GloWE-8D: a global long-term 8-day wind erosion dataset from 1982 to 2020

Zhang, Hanbing; Peng, Jian; Dong, Jianquan; Yang, Zhiwei; Lin, Yifan; Liu, Xuebang

doi:10.5194/essd-2026-66

Preprints

https://doi.org/10.5194/essd-2026-66

Preprints

10 Feb 2026

| 10 Feb 2026

Status: this preprint is currently under review for the journal ESSD.

GloWE-8D: a global long-term 8-day wind erosion dataset from 1982 to 2020

Hanbing Zhang, Jian Peng, Jianquan Dong, Zhiwei Yang, Yifan Lin, and Xuebang Liu

Abstract. Wind erosion constitutes a critical driver of global land degradation and dust emission, posing persistent threats to ecological security, agricultural production, and human health. Although regional-scale wind erosion assessments exist, there remains a lack of long-term, high spatiotemporal resolution, and publicly available global-scale wind erosion datasets, which has constrained a deeper understanding of its dynamic processes and driving mechanisms. Based on the Revised Wind Erosion Equation (RWEQ), this study constructed a global wind erosion dataset from 1982 to 2020 with an 8-day temporal resolution and a 0.05° spatial resolution. By introducing a residue factor scheme based on growing season identification, the characterization accuracy of wind erosion suppression during vegetation cover periods was enhanced, enabling a more refined depiction of episodic wind erosion features. The dataset revealed that the global annual average wind erosion total from 1982 to 2020 was 539.13 Pg, with severe erosion areas concentrated in the arid and semi-arid regions of the Northern Hemisphere. The wind erosion exhibited a slowly increasing trend, although with significant regional variations. Data validation demonstrated a high spatial consistency between this dataset and the MERRA-2 dust emission data (R²= 0.79), and a significant temporal correlation with coarse-mode aerosol optical depth observations from AERONET stations. Furthermore, comparisons indicated that the results of this study were within the same order of magnitude and showed high correlation with existing regional research. As the first publicly available long-term, high spatiotemporal resolution global wind erosion data product, this dataset provides crucial data support for global and regional dust emission estimation, research on wind erosion process mechanisms, land degradation prevention, and climate change response. The dataset is publicly accessible at https://zenodo.org/records/18245214 (Zhang et al., 2026).

Received: 25 Jan 2026 – Discussion started: 10 Feb 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Hanbing Zhang, Jian Peng, Jianquan Dong, Zhiwei Yang, Yifan Lin, and Xuebang Liu

Status: final response (author comments only)

RC1:
'Comment on essd-2026-66', Anonymous Referee #1, 03 Mar 2026
The authors have successfully constructed and made publicly available the long-term, high spatiotemporal resolution global wind erosion dataset, filling a critical gap in the field. The manuscript is well-written, logically structured, scientifically sound in its methodology, and thorough in its validation. The results are reliable and the discussion is insightful. The public availability of this dataset will provide invaluable support for research in wind erosion, dust cycles, land degradation, and climate change. Suggested to be published after revisions. The following are the questions and some mistakes in this manuscript:
The introduction does not explicitly state the research objective of this manuscript.

The abstract highlights the novel “residue factor scheme based on growing season identification.” To better prepare the reader, it would be beneficial to briefly mention this core improvement at the end of the introduction. This would frame it as the key solution to addressing the limitations of previous models in simulating wind erosion suppression during vegetated periods.

Equation (5) assumes a constant threshold wind speed of 5 m/s. This is a critical parameter known to vary significantly with surface characteristics globally. Please provide a justification for this single value (e.g., citing key literature that supports this as a global average) or discuss the potential uncertainties introduced by this simplification.

The use of a “50% threshold” to define growing season timing is stated. While a reference is provided, it would strengthen the methodology to briefly explain why this specific threshold was chosen and comment on its applicability across diverse global ecosystems (e.g., forests vs. grasslands).

Equation (6) is used twice. Please renumber the subsequent equations to correct this.

The Wind Erosion Concentration Period (WECP) is defined as the peak period plus the five preceding and five subsequent periods (an ~88-day window). What is the basis for selecting this specific window size? Was a sensitivity analysis performed, or is it based on the typical duration of wind erosion events in key regions? A brief explanation would increase methodological transparency.

