Xu et al. (2025) Assessing the compound impacts of urbanization and climate change on flood hazards in rapid urbanized basin: A case study of Chebei River Basin, South China

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Date: 2025-11-20
• Authors: Qipeng Xu, Sijing He, Jie Jiang, Zhaoyang Zeng, Zhaoli Wang, Huanfeng Duan
• DOI: 10.1016/j.ejrh.2025.102942

Research Groups

• School of Civil Engineering and Transportation, State Key Laboratory of Subtropical Building and Urban Science, South China University of Technology, Guangzhou, China
• Pazhou Lab, Guangzhou, China
• College of Water Conservancy and Civil Engineering, South China Agricultural University, Guangzhou, China
• Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, China

Short Summary

This study quantifies the compound impacts of urbanization and climate change on flood hazards in the Chebei River Basin (1980-2050) using a coupled hydrodynamic model and interpretable AI, revealing a critical urbanization threshold of approximately 65% that fundamentally shifts the hydrological response to rainfall.

Objective

• To systematically investigate and quantitatively disentangle the non-linear, compound impacts of long-term urbanization and climate change (rainfall intensity) on urban flood hazards in the Chebei River Basin, identifying underlying driving mechanisms and potential hydrological regime shifts.

Study Configuration

• Spatial Scale: Chebei River Basin, Guangzhou City, Guangdong Province, South China. The basin spans an area of 74 km², located between 113°20′ E to 113°26′ E and 23°6′ N to 23°15′ N.
• Temporal Scale: Historical analysis from 1980 to 2025, with future projections extending to 2050 under SSP2-RCP4.5 and SSP5-RCP8.5 climate and land-use scenarios.

Methodology and Data

• Models used:
  • Coupled 1D-2D Hydrodynamic Model: TSWM (TELEMAC-SWMM one-way coupled model)
    • SWMM (Storm Water Management Model) for 1D pipe network simulation.
    • TELEMAC-2D for 2D surface hydrodynamic simulation.
  • Land Use Classification: Support Vector Machines (SVM).
  • Future Land Use Projection: FLUS (Future Land Use Simulation) model, based on CMIP6 Land-Use Harmonization (LUH2) project.
  • