Zhang et al. (2026) Attribution and Uncertainty Analysis of River Runoff Changes: A Case Study of the Upstream Yangtze River in China

Identification

• Journal: Lecture notes in civil engineering
• Year: 2026
• Date: 2026-01-01
• Authors: Yu Zhang, Xiao Zhang, Jing Zhang
• DOI: 10.1007/978-981-95-4889-7_26

Research Groups

• Nanjing Hydraulic Research Institute, Nanjing, Jiangsu, China
• Bureau of Hydrology, Changjiang Water Resources Commission, Wuhan, China

Short Summary

This study develops a new large-scale hydrological modeling method to attribute streamflow changes in the Upstream Yangtze River (UYZR) to climate change and human activities, explicitly incorporating uncertainty analysis. It reveals a significant streamflow decrease in the UYZR, primarily driven by precipitation changes, with varying impacts from other climatic factors and land use/cover change.

Objective

• To develop and apply a new attribution analysis method based on large-scale hydrological modeling to quantify the impacts of climate change and human activities on streamflow in the Upstream Yangtze River (UYZR), while explicitly considering and evaluating uncertainties in the process.

Study Configuration

• Spatial Scale: Upstream Yangtze River (UYZR) in China, covering an area of 1 million square kilometers (97.37°E - 110.18°E and 21.13°N - 34.33°N), divided into 99 sub-basins.
• Temporal Scale: Data from 1951 to 2013. The study period is divided into a baseline period (1951–1993) and an altered period (1994–2013). Model calibration was performed for 1960–1980 and validated for 1981–1993.

Methodology and Data

• Models used:
  • Hydrological model: Soil and Water Assessment Tool (SWAT), specifically ArcSWAT 2012 interface.
  • Uncertainty analysis and parameter inference: Differential Evolution Adaptive Metropolis (DREAM) algorithm for Bayesian inference and Markov Chain Monte Carlo (MCMC) simulation.
  • Sensitivity analysis: Latin-hypercube One-at-a-time (LH-OAT) approach.
  • Time series analysis: Mann-Kendall test, Spearman’s Rho test, Pettitt’s test, and Sequential clustering method.
  • Reference evapotranspiration estimation: FAO Penman-Monteith method.
• Data sources:
  • Hydrological data: Monthly streamflow series from hydrological stations (e.g., Yichang, Zhimenda, Xiaodeshi, Pingshan) provided by the Bureau of Hydrology, Changjiang Water Resources Commission.
  • Meteorological data: Daily observations of precipitation, maximum and minimum air temperatures, relative humidity, wind speed, and sunshine duration from National Meteorological Observatory (NMO) stations, collected from the National Meteorological Information Centre of China (NMIC) of the China Meteorological Administration (CMA).
  • Spatial data: Digital Elevation Model (DEM) with a 90 meter spatial resolution from the Consultative Group on International Agricultural Research (CGIAR) Consortium for Spatial Information (CGIAR-CSI).
  • Geological characteristics data: Land use/cover maps (1:100000 scale for 1990, 1995, 2000, 2005, 2010) and a soil type map (1:1000000 scale) from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (RESDC).

Main Results

• The annual streamflow in the Upstream Yangtze River (UYZR) significantly decreased by 8.9 cubic kilometers per decade (km³/10a) during the study period.
• The year 1993 was identified as an abrupt change point in the UYZR's streamflow series.
• Spatially, streamflow increased in the source area (SA) but decreased in other tributaries, with particularly significant decreases observed in the Minjiang River (MJR), Tuojiang River (TJR), and Jialing River (JLR) basins.
• Precipitation, maximum temperature, wind speed, and land use/cover change (LUCC) generally contributed to decreasing streamflow in the UYZR, while minimum temperature and relative humidity tended to increase it.
• Precipitation was identified as the most dominant driving factor influencing streamflow change across the UYZR.
• The SWAT model demonstrated good performance in simulating streamflow, with coefficients of determination (R²) ranging from 0.66 to 0.95 and Nash–Sutcliffe efficiencies (E_NS) from 0.51 to 0.94 during calibration.
• Parameter uncertainty analysis showed that, on average, 66% of observed monthly streamflow fell within the 95% simulation uncertainty ranges during calibration, indicating that parameter uncertainty alone does not fully capture all modeling uncertainties.
• Precipitation exhibited the largest absolute value of attribution uncertainty in the UYZR. Relative attribution uncertainties for precipitation and LUCC were the most evident, with distinct spatial variations across different river basins.

Contributions

• Developed a novel attribution analysis method based on large-scale hydrological modeling, integrating time series analysis, the SWAT model, and the DREAM algorithm within a comprehensive framework.
• Incorporated explicit evaluation of parameter uncertainty in hydrological modeling and its propagation into attribution analysis uncertainty using Bayesian inference.
• Introduced a multi-routes-based approach to calculate the contributions of driving factors, averaging influences across different scenarios to account for factor interactions and reduce inconsistencies.
• Enabled detailed spatial attribution analysis at the sub-basin scale using a physically-based distributed hydrological model, providing more localized and practical guidance for water resource management.
• Provided a comprehensive case study of the Upstream Yangtze River, offering enhanced understanding of the complex impacts of climate change and human activities on streamflow in a critical water resource region.

Funding

• National Key Research and Development Program of China (Grant No. 2022YFC3002705)
• National Natural Science Foundation of China (Grant No. 52209032)

Citation

@article{Zhang2026Attribution,
  author = {Zhang, Yu and Zhang, Xiao and Zhang, Jing},
  title = {Attribution and Uncertainty Analysis of River Runoff Changes: A Case Study of the Upstream Yangtze River in China},
  journal = {Lecture notes in civil engineering},
  year = {2026},
  doi = {10.1007/978-981-95-4889-7_26},
  url = {https://doi.org/10.1007/978-981-95-4889-7_26}
}