Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 77 of the 77 references it cites were found in a registry or an index; 1 of the 18 sources its claims cite were checked for support.

A Global Evaluation of Radar‐Derived Digital Elevation Models: SRTM, NASADEM, and GLO‐30

Open access
Simard, Marc, Denbina, Michael, Marshak, Charles, Neumann, Maxim · 2024 · Journal of Geophysical Research: Biogeosciences

Abstract

This study evaluates global radar‐derived digital elevation models (DEMs), namely the Shuttle Radar Topography Mission (SRTM), NASADEM and GLO‐30 DEMs. We evaluate their accuracy over bare‐earth terrain and characterize elevation biases induced by forests using global Lidar measurements from the Ice, Cloud, and Land Elevation Satellite (ICESat)'s Geoscience Laser Altimeter System (GLAS), the Global Ecosystem Dynamics Investigation (GEDI) and the ICESat‐2 Advanced Topographic Laser Altimeter System (ATLAS) instruments collected on locally flat terrain. Our analysis is based on error statistics calculated for each DEM tile, which are then summarized as global error percentiles, providing a regional characterization of DEM quality. We find NASADEM to be a significant improvement upon the SRTM V3. Over bare ground areas, the mean elevation bias and root mean square error (RMSE) improved from 0.68 to 2.50 m respectively to 0.00 and 1.5 m as compared to ICESat/GLAS. GLO‐30 is more accurate with bare ground elevation bias and RMSE were below 0.05 and 0.55 m. Similar improvements were observed when compared to GEDI and ICESat‐2 measurements. The DEM biases associated with the presence of vegetation vary linearly with canopy height, and more closely follow the percentile of Lidar Relative Height (RH50). Other factors such as canopy density, radar frequency and Lidar technology also contribute to observed elevation biases. This global analysis highlights the potential of various technologies for mapping of Earth's topography, and the need for more advanced remote sensing observations that can resolve vegetation structure and sub‐canopy ground elevation.

Overview

CoChat Summary

TL;DR

This study globally evaluates SRTM V3, NASADEM, and GLO-30 radar-derived DEMs against spaceborne Lidar from ICESat/GLAS, GEDI, and ICESat-2/ATLAS over bare ground and vegetated terrain. NASADEM improves substantially on SRTM V3, GLO-30 is the most accurate over bare ground, and vegetation biases scale with canopy height and are generally closest to RH50.

Key findings

  • Over bare ground versus ICESat, NASADEM reduced mean bias and RMSE from SRTM V3’s 0.68 m and 2.50 m to 0.00 m and 1.50 m; GLO-30 had mean bias 0.05 m and RMSE 0.55 m. Similar improvements appeared versus GEDI and ICESat-2: GLO-30 mean bias/RMSE was 0.15/0.92 m versus GEDI and 0.07/0.69 m versus ICESat-2, while NASADEM was 0.06/1.81 m versus GEDI and 0.01/1.60 m versus ICESat-2.
  • NASADEM’s mean tile bias was more than 10 times lower than SRTM V3 for all Lidar data sets; mean RMSE was 1 m lower than SRTM V3 with ICESat and ICESat-2, and 0.86 m lower with GEDI. For 99% of tiles versus ICESat, SRTM V3 had MAE below 4.69 m and RMSE below 4.93 m, while NASADEM had equivalent values of 2.08 m and 2.50 m.
  • GLO-30 was about 2 times more accurate than NASADEM. Across all Lidar data sets, GLO-30 mean bias and RMSE ranged between 5–15 cm and 55–92 cm, respectively; 90% of GLO-30 tiles had MAE less than 1 m, lower by 92 cm than NASADEM’s 1.93 m, and 99% of GLO-30 tiles had RMSE never exceeding 2 m. Versus GEDI, 80% of GLO-30 tiles had absolute bias less than 50 cm and RMSE less than 1.16 m.
  • Vegetation biases were positively correlated with canopy height and generally closest to RH50. NASADEM bias versus ICESat RH50 had slope 0.87 m/m and intercept 1.52 m; GLO-30 bias versus ICESat RH50 had slope 1.10 m/m and intercept 2.21 m. GLO-30 vegetation bias tended to be larger than SRTM or NASADEM, likely due to shallower X-band microwave penetration and snow during SRTM acquisition.

