CoChat

The DOI matches its Crossref record; 56 of the 56 references it cites were found in a registry or an index; none of the 23 sources its claims cite is checked for support yet.

High Spatiotemporal Resolution River Networks Mapping on Catchment Scale Using Satellite Remote Sensing Imagery and DEM Data

Open access
Li, Peng, Zhang, Yun, Liang, Cunren, Wang, Houjie, Li, Zhenhong · 2024 · Geophysical Research Letters

Published version of a preprint

Abstract

Characterizing and understanding the changes in the flow regimes of rivers have been challenging. Existing global river network data sets are not updated and can only identify rivers wider than 30 m. We propose a novel automated method to map river networks on a monthly basin scale for the first time at 10‐m resolution using Sentinel‐1 Synthetic Aperture Radar, Sentinel‐2 multispectral images, and the AW3D30 Digital Surface Model. This method achieved an overall accuracy of 95.8%. The total length of the Yellow River network produced is 40,280 km, approximately 3.2 times that of the Global River Widths from Landsat (GRWL) database, more effectively covering small and medium rivers. The monthly river geometry revealed a positive correlation between river network area and precipitation. This study is expected to provide a cost‐effective alternative to accurately mapping global river networks and advance our understanding of the changes and drivers of river systems.

Overview

CoChat Summary

TL;DR

The paper presents an automated method that fuses Sentinel-1 SAR, Sentinel-2 multispectral imagery, and AW3D30 DSM to map monthly river networks at 10-m resolution across the Yellow River Basin. It reports 95.8% overall accuracy and a 40,280-km network, about 3.2 times the GRWL database, with monthly area changes positively correlated with precipitation.

Key findings

  • The method achieved an overall accuracy of 95.77% (abstract: 95.8%), with user accuracy 95.83% and producer accuracy 95.84%, based on 2,556 random “true water” and 2,430 random “true non-water” validation points.
  • The Yellow River network total length was 40,280 km; the abstract states this is approximately 3.2 times the GRWL database, while the discussion reports it as 28,587 km longer than all rivers in GRWL and 19,389 km longer than GRN. The OWF index was approximately 3.2 times that of GRWL, and drainage density was approximately three times higher than GRWL and GSW.
  • Monthly 2019 maps showed river density increasing from 0.038 km⁻¹ in January–February to 0.042 km⁻¹ in March–April, reaching an annual maximum of 0.053 km⁻¹ in July with total length 40,280 km. Minimum network area was 8,306 km² in January and maximum was 10,267 km² in August; river network area was positively correlated with precipitation and primarily controlled by seasonal precipitation because monthly precipitation is 10–20 times evaporation.
  • The method better suppressed shadow noise in urban high-rise and mountainous areas and extracted more small and medium rivers than MNDWI, Zou et al. (2018), and APWC; connected rivers of order ≥4 were longer than in GRWL and GRN.

Methods

  • Study area: Yellow River Basin; period January–December 2019; data included 6,357 Sentinel-2 images, Sentinel-1 GRD supplements, AW3D30 DSM, GSW, ESRI 2020 Land Cover, and ERA5 climate data.
  • Water extraction: cloud/shadow masking and monthly median composites; SNIC superpixels with a decision tree trained on 10,093 water and 9,905 non-water samples; threshold rule ((AWEIₙₛₕ > 0.6 or AWEIₛₕ > 0) and (MNDWI > EVI or MNDWI > NDVI)); SAR refined Lee filter, SDWI, and OTSU thresholding; paddy fields removed using NDVI time series.
  • River network extraction: RivWidthCloud centerlines; AW3D30 DSM depression filling, D8 flow direction, confluence accumulation, and buffers to connect fragmented networks; accuracy assessed with random points and compared with GRWL, GSW, and GRN; climate correlations used ERA5 precipitation, temperature, and evaporation.

Limitations

  • Stated: river networks were most accurate and complete when river width was greater than 30 m, broken river lines may exist in areas less than 30 m wide, and ice/snow cover effects on extraction errors cannot be completely eliminated in the Qinghai-Tibet Plateau. The GSW data set contained artificial wetlands not considered part of river networks, so its OWF index was higher than the study’s results.
  • Evident: the demonstration is limited to one basin (Yellow River) and one year (2019), so global applicability and long-term performance are not established.
  • Evident: monthly mapping depends on satellite data availability and cloud gaps, and the threshold rules were calibrated for this region; potential uncertainty from DEM vertical errors and transferability of thresholds is not quantified.

