Zang et al. (2026) High-Precision River Network Mapping Using River Probability Learning and Adaptive Stream Burning

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Remote Sensing
• Year: 2026
• Date: 2026-01-21
• Authors: Yufu Zang, Zhaocai Chu, Zhen Cui, Zhuokai Shi, Qihan Jiang, Yichen Shen, Jue Ding
• DOI: 10.3390/rs18020362

Research Groups

Not explicitly stated in the provided text.

Short Summary

This study proposes a novel method integrating a river-oriented Gradient Boosting Tree model and an adaptive stream burning algorithm for high-precision and topologically consistent river network extraction. The method achieves superior positional accuracy and network continuity, particularly for complex river morphologies, with errors predominantly within a 0–40 meter range.

Objective

• To develop a novel method integrating a river-oriented Gradient Boosting Tree model (RGBT) and an adaptive stream burning algorithm to overcome limitations of conventional methods (river network discontinuity and poor representation of morphologically complex rivers) for high-precision and topologically consistent river network extraction.

Study Configuration

• Spatial Scale: Regional (evaluated across eight representative regions with diverse topographical characteristics).
• Temporal Scale: Static (focuses on mapping river networks at a specific point in time).

Methodology and Data

• Models used: River-oriented Gradient Boosting Tree model (RGBT), direction-constrained region growing strategy, adaptive stream burning algorithm.
• Data sources: Water-oriented multispectral indices (derived from multispectral imagery), multi-scale linear geometric features, digital elevation models (DEMs).

Main Results

• The proposed method achieves the highest positional accuracy and network continuity compared to reference networks and mainstream hydrographic products.
• Positional errors are mainly concentrated within a 0–40 meter range.
• Significant improvements are observed for narrow tributaries, highly meandering rivers, and braided channels.
• The method provides a reliable solution for high-resolution river network mapping in complex environments.

Contributions

• Introduction of a novel integrated method combining a river-oriented Gradient Boosting Tree model and an adaptive stream burning algorithm for river network extraction.
• Addresses and significantly improves issues of river network discontinuity and poor representation in morphologically complex rivers.
• Demonstrates superior performance in positional accuracy and network continuity over existing conventional methods and products.
• Offers a robust solution for high-resolution river network mapping in challenging geographical settings.

Funding

Not explicitly stated in the provided text.

Citation

@article{Zang2026HighPrecision,
  author = {Zang, Yufu and Chu, Zhaocai and Cui, Zhen and Shi, Zhuokai and Jiang, Qihan and Shen, Yichen and Ding, Jue},
  title = {High-Precision River Network Mapping Using River Probability Learning and Adaptive Stream Burning},
  journal = {Remote Sensing},
  year = {2026},
  doi = {10.3390/rs18020362},
  url = {https://doi.org/10.3390/rs18020362}
}