Methods

  • Design: global validation of SRTM V3, NASADEM, and GLO-30 against ICESat/GLAS, GEDI, and ICESat-2/ATLAS Lidar, performed separately over bare ground and vegetated footprints. The analysis focused on locally flat terrain, excluding slopes >5°, glaciers, forest clearings/growth, water, snow, and the 1st percentile of Lidar points with highest absolute deviation from the DEM.
  • Data/sample: SRTM V3, NASADEM, and GLO-30 DEMs; ICESat GLA14, GEDI L2A first 8 months, and ICESat-2 ATL08. Bare ground was defined as <10% tree cover and vegetated as >30% tree cover. Tiles were 1° × 1° and included only if they contained at least 1,000 ICESat or GEDI footprints, or at least 500 ICESat-2 footprints.
  • Analysis: each Lidar elevation estimate was compared with the DEM pixel containing the footprint center. Per-tile mean bias, absolute bias, and RMSE were calculated and summarized as global percentiles. Vegetation bias was modeled as y = h_DEM − h_Lidar Ground versus x = h_RH Metric using linear fits, with slope, intercept, r², and RMSD reported for RH50, RH75, and RH100/RH98.

Limitations

  • The paper states that the analysis focuses on nearly flat local terrain and does not include evaluation of radar-derived DEM performance over slopes, which is best assessed with airborne Lidar surveys. It also states that the global linear fits are not necessarily representative of specific forest types, and that comparing canopy height and ground-finding accuracy among ICESat, ICESat-2, and GEDI should be a follow-up study.
  • The text makes evident additional constraints: footprint selection excluded large regions in tropical rain forests and mountainous areas; ICESat cloud cover can obstruct forest analysis near the equator; GEDI used only the first 8 months of data; full-waveform Lidars have limited ground detection in tropical forests; GEDI may not accurately measure ground elevation in tall (>60 m) canopies; and ICESat-2’s photon-counting Lidar has different canopy penetration and ground-detection characteristics.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1029/2023jg007672

Cites

77 of 77 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Correct label 'e' Example of height differences between SRTM V3 and NASADEM over the Florida Everglades, U.S. (left column) and New South Wales, Australia (right column). Heights are with respect to the EGM96 geoid. SRTM V3 elevation in (a) the Everglades and (b) New South Wales. NASADEM elevation in (c) the Everglades and (d) New South Wales. Difference NASADEM-SRTM V3 in (e) the Everglades and (f) New South Wales. Height ripples and artifacts can be observed in the SRTM elevation images, which have been corrected in NASADEM. Also note the different colorbars in the Florida elevation maps are due to the bias in the SRTM data corrected in the NASADEM.
Figure 1 Correct label 'e' Example of height differences between SRTM V3 and NASADEM over the Florida Everglades, U.S. (left column) and New South Wales, Australia (right column). Heights are with respect to the EGM96 geoid. SRTM V3 elevation in (a) the Everglades and (b) New South Wales. NASADEM elevation in (c) the Everglades and (d) New South Wales. Difference NASADEM-SRTM V3 in (e) the Everglades and (f) New South Wales. Height ripples and artifacts can be observed in the SRTM elevation images, which have been corrected in NASADEM. Also note the different colorbars in the Florida elevation maps are due to the bias in the SRTM data corrected in the NASADEM.CC-BY · From the PDF
Image 2
CC-BY · From the PDF
Figure 2 The log-scaled number of bare ground footprints remaining within each 1 ° × 1 ° tile after filtering, for (a) ICESat, (b) ICESat-2, and (c) GEDI. All tiles with at least 10 Lidar footprints are shown. Only tiles with 1,000 (i.e., 3 in log scale) are analyzed.
Figure 2 The log-scaled number of bare ground footprints remaining within each 1 ° × 1 ° tile after filtering, for (a) ICESat, (b) ICESat-2, and (c) GEDI. All tiles with at least 10 Lidar footprints are shown. Only tiles with 1,000 (i.e., 3 in log scale) are analyzed.CC-BY · From the PDF
Image 4
CC-BY · From the PDF
Figure 3 Global DEMs versus ICESat bias for each 1 ° × 1 ° tile, evaluated over bare ground areas for (a) SRTM V3, (b) NASADEM and (c) GLO-30 DEM.
Figure 3 Global DEMs versus ICESat bias for each 1 ° × 1 ° tile, evaluated over bare ground areas for (a) SRTM V3, (b) NASADEM and (c) GLO-30 DEM.CC-BY · From the PDF
Image 6
CC-BY · From the PDF
Figure 4 Global map of NASADEM versus ICESat RMSE for each NASADEM tile, evaluated over bare ground areas.
Figure 4 Global map of NASADEM versus ICESat RMSE for each NASADEM tile, evaluated over bare ground areas.CC-BY · From the PDF
Image 8
CC-BY · From the PDF
Figure 5 Global map of bias between NASADEM and (a) GEDI and (b) ICESat-2 for each NASADEM tile, evaluated over bare ground areas.
Figure 5 Global map of bias between NASADEM and (a) GEDI and (b) ICESat-2 for each NASADEM tile, evaluated over bare ground areas.CC-BY · From the PDF
Figure 6 Global map of tiled RMSE between NASADEM and (a) GEDI and (b) ICESat-2 evaluated over bare ground areas.
Figure 6 Global map of tiled RMSE between NASADEM and (a) GEDI and (b) ICESat-2 evaluated over bare ground areas.CC-BY · From the PDF
Image 11
CC-BY · From the PDF
Figure 7 Global map of (a) NASADEM and (b) GLO-30 DEM mean vegetation bias in each tile compared to ICESat. Panel (c) shows the mean ICESat ground return energy in each tile. All analysis is based on vegetated footprints only.
Figure 7 Global map of (a) NASADEM and (b) GLO-30 DEM mean vegetation bias in each tile compared to ICESat. Panel (c) shows the mean ICESat ground return energy in each tile. All analysis is based on vegetated footprints only.CC-BY · From the PDF
Image 13
CC-BY · From the PDF
Figure 8 Global map of NASADEM mean vegetation bias in each tile compared to (a) ICESat-2 and (b) GEDI. All analysis is based on vegetated footprints only.
Figure 8 Global map of NASADEM mean vegetation bias in each tile compared to (a) ICESat-2 and (b) GEDI. All analysis is based on vegetated footprints only.CC-BY · From the PDF
Image 15
CC-BY · From the PDF
Figure 9 Global map of GLO-30 mean vegetation bias in each tile compared to (a) ICESat-2 and (b) GEDI. All analysis is based on vegetated footprints only.
Figure 9 Global map of GLO-30 mean vegetation bias in each tile compared to (a) ICESat-2 and (b) GEDI. All analysis is based on vegetated footprints only.CC-BY · From the PDF
Figure 10 Correct legend 'GLO-30' DEM elevation relative to the Lidar ground calculated for 10 m canopy height bins. The marker shows the mean value μ of the DEM elevation over the Lidar ground and the error bars show μ ± σ , with σ the standard deviation of the heights in each bin.
Figure 10 Correct legend 'GLO-30' DEM elevation relative to the Lidar ground calculated for 10 m canopy height bins. The marker shows the mean value μ of the DEM elevation over the Lidar ground and the error bars show μ ± σ , with σ the standard deviation of the heights in each bin.CC-BY · From the PDF
Image 18
CC-BY · From the PDF