Table 1 shows some notable differences for specific regions (e.g., Mu Us Sandy Land, Mongolia). While differences in data sources are mentioned, could a more specific hypothesis be offered for these particular cases? For instance, does the higher spatial resolution of this study capture localized, high-erosion hotspots that coarser studies averaged out? A more targeted discussion would strengthen the analysis of the dataset's characteristics.

The discussion on applications for climate models and policy is good but could be slightly expanded. A concrete example of how the 8-day data could be used to improve dust emission parameterizations in a global model, or how it could inform the timing of seasonal sand-fixing measures, would powerfully illustrate the dataset's practical utility.

Please improve English expression in the manuscript.
Citation: https://doi.org/10.5194/essd-2026-66-RC1
RC2: 'Comment on essd-2026-66', Anonymous Referee #2, 15 Mar 2026

This manuscript presents GloWE-8D, a global wind erosion dataset spanning 1982–2020 at 8-day temporal resolution and 0.05° spatial resolution, derived from the Revised Wind Erosion Equation (RWEQ). The topic is timely and the ambition to produce a long-term, high-spatiotemporal-resolution global wind erosion product is commendable. However, the manuscript has substantial deficiencies in uncertainty quantification that prevent it from meeting ESSD's standards for a data description paper
Minor Concerns
The claim that GloWE-8D is "the first publicly available long-term, high spatiotemporal resolution global wind erosion data product" should be more carefully substantiated, given that Chu et al. (2024), Sun et al. (2024), and Yang et al. (2021) are cited as closely related global products. A clearer differentiation in terms of temporal coverage, resolution, and open accessibility is warranted.
Major Concern: Inadequate Uncertainty Quantification
ESSD explicitly requires that data description papers provide transparent uncertainty assessment. This manuscript does not adequately fulfil this requirement. The RWEQ framework involves numerous parameters and input datasets, each carrying its own uncertainties, yet no systematic uncertainty analysis is presented. The following specific issues require attention:
1 Threshold Wind Speed
The universal threshold wind speed of 5 m/s is applied uniformly across all land surface types and geographic regions globally. The authors acknowledge in Section 4.3 that this parameter exhibits spatial heterogeneity depending on land use type, surface roughness, and topsoil conditions, yet no sensitivity analysis is provided to assess its impact on regional or global wind erosion estimates. Given that the threshold wind speed fundamentally controls whether erosion is initiated at all, this is arguably the single most consequential simplification in the entire model chain. Even a basic sensitivity test would substantially strengthen confidence in the dataset.
2 Propagation of Input Data Uncertainties
The dataset integrates multiple input products — ERA5 meteorological reanalysis, GLASS FVC, GLC_FCS30D land use, SoilGrids soil properties, and Copernicus DEM — each carrying their own spatial and temporal uncertainties. No formal uncertainty propagation is attempted across these inputs. Furthermore, the temporal homogeneity of the 39-year record deserves scrutiny: ERA5 assimilates different observational data streams over time, and the GLC_FCS30D land use product has limited temporal resolution prior to 2000, with the authors using the temporally nearest available data as a substitute. These discontinuities could introduce spurious trends or step changes in the wind erosion time series that are artefacts of input data transitions rather than genuine environmental signals. The authors should address this risk explicitly, ideally by examining whether identifiable breakpoints in the time series coincide with known transitions in the input data record.

Citation: https://doi.org/10.5194/essd-2026-66-RC2
CC1: 'Comment on essd-2026-66', Xuesong Wang, 24 Mar 2026

Publisher’s note: the content of this comment was removed on 27 March 2026 since the comment was posted by mistake.

Citation: https://doi.org/10.5194/essd-2026-66-CC1
RC3: 'Comment on essd-2026-66', Anonymous Referee #3, 20 Apr 2026

