), and 0–30 cm (full column). These approaches involved the generalized additive model and machine learning techniques, i.e., artificial neural networks, random forest, and XGBoost. The soil samples used for the model evaluation and selection consisted of the newly collected samples in 2020–2022 and the Rapid Carbon Assessment (RaCA) legacy samples collected in 2010–2011. Environmental covariates corresponding to these SOC measurements were used in model training, including long-term physical climate, short-term weather, topographic and edaphic, and remotely sensed variables. Among the models evaluated in this study, the XGB regression model with a full column depth assignment strategy yielded the best prediction performance for 0–30 cm SOC content, with an r2 (squared Pearson correlation coefficient) of 0.48, an RMSE (root mean square error) of 0.29%, an ME (mean error) of 0.06%, an MAE of 0.25%, and an MEC (modeling efficiency coefficient) of 0.36 at the pixel level and an r2 of 0.64, an RMSE of 0.32%, an ME of −0.20%, an MAE of 0.28%, and an MEC of 0.48 at the field level. This study highlights that machine learning models with a full column depth strategy should be used to quantify 0–30 cm SOC content in agricultural soils over the continental United States (CONUS).

Pioneering work over the past decade has demonstrated the potential to map and monitor individual canopy trees using a wide range of sensor data, including digital airborne images, high-density terrestrial laser scanning, drone lidar, airborne imaging spectroscopy and multispectral satellite observations. Studies on Barro Colorado Island have identified canopy individuals of 11 species of tropical trees in 7 families. Improvements in spatial resolution, spectral resolution, spectral range, temporal frequency, and radiometric accuracy will increase the number of species that can be mapped and monitored using remotely sensed data. Over the next decade, whole-tree segmentation algorithms using very-high-density point clouds from lidar and digital photogrammetry will augment ground-based tree inventories, expanding the size of field-inventory plots and the types of measurements that are collected. Constellations of small satellites will enable analysis of tree populations throughout entire species ranges. Critical to all of these efforts is the acquisition of high-quality and spatially representative training data, as well as investment in automated algorithms with clearly defined sources of uncertainty.

Space-based laser altimetry has revolutionized our capacity to characterize terrestrial ecosystems through the direct observation of vegetation structure and the terrain beneath it. Data from NASA's ICESat-2 mission provide the first comprehensive look at canopy structure for boreal forests from space-based lidar. The objective of this research was to create ICESat-2 aboveground biomass density (AGBD) models for the global entirety of boreal forests at a 30 m spatial resolution and apply those models to ICESat-2 data from the 2019–2021 period. Although limited in dense canopy, ICESat-2 is the only space-based laser altimeter capable of mapping vegetation in northern latitudes. Along each ICESat-2 orbit track, ground and vegetation height is captured with additional modeling required to characterize biomass. By implementing a similar methodology of estimating AGBD as GEDI, ICESat-2 AGBD estimates can complement GEDI's estimates for a full global accounting of aboveground carbon. Using a suite of field measurements with contemporaneous airborne lidar data over boreal forests, ICESat-2 photons were simulated over many field sites and the impact of two methods of computing relative height (RH) metrics on AGBD at a 30 m along-track spatial resolution were tested; with and without ground photons. AGBD models were developed specifically for ICESat-2 segments having land cover as either Evergreen Needleleaf or Deciduous Broadleaf Trees, whereas a generalized boreal-wide AGBD model was developed for ICESat-2 segments whose land cover was neither. Applying our AGBD models to a set of over 19 million ICESat-2 observations yielded a 30 m along-track AGBD product for the pan-boreal. The ability demonstrated herein to calculate ICESat-2 biomass estimates at a 30 m spatial resolution provides the scientific underpinning for a full, spatially explicit, global accounting of aboveground biomass.

Drone lidar has the potential to provide detailed measurements of vertical forest structure throughout large areas, but a systematic evaluation of unsampled forest structure in comparison to independent reference data has not been performed. Here, we used ray tracing on a high-resolution voxel grid to quantify sampling variation in a temperate mountain forest in the southwest Czech Republic. We decoupled the impact of pulse density and scan-angle range on the likelihood of generating a return using spatially and temporally coincident TLS data. We show three ways that a return can fail to be generated in the presence of vegetation: first, voxels could be searched without producing a return, even when vegetation is present; second, voxels could be shadowed (occluded) by other material in the beam path, preventing a pulse from searching a given voxel; and third, some voxels were unsearched because no pulse was fired in that direction. We found that all three types existed, and that the proportion of each of them varied with pulse density and scan-angle range throughout the canopy height profile. Across the entire data set, 98.1% of voxels known to contain vegetation from a combination of coincident drone lidar and TLS data were searched by high-density drone lidar, and 81.8% of voxels that were occupied by vegetation generated at least one return. By decoupling the impacts of pulse density and scan angle range, we found that sampling completeness was more sensitive to pulse density than to scan-angle range. There are important differences in the causes of sampling variation that change with pulse density, scan-angle range, and canopy height. Our findings demonstrate the value of ray tracing to quantifying sampling completeness in drone lidar.