Verification

Claims the paper makes

1 Valid · 0 Not valid · 13 Not checked · 80 Own results and interpretations

47 more claims

Create a free account to see all 94 claims and to verify the 14 references still open.

See all claims — free account
  • It may state two separate claims as one. 25
  • Probably not a substantial claim. 24

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 76 references

Research works printed but not linked: 5

  1. Buckley S., 2020, NASADEM user guide
  2. Donnellan A., 2021, NASA surface topography and vegetation incubation study, 210
  3. Gallant J., 2009, Proc. geomorphometry, 149
  4. JPL N., 2020, NASADEM Merged DEM Global 1 arc second V001
  5. NISAR, 2018, NASA‐ISRO SAR (NISAR) mission science users’ handbook

Cited by

Metadata

Authors
Simard, M., Denbina, M., Marshak, C., & Neumann, M.
Affiliations
  • Jet Propulsion Laboratory (US)
  • Jet Propulsion Laboratory California Institute of Technology Pasadena CA USA
Corresponding author
Simard, Marc
Venue
Journal of Geophysical Research: Biogeosciences
Published
2024-10-29
Volume
129
Issue
11
ISSN
  • 2169-8953 (ISSN-L)
  • 2169-8961
OpenAlex
W4403879929
Cited by
54
References
76
Citations by year
2026: 38 · 2025: 15 · 2024: 1
FWCI
4.85
Type
journal article
Language
en
Open access
Locations
Journal of Geophysical Research Biogeosciences (publishedVersion, cc-by)
Funders
  • National Aeronautics and Space Administration (ICESat-2, 80NM0018D0004)
  • California Institute of Technology (80 NM0018D0004)
  • Jet Propulsion Laboratory (80NM0018D0004)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Satellite Image Processing and Photogrammetry
  • Physical Sciences › Earth and Planetary Sciences › Geophysics › Synthetic Aperture Radar (SAR) Applications and Techniques
  • Life Sciences › Agricultural and Biological Sciences › Forestry › Remote Sensing and LiDAR Applications
Keywords
SRTM, NASADEM, DEM accuracy assessment, bare-earth elevation, elevation bias, ICESat/GLAS, GEDI, canopy height, canopy density, radar frequency, global topography

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-11-19, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-19 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.