Khuong et al. (2025) A new framework for quantifying the impacts of climate variability and human activities on streamflow variation with an application to the upper Da river basin

Identification

• Journal: Climatic Change
• Year: 2025
• Date: 2025-12-01
• Authors: Hai Van Khuong, Dung Nguyen Thanh, Bich Thi Ngoc, Minh Hoang Dang, Tien Giang Nguyen
• DOI: 10.1007/s10584-025-04072-6

Research Groups

• Center for Water Resources Software, Vietnam Academy for Water Resources
• MRC Regional Flood and Drought Management Centre
• Water Resources Institute
• Faculty of Hydrology, Meteorology and Oceanography, VNU University of Science, Vietnam National University

Short Summary

This paper introduces a new framework, combining a physics-based hydrological model and an Extended Impact Factor Formula (EIFF), to quantify the impacts of climate variability and human activities on streamflow at annual and seasonal scales in transboundary river basins. Applied to the upper Da River basin, the framework revealed that human activities have lessened annual streamflow downstream since 2009, reducing flood season flow while enhancing dry season flow, with climate variability generally dominating impacts in the initial period.

Objective

• To develop a new framework for quantifying the impacts of climate variability and anthropogenic interventions on streamflow changes at both annual and seasonal time scales in transboundary river basins.

Study Configuration

• Spatial Scale: Upper Da River basin, a transboundary river spanning approximately 980 km (540 km in Vietnam, 440 km in China), covering 50,600 km² (52.4% Vietnam, 45.5% China, 2.1% Laos). Located between 103.00° to 105.00° East Longitude and 21.00° to 25.00° North Latitude.
• Temporal Scale:
  • Observed streamflow: 1980–2021.
  • Baseline period: 1981–2008.
  • Impacted period 1: 2009–2015.
  • Impacted period 2: 2016–2020.
  • Model calibration: 1981–1995 (daily data aggregated to monthly).
  • Model validation: 1996–2008 (monthly data).
  • Analysis conducted at annual and seasonal (flood: June-October; dry: November-May) time scales.

Methodology and Data

• Models used:
  • Variable Infiltration Capacity (VIC) model (physics-based hydrological model).
  • Extended Impact Factor Formula (EIFF) (newly developed by authors, using vector algebra to separate impacts and account for model errors).
  • Statistical tests: Pettitt’s Test, Standard Normal Homogeneity Test, Buishand’s Range Test, Buishand’s Likelihood Ratio Test, Buishand’s U Test (for change point detection); Mann-Kendall and Sen’s Slope tests (for trend detection).
  • SCE-UA optimization algorithm for VIC model parameter calibration.
• Data sources:
  • Land use and land cover maps.
  • Vegetation properties (Leaf Area Index - LAI, albedo).
  • Regional meteorological forcing data (daily rainfall, wind speed, maximum and minimum temperatures).
  • Digital elevation model (ALOS World 3D, 30 m resolution).
  • Flow direction map.
  • Soil map (incorporating soil properties).
  • Observed streamflow time series at Muong Lay gauging station.
  • Data on reservoirs in the upper Da River basin (locations, total storage, operational years) from various online sources.
  • Most data obtained from open data sources.

Main Results

• The framework was successfully tested in the upper Da River basin, demonstrating high interpretability, reliability, and computational efficiency.
• Annual and flood season streamflow at Muong Lay gauging station showed an abrupt change point around 2009.
• Impacts on Annual Streamflow:
  • Period 1 (2009–2015): Climate variability contributed 60.4% and human activities 39.6% to streamflow change. Human activities (reservoir infilling, consumptive uses) significantly impacted flows, especially in 2015 (Lai Chau Reservoir infilling).
  • Period 2 (2016–2020): Climate variability contributed 75.5% and human activities 24.5%. Human impacts were generally smaller than climatic impacts, but played a role in regulating flow during drought (storage) and wet years (release).
• Impacts on Seasonal Streamflow:
  • Flood Season: Human activities consistently reduced streamflow (retaining water upstream).
    • Period 1: Climate variability 51.9%, human activities 48.1%.
    • Period 2: Climate variability 59.8%, human activities 41.2%.
  • Dry Season: Human activities generally enhanced streamflow (releasing water from reservoirs).
    • Period 1: Climate variability 71.24%, human activities 28.76%.
    • Period 2: Human activities 69.98%, climate variability 30.02%. The Lai Chau reservoir's stable operation after 2015 significantly increased dry-season streamflow.
• Overall: Human activities upstream have lessened the annual streamflow to the downstream since 2009. During flood seasons, human activities reduced streamflow, while during dry seasons, they enhanced it.
• Climate variability dominated over human activities during the first impacted period (2009–2015) for annual and flood season streamflow.
• Upstream reservoirs (total storage over 4 billion m³) played an insignificant role in mitigating extreme drought and wet year impacts, as their active capacity (2.32 billion m³) is about 8.1% of flood season volume and 32.6% of dry season volume.
• VIC model calibration (1981–1995) yielded a Pbias of -1.7% and Nash index of 0.92. Validation (1996–2008) yielded a Pbias of 5% and Nash index of 0.91, indicating reliable hydrological process representation.

Contributions

• Proposes a novel framework for quantifying climate variability and human activity impacts on streamflow, offering high interpretability, reliability, and computational efficiency.
• Introduces an Extended Impact Factor Formula (EIFF) using vector algebra, which can handle situations where climatic and human impacts have opposite influential directions, a limitation of previous methods.
• Incorporates model errors into the reconstruction of natural streamflow, allowing for bias corrections, especially for seasonal simulations.
• Addresses data scarcity in transboundary basins by leveraging open data sources and satellite data.
• Enhances reliability of period division by simultaneously employing five change-point tests, mitigating uncertainty from single-test approaches.
• Validates attribution results against empirical evidence from hydroclimatic trend tests, actual human activity data (e.g., reservoir operations), and comparisons with other studies.
• Provides a robust understanding of streamflow dynamics for effective water resource management and adaptive policy development in transboundary river basins.

Funding

• Vietnam Ministry of Science and Technology (project No. ĐTĐL.CN–06/23)

Citation

@article{Khuong2025new,
  author = {Khuong, Hai Van and Thanh, Dung Nguyen and Ngoc, Bich Thi and Dang, Minh Hoang and Nguyen, Tien Giang},
  title = {A new framework for quantifying the impacts of climate variability and human activities on streamflow variation with an application to the upper Da river basin},
  journal = {Climatic Change},
  year = {2025},
  doi = {10.1007/s10584-025-04072-6},
  url = {https://doi.org/10.1007/s10584-025-04072-6}
}