Xu et al. (2026) Compound effects of dams and levees reshape Yangtze flood dynamics and reveal substantial risk misestimations from ignoring levees

Identification

• Journal: Scientific Reports
• Year: 2026
• Date: 2026-03-12
• Authors: Shuyuan Xu, Haodong Sun, Lingyun Zhang, Juan Han, Ruikai Pan, Liwen Yang, Xinghua Xiang, Li Tang
• DOI: 10.1038/s41598-026-41694-x

Research Groups

• Shanxi Institute of Energy, Jinzhong, China
• Geological Environments and Disaster Prevention and Reduction Research Center, Shanxi Institute of Energy, Jinzhong, China
• Shanxi Provincial Geological Prospecting Bureau, Taiyuan, China
• Key Laboratory of Transparent Mine Geology and Digital Twin Technology, National Mine Safety Administration, Beijing, China
• Wuhan University of Science and Technology, Wuhan, China

Short Summary

This study assesses the compound effects of dams and levees on Yangtze River flood dynamics using the CaMa-Flood model, revealing their distinct and complementary roles in regulating flow and inundation, and demonstrating that ignoring levees leads to a significant overestimation of flood risk.

Objective

• To provide the first assessment of the individual and compound contributions of dams and levees to historical flood mitigation in the Yangtze River Basin, clarifying their integrated roles and revealing flood risk misestimations from ignoring levees.

Study Configuration

• Spatial Scale: Yangtze River Basin, China (1.8 million km²), simulated at 3 arcmin (approximately 5 km) resolution.
• Temporal Scale: Daily resolution, covering the historical period from 1980 to 2019.

Methodology and Data

• Models used: CaMa-Flood (Catchment-based Macro-scale Floodplain model) with integrated dam and levee modules. Runoff inputs derived from a bias-corrected VIC (Variable Infiltration Capacity) hydrological model.
• Data sources:
  • Runoff Inputs: VIC Model (bias-corrected, global simulations) at 0.25° resolution.
  • River/Floodplain Topography: MERIT Hydro global hydrography dataset (90 m resolution).
  • Reservoir Parameters: GRanD v1.3 (Global Reservoir and Dam database), GRSAD (Global Reservoir and Dam Surface Area Database), and ReGeom (Reservoir Geometry Database) for 327 major Yangtze River Basin reservoirs.
  • Levee Parameters: High-resolution levee parameterization maps from Zhao et al. (2025), refined from 15 arcmin to 3 arcmin.
  • In situ discharge Validation: Daily discharge observations from 32 gauging stations provided by Yangtze Water Management (China Hydrology Yearbook).

Main Results

• Dams and levees exhibit distinct and complementary hydrological roles: dams primarily enhance low-flow accuracy (median logNSE 0.71), while levees rectify peak-flow fidelity (median NSE 0.62).
• Ignoring levees results in a systemic ~15% overestimation of annual maximum inundation area relative to dam-only assessments, which can exceed 30% during extreme flood events like 1998.
• The combined DAMLEVEE scenario provides the most robust and spatially consistent model performance (median KGE 0.68, NSE 0.59) by effectively removing extreme negative outliers across the basin.
• Dams alone reduce yearly maximum inundation by an average of 51%, levees by 67%, and the combined DAMLEVEE scenario by 65%.
• Upstream tributaries generally experience increases in Mean Annual Flood (MAF) by 2–36% due to reservoir-controlled peak attenuation, while several mid-basin stations show reductions in MAF up to 63%.
• Long-term mean discharge shows variations mostly within ±5%, but some lower-basin stations exhibit significant negative deviations (up to –19% to –61%), indicating reduced baseflow persistence.
• A "diminishing return" effect is observed: in many reaches, the levee-only scenario is nearly as effective as the combined framework (e.g., 67% vs 65% reduction in yearly maximum inundation), suggesting that once robust levees are in place, the incremental benefit of upstream dam storage for reducing the spatial extent of flooding becomes marginal.

Contributions

• Provides the first large-scale assessment explicitly simulating the joint influence of dams and levees on historical flood dynamics in the Yangtze River Basin.
• Quantifies the distinct and complementary roles of dams (low-flow regulation and volumetric buffering) and levees (peak-flow confinement and conveyance) in reshaping river hydrology and flood inundation.
• Reveals and quantifies a systemic misestimation of flood risk (up to 15% overestimation of annual maximum inundation area) in models that neglect levee systems.
• Establishes a critical physical baseline for designing resilient, basin-wide adaptation strategies in human-regulated river systems under changing climatic conditions.
• Identifies a non-linear synergy and "diminishing return" effect between dams and levees, where the incremental benefit of one infrastructure for flood mitigation becomes marginal if the other is already highly effective.

Funding

• Fundamental Research Program of Shanxi Province (Grant numbers [202303021211196])
• Teaching Reform and Innovation Project of Higher Education Institutions in Shanxi Province in 2024 (No. J20241550)
• Key Laboratory of Transparent Mine Geology and Digital Twin Technology, National Mine Safety Administration Open Topics (No.SYSKT-2025-6)
• Research Program of Northwest Institute of Eco-environment Resources of Chinese Academy of Sciences (Project No. HF2025-R&D-722)

Citation

@article{Xu2026Compound,
  author = {Xu, Shuyuan and Sun, Haodong and Zhang, Lingyun and Han, Juan and Pan, Ruikai and Yang, Liwen and Xiang, Xinghua and Tang, Li},
  title = {Compound effects of dams and levees reshape Yangtze flood dynamics and reveal substantial risk misestimations from ignoring levees},
  journal = {Scientific Reports},
  year = {2026},
  doi = {10.1038/s41598-026-41694-x},
  url = {https://doi.org/10.1038/s41598-026-41694-x}
}