Zhao et al. (2025) A systematic review of methods for identifying drought-flood abrupt alternation: advances and future directions

Identification

• Journal: Frontiers in Environmental Science
• Year: 2025
• Date: 2025-11-11
• Authors: Lin Zhao, Kadaruddin Aiyub, Frankie Marcus Ata, Lam Kuok Choy, Mengzhu Sun
• DOI: 10.3389/fenvs.2025.1590613

Research Groups

• Geography Program, Centre for Research in Development, Social and Environment, Faculty of Social Sciences and Humanities, Universiti Kebangsaan Malaysia, Bangi, Selangor, Malaysia
• Faculty of Geography, Yunnan Normal University, Kunming, Yunnan, China
• The Engineering Research Center of GIS Technology in Western China of Ministry of Education of China, Yunnan Normal University, Kunming, Yunnan, China

Short Summary

This systematic review synthesizes 55 publications on Drought-Flood Abrupt Alternation (DFAA) event identification, proposing a unified definition framework, evaluating methodological advances, and outlining future directions to improve DFAA identification under climate change. It finds that while traditional indices are common, advanced methods incorporating transition characteristics are emerging, and future research needs multi-source data integration and dynamic time windows for better accuracy and policy alignment.

Objective

• To systematically review methods for identifying Drought-Flood Abrupt Alternation (DFAA) events, focusing on emerging concepts, characterization variables, and methodological advancements under climate change.
• To propose a unified definition framework for DFAA events.
• To synthesize identification and characterization methods for DFAA events.
• To evaluate recent methodological advances and outline future directions for improving DFAA identification.

Study Configuration

• Spatial Scale: Global, regional (e.g., South Asia, Central Asia, Asia), national (e.g., China, United States of America, Canada, South Korea), and basin-specific.
• Temporal Scale: Literature published between 1 January 2000 and 31 December 2024. The reviewed studies utilize monthly-scale, daily-scale, and sub-daily (future direction) data/indices, with DFAA events characterized by transitions from days to weeks.

Methodology and Data

• Models used: Hydrological models (e.g., VIC, SWAT), Machine Learning (ML) and Deep Learning (DL) models (e.g., Long Short-Term Memory, Convolutional Neural Network, Generative Adversarial Network), Climate Models (e.g., CMIP5, CMIP6, GCMs), Bluecat and e-Bluecat frameworks.
• Data sources: SCOPUS and Web of Science (for the review itself), ground-based station observations, reanalysis datasets (e.g., ERA5, MERRA, NCEP), remote sensing data (e.g., TRMM precipitation, MODIS land surface temperature, SMAP soil moisture, Sentinel-2, Landsat), climate model outputs.

Main Results

• A unified definition framework for DFAA events is proposed, based on transition direction, transition time (days to weeks), alternation point, and geographical consistency.
• Research on DFAA events has significantly increased, particularly between 2019 and 2024, with a predominance of studies in China (58%).
• 58% of reviewed studies use traditional drought-flood indices (e.g., SPEI, SPI, SWAP, SAPEI), while 42% propose DFAA-specific indices (e.g., LDFAI, SDFAI, DWAAI, MSDFAI, MSDFI, SMCI).
• Traditional methods often treat droughts and floods as separate events, overlooking key DFAA characteristics like alternation points, transition time, and transition speed.
• Advanced methods (e.g., MSDFAI, MSDFI, SMCI) incorporate features like transition speed and multiple hydroclimatic variables, but challenges remain in accurately capturing abrupt transitions and ensuring broad applicability.
• Characterization variables like frequency (100%), intensity (69%), and duration (35%) are commonly used, but alternation point (4%), transition time (24%), transition speed (9%), severity (5%), and spatial extent (7%) are underutilized despite their importance for understanding abruptness and impacts.
• Precipitation is the dominant climatic driver of DFAA events, but temperature and potential evapotranspiration play critical roles in arid/semi-arid regions. The influence of climatic drivers is highly region-dependent.
• Future directions include integrating multi-source datasets (ground, reanalysis, remote sensing), applying dynamic time windows with ML/DL models, and using advanced bias correction/uncertainty quantification frameworks (e.g., Bluecat) to improve DFAA identification and risk assessment under climate change.

Contributions

• Provides a systematic and unbiased analysis of 55 publications on DFAA event identification methods.
• Proposes a unified definition framework for DFAA events, emphasizing abruptness and regional consistency through key terms like transition direction, alternation point, transition time, and geographical consistency.
• Comprehensively reviews and categorizes existing DFAA identification and characterization methods, highlighting their strengths and limitations.
• Evaluates state-of-the-art DFAA identification indices (MSDFAI, MSDFI, SMCI) using a SWOT analysis.
• Synthesizes the influence of major climatic drivers on DFAA events, revealing region-dependent patterns through meta-analysis.
• Outlines concrete future directions for improving DFAA identification, including multi-source data integration, dynamic time windows, and advanced modeling techniques, to enhance early warning and risk management under climate change.

Funding

• The author(s) declare that no financial support was received for the research and/or publication of this article.

Citation

@article{Zhao2025systematic,
  author = {Zhao, Lin and Aiyub, Kadaruddin and Ata, Frankie Marcus and Choy, Lam Kuok and Sun, Mengzhu},
  title = {A systematic review of methods for identifying drought-flood abrupt alternation: advances and future directions},
  journal = {Frontiers in Environmental Science},
  year = {2025},
  doi = {10.3389/fenvs.2025.1590613},
  url = {https://doi.org/10.3389/fenvs.2025.1590613}
}