The velocity of climate change and its subsequent impact on vegetation has been well characterized at high elevations and latitudes, including the Arctic. But whether species and ecosystems are keeping pace with the velocity of temperature change is not as well documented. Some evidence indicates that species are less able to keep pace with the velocity of climate change along elevational gradients than latitudinal ones. If substantiated this finding could warrant reconsideration of a current cornerstone of conservation planning. Here we use 27 years of high-resolution satellite data to quantify changes in vegetation cover across elevation within nine mountain ranges in western North America, spanning tropical Mexico to subarctic Canada and from coastal California to interior deserts. Across these ranges we show a uniform pattern at the highest elevations in each range, where increases in vegetation have occurred ubiquitously over the past three decades. At these highest elevations, the realized velocity of vegetation varies among mountain ranges from 19.8–112.8 m · decade -1 (mean = 67.3 m · decade -1 ). This is equivalent, with respect to gradients in temperature, to a 14.4–104.3 km · decade -1 poleward shift (mean = 56.1 km · decade -1 ). This realized velocity is 4.4 times larger than previously reported for plants, and is among the fastest rates predicted for the velocity of climate change. However, in three of the five mountain ranges with long-term climate data, realized velocities fail to keep pace with changes in temperature, a finding with important implications for conservation of biological diversity.

Remote sensing of solar-induced chlorophyll fluorescence (SIF) advances our ability to monitor gross primary productivity. Because SIF is usually <5% of recorded canopy radiance, sources of measurement error must be reduced for accurate SIF retrieval. Here we quantify the impact of spectral stray light on SIF retrieval using a high-resolution imaging spectrometer. We first derived the FLD retrieval solution in the presence of stray light under the simplifying assumptions that SIF, reflectance, and stray light are constant across the region of the spectrum used in the retrieval. We show that stray light contributions from the canopy and reference spectra could cancel out if obtained from the same instrument. In practice however, the simplifying assumptions are unlikely to hold because SIF and reflectance vary across regions of the spectrum commonly exploited by the FLD retrieval, and stray light can also vary across the spectrum, as we demonstrate with an empirical example. To quantify the impact of spectral stray light on SIF retrievals in more realistic scenarios, we performed a sensitivity analysis. We used a stray light signal distribution function and measurements of spectral radiance and SIF to generate spectra with known quantities of SIF and stray light over four orders of magnitude of stray light. We then quantified the bias in retrieved SIF using four SIF retrieval approaches—standard FLD, 3FLD, a spectral fitting method (SFM), and a method based on singular value decomposition (SVD)—applied to the red and/or far-red spectral domains. We found that spectral stray light can cause either a positive or negative bias in SIF estimates. In the highest stray light scenario, with spectral stray light one order of magnitude smaller than the radiance signal, the stray light error ranged from < ± 1% of SIF (for O2-A band retrievals with FLD, 3FLD, and SFM methods, and for far-red retrievals in the 748 nm to 756 nm domain with SVD-based and 3FLD methods applied to Fraunhofer lines) to >30% of SIF (O2-B band retrieval with SFM, and 3FLD applied to Fraunhofer lines in the red domain). For context, a bias of 27% of SIF is a comparable magnitude to seasonal variation in SIF observed in the Amazon basin, and to SIF differences among plant functional types. However, when spectral stray light was three orders of magnitude smaller than the radiance signal, mean bias was < ± 2% for all retrieval methods. Our analysis informs when spectral stray light corrections must be performed for SIF remote sensing.