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

Integrity signals

Registry

Crossref · 10.1029/2023gl107956

Cites

56 of 56 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Location of the study area. (a) Upper reaches, (b) middle reaches, and (c) lower reaches in the Yellow River Basin were derived from Sentinel-2 RGB imagery. Note that, RGB bands include band 4 (red), band 3 (green), and band 2 (blue).
Figure 1 Location of the study area. (a) Upper reaches, (b) middle reaches, and (c) lower reaches in the Yellow River Basin were derived from Sentinel-2 RGB imagery. Note that, RGB bands include band 4 (red), band 3 (green), and band 2 (blue).CC-BY-NC-ND · From the PDF
Image 2
CC-BY-NC-ND · From the PDF
Figure 2 Workflow of river networks extraction in the Yellow River Basin.
Figure 2 Workflow of river networks extraction in the Yellow River Basin.CC-BY-NC-ND · From the PDF
Image 4
CC-BY-NC-ND · From the PDF
Figure 3 Comparisons of the results of this method with other methods. Typical city (a) and mountain area (b) correspond to Figures 1a and 1b; (c) and (g) show the results of our method; (d) and (h) present the results of MNDWI; (e) and (i) are the results of Zou et al. (2018); (f) and (j) are the results of the APWC method proposed by Slinski et al. (2019).
Figure 3 Comparisons of the results of this method with other methods. Typical city (a) and mountain area (b) correspond to Figures 1a and 1b; (c) and (g) show the results of our method; (d) and (h) present the results of MNDWI; (e) and (i) are the results of Zou et al. (2018); (f) and (j) are the results of the APWC method proposed by Slinski et al. (2019).CC-BY-NC-ND · From the PDF
Image 6
CC-BY-NC-ND · From the PDF
Figure 4 Comparisons with different river data sets and our results. The base map shows Yellow River networks results and the existing the Global River Widths from the Landsat (GRWL) database. (a), (c), and (e) Water surface results of the existing Global Surface Water (GSW) data set. (b), (d), and (f) Water surface results of our method. (g) Confusion matrix for the automated accuracy assessment of our method. (h) Comparison of the river length (km) in the results of this study, GRWL data set and Global River Networks (GRN) data set. (i) The relationship between the surface area (yellow line), precipitation (blue column), temperature (red line), and evaporation (green line) in the Yellow River Basin.
Figure 4 Comparisons with different river data sets and our results. The base map shows Yellow River networks results and the existing the Global River Widths from the Landsat (GRWL) database. (a), (c), and (e) Water surface results of the existing Global Surface Water (GSW) data set. (b), (d), and (f) Water surface results of our method. (g) Confusion matrix for the automated accuracy assessment of our method. (h) Comparison of the river length (km) in the results of this study, GRWL data set and Global River Networks (GRN) data set. (i) The relationship between the surface area (yellow line), precipitation (blue column), temperature (red line), and evaporation (green line) in the Yellow River Basin.CC-BY-NC-ND · From the PDF
Image 8
CC-BY-NC-ND · From the PDF

Verification

Claims the paper makes

0 Valid · 3 Not valid · 16 Not checked · 54 Own results and interpretations

44 more claims

Create a free account to see all 73 claims and to verify the 15 references still open.

See all claims — free account
  • It may state two separate claims as one. 16
  • Probably not a substantial claim. 11
  • The quote only partly states this claim. 1

Create a free account to see these claims.

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.

3 of 3

Citations

Cites

Show all 55 references

Research works printed but not linked: 1

  1. 10.16232/j.cnki.1001‐4179.2019.02.038 · doi:10.16232/j.cnki.1001-4179.2019.02.038

Cited by

Metadata

Authors
Li, P., Zhang, Y., Liang, C., Wang, H., & Li, Z.
Affiliations
  • Qingdao National Laboratory for Marine Science and Technology (CN)
  • Peking University (CN)
  • Chang'an University (CN)
Corresponding author
Wang, Houjie
Venue
Geophysical Research Letters
Published
2024-03-22
Volume
51
Issue
6
ISSN
  • 0094-8276 (ISSN-L)
  • 1944-8007
OpenAlex
W4393090964
Cited by
30
References
55
Citations by year
2026: 18 · 2025: 10 · 2024: 2
FWCI
3.20
Type
journal article
Language
en
Open access
Licence
Locations
  • Geophysical Research Letters (publishedVersion, cc-by-nc-nd)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
National Natural Science Foundation of China (42041005‐4)
Topics
  • Physical Sciences › Environmental Science › Global and Planetary Change › Flood Risk Assessment and Management
  • Physical Sciences › Environmental Science › Water Science and Technology › Hydrology and Watershed Management Studies
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Automated Road and Building Extraction
Keywords
river network mapping, high spatial and temporal resolution, Sentinel-1 SAR, Sentinel-2 imagery, small and medium rivers, Yellow River, flow regime alteration

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.

A language model drew the claims from the paper on 2026-11-18, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-18 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

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.