A worldwide event-based debris flow barrier dam dataset from 1800 to 2023

Haiguang Cheng, Kaiheng Hu, Shuang Liu, Xiaopeng Zhang, Hao Li, Qiyuan Zhang, Ning Lan, Manish Raj Gouli, Pu Li, Anna Yang, Peng Zhao, Junyu Liu, Li WeiEarth 2025 · DOI

Abstract

Debris flows frequently block rivers to form barrier dams, triggering upstream flooding and catastrophic outburst floods. Despite their danger, debris flow barrier dams (DFBDs) are poorly understood compared to other landslide dams (LDs). To address this, researchers reviewed over 2,500 sources to compile a global dataset of 555 historical DFBDs from 1800 to 2023. The dataset includes 38 attributes covering dam geometry, failure modes, and climate data. Analysis reveals that DFBDs are uniquely unstable, often failing via overtopping within a year of formation. This work provides a critical foundation for improving disaster prediction and understanding the evolution of river-blocking events.

Introduction: Distinguishing Debris Flows from Landslides

In geomorphology, the distinction between a standard landslide and a debris flow is critical. Debris flows are rapid, saturated slurries of rock and mud that behave as "two-phase" flows. Consequently, the dams they form have higher water content and gentler slopes than those created by dry rockfalls or earth slides.

Real-world example of a debris flow event in the Bailong River, showing the formation of a submerged dam and subsequent mitigation efforts.

Historically, DFBDs have been lumped into general landslide dam inventories, obscuring their unique life cycles. This lack of specificity makes it difficult for engineers to predict dam stability or the scale of potential outburst floods. This new dataset aims to fill that gap by focusing exclusively on the characteristics and behaviors of debris flow-induced blockages.

Methodology: Building a High-Quality Inventory

The authors employed a rigorous multi-step process to ensure data accuracy:

  1. Extensive Review: Screening 2,519 sources, including peer-reviewed literature and government reports in multiple languages.
  2. Validation: Every event was cross-referenced using Google Earth to verify coordinates and formation dates.
  3. Conflict Resolution: A reliability hierarchy prioritized high-impact peer-reviewed journals and official records when data points (like dam height) conflicted.
  4. Model Evaluation: Six existing stability models and 21 peak-discharge models were tested against the new data to identify the most accurate predictive tools.

Key Insights: The "China Hotspot" and Rapid Failure

The dataset highlights several critical trends in DFBD formation and evolution:

  • Geographic Concentration: China accounts for 376 of the 555 recorded dams, driven by active tectonics, monsoon rains, and steep topography.
  • High Instability: DFBDs are notoriously short-lived; 73% fail within a single year, and 33% fail within just 24 hours.
  • Recurrence: Approximately 15% of river channels in the dataset experienced recurrent debris flows, accounting for 35% of all recorded dams.
  • Failure Mode: 98% of failures are caused by overtopping, where water flows over the dam and rapidly erodes the loose, saturated material.

Results: Blockage Modes and Model Performance

The study categorized blockages into three modes: complete blockage (69.8%), partial blockage (28.1%), and submerged dams (2.2%). Submerged dams are particularly difficult to detect and are often underreported in historical records.

Schematic representation of a complete river blockage by a debris flow barrier dam (DFBD).

Regarding stability assessment, the study found that the $L_s$ (AHV) model is the most reliable, achieving 100% conservative accuracy. Conversely, most empirical models used for general landslide dams consistently overestimate the "peak discharge" for DFBDs, suggesting that specialized models are necessary to avoid over-engineering emergency responses.

Conclusion and Future Work

While this dataset is the most comprehensive to date, historical gaps remain for events prior to 1900. Future research will focus on integrating advanced remote sensing to map dam geometries more accurately and conducting field investigations to identify remote or "hidden" DFBDs. This dataset is now publicly accessible via Zenodo to support global disaster mitigation efforts.

Reference: Cheng, H., et al. (2025). A worldwide event-based debris flow barrier dam dataset from 1800 to 2023. Earth System Science Data, 17, 1573–1593. https://doi.org/10.5194/essd-17-1573-2025

Summary updated

Abstract

Debris flows, as a special kind of landslide, often block rivers to form barrier dams and trigger a series of disasters such as upstream aggradation and outburst floods. The understanding of debris flow barrier dams (DFBDs) is poor, mostly due to existing studies focusing on individual events and a lack of summarization of multiple DFBD events. The existing global or regional datasets of landslide barrier dams (LDs) contain only a few cases of DFBDs and ignore the differences between DFBDs and other landslide barrier dams (LDs), such as the dams of rock slides, debris avalanches, or earth slides. To fill this gap, we reviewed 2519 high-quality literature and media reports. Focusing on identified debris flow damming events, a rigorous data review and validation process was conducted using Google Earth. A systematic approach was employed to prioritize conflicting information from various data sources. Consequently, a global dataset was compiled, encompassing 555 historical DFBDs from 1800 to 2023. This pioneering global dataset includes 6 categories and 38 attributes, detailing DFBDs. It captures basic information (location, the date of formation, etc.), dam characteristics (height, length, volume, etc.), lake characteristics (area, capacity, length), debris flow characteristics (velocity, discharge, volume, etc.), failure characteristics (peak discharge, loss of life, etc.), and climate characteristics (precipitation and temperature). Our dataset elucidates that DFBDs exhibit key features of instability, complete blockage, and overtopping failure. The number of such dams has notably increased, especially in China. A total of 15 % of channels showed recurrent debris flows, resulting in DFBDs that make up 35 % of all DFBDs. Further analysis suggests that the Ls (AHV) model should be used for priority use, followed by the DBI model, in the stability assessment of DFBDs. Compared to other barrier dam datasets, our dataset is more targeted places greater emphasis on the review of raw data, and stresses the unification of terminology and concepts (i.e., blockage modes and stability), ensuring the consistency and accuracy of the data. The dataset and results in this work may help to deepen the understanding of DFBD formation, distribution, and evolution. The DFBD dataset can be accessed through this link: https://doi.org/10.5281/zenodo.14766647 (Cheng et al., 2025).
BibTeX
@inproceedings{cheng2025worldwide,
  author = {Haiguang Cheng and Kaiheng Hu and Shuang Liu and Xiaopeng Zhang and Hao Li and Qiyuan Zhang and Ning Lan and Manish Raj Gouli and Pu Li and Anna Yang and Peng Zhao and Junyu Liu and Li Wei},
  title = {{A worldwide event-based debris flow barrier dam dataset from 1800 to 2023}},
  booktitle = {Earth},
  year = {2025},
  doi = {10.5194/essd-17-1573-2025},
}
Figure