The Global Ecosystem Dynamics Investigation (GEDI) lidar is a multibeam laser altimeter on the International Space Station (ISS). GEDI is the first spaceborne instrument designed to measure vegetation height and to quantify aboveground carbon stocks in temperate and tropical forests and woodlands. This document describes the algorithm theoretical basis underpinning the development of the GEDI Level-4A (GEDI04_A) footprint aboveground biomass density (AGBD) data product. The GEDI04_A data product contains estimates of AGBD for individual GEDI footprints and associated prediction intervals. The algorithm uses GEDI02_A relative height (RH) metrics and 13 linear models to predict AGBD in 32 combinations of plant functional type (PFT) and world region within the observation limits of the ISS. GEDI04_A models for the release 1 and release 2 data products were developed using 8,587 quality-filtered simulated GEDI waveforms associated with field estimates of AGBD in 21 countries. Although this is the most geographically comprehensive data available for the development of AGBD models using lidar remote sensing, important regions are underrepresented, including the forests of continental Asia, deciduous broadleaf forests and savannas of the dry tropics, and evergreen broadleaf forests north of Australia. We describe the scientific and mathematical assumptions required to develop globally representative estimates of AGBD using GEDI lidar, including generalization beyond training data, and exclusion of GEDI02_A observations that do not meet requirements of the GEDI04_A algorithm. The footprint-level predictions generated by this process provide globally comprehensive estimates of AGBD. These footprint-level predictions are a prerequisite for the GEDI GEDI04_B gridded AGBD data product.

•Assessment of NASA GEDI biomass models at forest type level.•We used National Forest Inventory data for Spain for benchmarking.•Airborne laser data used to simulate GEDI laser full waveforms.•We present model strategies to correct GEDI L4A biomass estimates.•GEDI AGBD models perform reasonably in Mediterranean forests.

The Global Ecosystem Dynamics Investigation (GEDI) is a waveform lidar instrument on the International Space Station used to estimate aboveground biomass density (AGBD) in temperate and tropical forests. Algorithms to predict footprint AGBD from GEDI relative height (RH) metrics were developed from simulated waveforms with leaf-on (growing season) conditions. Leaf-off GEDI data with lower canopy cover are expected to have shorter RH metrics, and are therefore excluded from GEDI’s gridded AGBD products. However, the effects of leaf phenology on RH metric heights, and implications for GEDI footprint AGBD models that can include multiple nonlinear RH predictors, have not been quantified. Here, we test the sensitivity of GEDI data and AGBD predictions to leaf phenology. We simulated GEDI data using high-density drone lidar collected in a temperate mountain forest in the Czech Republic under leaf-off and leaf-on conditions, 51 d apart. We compared simulated GEDI RH metrics and footprint-level AGBD predictions from GEDI Level 4 A models from leaf-off and leaf-on datasets. Mean canopy cover increased by 31% from leaf-off to leaf-on conditions, from 57% to 88%. RH metrics < RH50 were more sensitive to changes in leaf phenology than RH metrics ⩾ RH50. Candidate AGBD models for the deciduous-broadleaf-trees prediction stratum in Europe that were trained using leaf-on measurements exhibited a systematic prediction difference of 0.6%–19% when applied to leaf-off data, as compared to leaf-on predictions. Models with the least systematic prediction difference contained only the highest RH metrics, or contained multiple predictor terms that contained both positive and negative coefficients, such that the difference from systematically shorter leaf-off RH metrics was partially offset among the multiple terms. These results suggest that, with consideration of model choice, leaf-off GEDI data can be suitable for AGBD prediction, which could increase data availability and reduce sampling error in some forests.

Habitat-suitability indices (HSI) have been employed in restoration to identify optimal sites for planting native species. Often, HSI are based on abiotic variables and do not include biotic interactions, even though similar abiotic conditions can favor both native and nonnative species. Biotic interactions such as competition may be especially important in invader-dominated habitats because invasive species often have fast growth rates and can exploit resources quickly. In this study, we test the utility of an HSI of microtopography derived from airborne LiDAR to predict post-disturbance recovery and native planting success in native shrub-dominated and nonnative, invasive grass-dominated dryland habitats in Hawai'i. The HSI uses high-resolution digital terrain models to classify sites' microtopography as high, medium, or low suitability, based on wind exposure and topographic position. We used a split-plot before-after-control-impact design to implement a disturbance experiment within native shrub (Dodonaea viscosa) and nonnative, invasive grass (Cenchrus clandestinus)-dominated ecosystems across three microtopography categories. In contrast to previous studies using the same HSI, we found that microtopography was a poor predictor of pre-disturbance conditions for soil nutrients, organic matter content, or foliar C:N, within both Dodonaea and Cenchrus vegetation types. In invader-dominated Cenchrus plots, microtopography helped predict cover, but not as expected (i.e., highest cover would be in high-suitability plots): D. viscosa had the greatest cover in low-suitability and C. clandestinus had the greatest cover in medium-suitability plots. Similarly, in native-dominated Dodonaea plots, microtopography was a poor predictor of D. viscosa, C. clandestinus, and total plant cover. Although we found some evidence that microtopography helped inform post-disturbance plant recovery of D. viscosa and total plant cover, vegetation type was a more important predictor. Important for considering the success of plantings, percent cover of D. viscosa decreased while percent cover of C. clandestinus increased within both vegetation types 20 months after disturbance. Our results are evidence that HSIs based on topographic features may prove most useful for choosing planting sites in harsh habitats or those already dominated by native species. In more productive habitats, competition from resident species may offset any benefits gained from "better" suitability sites.