This study presents a potentially useful global wind-erosion dataset with long temporal coverage and relatively high spatiotemporal resolution. The dataset could be valuable for future research on wind erosion, dust emission, and land degradation. However, several issues still require clarification or further improvement:
Lines 100–105: ERA5 meteorological data are typically provided at 0.25°, whereas ERA5-Land is provided at 0.1°. Please clarify the exact source and spatial resolution of the meteorological inputs used in this study. Since the Weather Factor is a major driver in RWEQ, and the meteorological data appear to have been resampled to 0.05°, it is unclear whether the resulting wind-erosion product can truly be regarded as a 0.05° assessment?
Line 106: For the GLASS FVC product at 0.05°, please specify which sensor-based product was used, i.e., AVHRR_0.05° or MODIS_0.05°. This is important because the MODIS-based product is only available after 2000.
Line 141: Please specify the value assigned to the distance from the upwind edge (x).
Line 149: Please provide the reference for converting 10 m wind speed to 2 m wind speed (Equation 6).
Line 154: The numbering of Equation (6) appears to be incorrect. In addition, this equation seems to be used to convert daily wind speed into hourly wind speed. However, ERA5 and ERA5-Land both provide hourly wind data. Please clarify why hourly data were not used directly and justify the need for this conversion.
Lines 202–203: The formatting of β is inconsistent, appearing as upright and italic in different places. Please standardize the formatting of this symbol and check the consistency of other variables throughout the manuscript.
Line 328: RWEQ simulates surface wind erosion, whereas dust-related datasets reflect atmospheric signals after emission, uplift, transport, dispersion, and deposition. Therefore, such datasets can only provide indirect validation rather than direct validation of modeled wind erosion itself. Please discuss this limitation and the associated uncertainties more explicitly.
Line 353: The discussion here appears to compare the present RWEQ-based results mainly with previous RWEQ-based studies. This type of comparison is useful for showing broad consistency, but it is closer to inter-model consistency than to independent validation. Please clarify the significance and limitations of this comparison. If possible, it would be helpful to include more direct observational evidence for validation.
Line 359: Is there direct evidence supporting this explanation? If not, this should be presented as a plausible interpretation rather than a near-confirmed mechanism. Please consider revising the wording accordingly.
Line 360: A finer spatial resolution may indeed improve the identification of local hotspots, but it may also better distinguish low-value areas or areas where wind erosion does not actually occur. Therefore, its effect on the global total is not necessarily unidirectional. I suggest presenting this point more cautiously.
Units of total wind erosion: When reporting total wind erosion in Pg, I suggest explicitly indicating the time scale, for example, Pg/year, to avoid ambiguity.
Figure 4a: Relatively humid or highly vegetated regions also appear to show widespread wind-erosion signals. Could the authors please clarify whether this might indicate a possible overestimation of the spatial extent of wind erosion?
Figure 5a: It would be helpful to explain the meaning of the x-axis numbers in the figure caption, since the current labeling is not immediately clear.

Citation: https://doi.org/10.5194/essd-2026-66-RC3

Hanbing Zhang, Jian Peng, Jianquan Dong, Zhiwei Yang, Yifan Lin, and Xuebang Liu

Data sets

GloWE-8D: a global long-term 8-day wind erosion dataset from 1982 to 2020 Hanbing Zhang https://doi.org/10.5281/zenodo.18245214

Hanbing Zhang, Jian Peng, Jianquan Dong, Zhiwei Yang, Yifan Lin, and Xuebang Liu

Viewed

Total article views: 520 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
297	199	24	520	32	39

HTML: 297
PDF: 199
XML: 24
Total: 520
BibTeX: 32
EndNote: 39

Views and downloads (calculated since 10 Feb 2026)

Month	HTML	PDF	XML	Total
Feb 2026	98	55	10	163
Mar 2026	104	78	8	190
Apr 2026	94	65	6	165
May 2026	1	1	0	2

Cumulative views and downloads (calculated since 10 Feb 2026)

Month	HTML	PDF	XML	Total
Feb 2026	98	55	10	163
Mar 2026	104	78	8	190
Apr 2026	94	65	6	165
May 2026	1	1	0	2

Viewed (geographical distribution)

Total article views: 533 (including HTML, PDF, and XML) Thereof 533 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 05 May 2026

Short summary

We produced a global long-term 8-day wind erosion dataset (GloWE-8D) with a spatial resolution of 0.05°, spanning from 1982 to 2020. Based on the Revised Wind Erosion Equation (RWEQ) and integrating multi-source data. GloWE-8D is the first publicly available global wind erosion product, offering reliable support for dust emission estimation, wind erosion mechanism studies, and land degradation monitoring at both global and regional scales.


Total:	0
HTML:	0
PDF:	0
XML:	0