Zhang et al. (2025) Influence of drought identification methods on analyzing and assessing responses of water quality to droughts

Identification

• Journal: Ecological Indicators
• Year: 2025
• Date: 2025-11-22
• Authors: Weijie Zhang, Jiefeng Wu, Huaxia Yao, Jie Wang
• DOI: 10.1016/j.ecolind.2025.114458

Research Groups

• School of Hydrology and Water Resources, Nanjing University of Information Science and Technology, Nanjing, China
• College of Civil Engineering, Hefei University of Technology, Hefei, China
• Inland Waters Unit, Environmental Monitoring and Reporting Branch, Ontario Ministry of Environment, Conservation and Parks, Dorset, Ontario, Canada
• The National Key Laboratory of Water Disaster Prevention, Nanjing Hydraulic Research Institute, Nanjing, China

Short Summary

This study compares fixed (FDT) and variable (VDT) drought identification methods to assess their influence on water quality responses in the Harp Lake catchment, Ontario. Findings reveal that the choice of drought identification method significantly impacts the assessment of water quality parameters (dissolved organic carbon, total nitrogen, total phosphorus) and their dynamics during drought events.

Objective

• To compare the differences between Fixed Drought Threshold (FDT) and Variable Drought Threshold (VDT) methods in hydrological drought (HD) identification and evaluate their impacts on assessing water quality parameters (dissolved organic carbon, total phosphorus, and total nitrogen) in terms of both concentration and flux.
• To analyze the sensitivity differences of threshold setting to optimize drought monitoring methods.
• To develop a coupled analysis framework for water quantity and quality assessment, clarifying potential effects of threshold selection in water quality assessment.

Study Configuration

• Spatial Scale: Harp Lake catchment, south-central Ontario, Canada (542 hectares), a snowmelt-dominated system. Four sub-catchments were studied: HP3 (26 hectares), HP3a (20 hectares), HP4 (119 hectares), and HP6 (10 hectares).
• Temporal Scale: Long-term (1978–2018), covering 41 years of monthly streamflow and water quality data.

Methodology and Data

• Models used:
  • Fixed Drought Threshold (FDT): Hydrological drought identified when streamflow falls below the 20th percentile of the multi-year flow series.
  • Variable Drought Threshold (VDT): Hydrological drought identified when monthly streamflow falls below the 20th percentile of historical flow for that specific month.
  • Linear regression models were used to quantify relationships between water quality (concentration and flux) and flow, as well as between cumulative concentration and drought characteristics (duration and severity).
• Data sources:
  • Long-term, high-frequency, and uninterrupted hydrological and water quality monitoring data from the Harp Lake catchment.
  • Data collected by the Ontario Ministry of Environment, Conservation and Parks.
  • Parameters included monthly streamflow, dissolved organic carbon (DOC) concentration (milligrams per liter), total nitrogen (TN) concentration (micrograms per liter), and total phosphorus (TP) concentration (micrograms per liter).

Main Results

• FDT identified fewer (e.g., 44 vs 53 for HP3), but longer (average duration 25.35% longer) and more severe (average severity 79.32% larger) drought events, predominantly concentrated in July through September.
• VDT detected more frequent, shorter droughts, evenly distributed throughout the year, capturing short-term wet-dry fluctuations.
• Drought events identified by FDT exhibited higher concentrations of DOC, TP, and TN than those by VDT (e.g., DOC concentration 23.11% higher for FDT).
• Conversely, fluxes of DOC, TP, and TN were significantly lower during FDT-identified droughts compared to non-drought periods and generally lower than VDT-identified droughts (e.g., DOC flux 80.15% lower for FDT in HP3).
• Both methods showed a significant positive correlation between drought duration and cumulative nutrient concentrations, with FDT exhibiting faster response rates (steeper slopes).
• A significant positive correlation was also found between drought severity and cumulative nutrient concentrations, but VDT generally showed higher response rates (steeper slopes) than FDT.
• Water quality concentrations generally decreased slightly with increasing flow during drought, but this correlation was not significant. Fluxes, however, significantly increased with flow for both methods (p < 0.01), with FDT showing a steeper rate of change.

Contributions

• This study systematically demonstrates that the choice of hydrological drought identification method (FDT vs. VDT) is a decisive factor influencing the assessment of water quality responses, addressing a previously overlooked aspect and a "hidden uncertainty" in the literature.
• It provides a mechanistic explanation for the "concentration-flux paradox" during droughts, showing its intrinsic link to how drought events are defined.
• The research offers a comparative analytical framework to quantify the influence of different drought identification methods on water quality assessment, providing an improved basis for optimizing drought monitoring.
• It highlights the implications of methodological divergence for regional water management policies, adaptation strategies, and efforts towards Sustainable Development Goal 6 (Clean Water and Sanitation).

Funding

• National Key Research and Development Program of China (Grant No. 2024YFC3211302)
• National Natural Science Foundation of China (Grant No. 52379013, 52279008)

Citation

@article{Zhang2025Influence,
  author = {Zhang, Weijie and Wu, Jiefeng and Yao, Huaxia and Wang, Jie},
  title = {Influence of drought identification methods on analyzing and assessing responses of water quality to droughts},
  journal = {Ecological Indicators},
  year = {2025},
  doi = {10.1016/j.ecolind.2025.114458},
  url = {https://doi.org/10.1016/j.ecolind.2025.114458}
}