The distribution of canopy heights in tropical rain forests directly affects carbon storage and the maintenance of biodiversity. We report results from a unique 20-yr record of annual monitoring of canopy-height distributions across an old-growth tropical rain forest landscape at the La Selva Biological Station in Costa Rica. Canopy heights to 15 m were measured annually in 18 0.50-ha plots at 231 points on a 5 x 5 m grid from 1999-2018 (nine plots in 1999), and heights >15 m were classified as "high canopy." During the study two major disturbance events (one immediately prior to the study) dominated the landscape-scale distribution of canopy heights. Height recovery from the 1997-1998 strong El Nino disturbance took approximately 15 yr. Frequency of canopy gaps varied an order of magnitude among years and 96% disappeared in <= 2 yr. High-canopy coverage and gap frequency varied substantially across the local gradients of soil nutrients and topography, and plot-level conditions and trends frequently differed from the landscape-level patterns. In contrast to the two major landscape-level disturbances, significant plot-level disturbances were common throughout two decades. Including a similar data set taken in 1992, canopy-height distributions for the last three decades over this old-growth tropical rain forest landscape are most parsimoniously interpreted as showing local disturbance and recovery and no unidirectional trends over time. Together these results suggest that understanding the landscape- and plot-level dynamics of tropical rain forest canopy-height distributions will require repeat sampling for multiple decades, while accurately measuring gap frequency and recovery will require sample intervals of <= 2 yr.

NASA’s Global Ecosystem Dynamics Investigation (GEDI) is collecting spaceborne full waveform lidar data with a primary science goal of producing accurate estimates of forest aboveground biomass density (AGBD). This paper presents the development of the models used to create GEDI’s footprint-level (~25 m) AGBD (GEDI04_A) product, including a description of the datasets used and the procedure for final model selection. The data used to fit our models are from a compilation of globally distributed spatially and temporally coincident field and airborne lidar datasets, whereby we simulated GEDI-like waveforms from airborne lidar to build a calibration database. We used this database to expand the geographic extent of past waveform lidar studies, and divided the globe into four broad strata by Plant Functional Type (PFT) and six geographic regions. GEDI’s waveform-to-biomass models take the form of parametric Ordinary Least Squares (OLS) models with simulated Relative Height (RH) metrics as predictor variables. From an exhaustive set of candidate models, we selected the best input predictor variables, and data transformations for each geographic stratum in the GEDI domain to produce a set of comprehensive predictive footprint-level models. We found that model selection frequently favored combinations of RH metrics at the 98th, 90th, 50th, and 10th height above ground-level percentiles (RH98, RH90, RH50, and RH10, respectively), but that inclusion of lower RH metrics (e.g. RH10) did not markedly improve model performance. Second, forced inclusion of RH98 in all models was important and did not degrade model performance, and the best performing models were parsimonious, typically having only 1-3 predictors. Third, stratification by geographic domain (PFT, geographic region) improved model performance in comparison to global models without stratification. Fourth, for the vast majority of strata, the best performing models were fit using square root transformation of field AGBD and/or height metrics. There was considerable variability in model performance across geographic strata, and areas with sparse training data and/or high AGBD values had the poorest performance. These models are used to produce global predictions of AGBD, but will be improved in the future as more and better training data become available.

Accurate estimation of aboveground forest biomass stocks is required to assess the impacts of land use changes such as deforestation and subsequent regrowth on concentrations of atmospheric CO2. The Global Ecosystem Dynamics Investigation (GEDI) is a lidar mission launched by NASA to the International Space Station in 2018. GEDI was specifically designed to retrieve vegetation structure within a novel, theoretical sampling design that explicitly quantifies biomass and its uncertainty across a variety of spatial scales. In this paper we provide the estimates of pan-tropical and temperate biomass derived from two years of GEDI observations. We present estimates of mean biomass densities at 1 km resolution, as well as estimates aggregated to the national level for every country GEDI observes, and at the sub-national level for the United States. For all estimates we provide the standard error of the mean biomass. These data serve as a baseline for current biomass stocks and their future changes, and the mission’s integrated use of formal statistical inference points the way towards the possibility of a new generation of powerful monitoring tools from space.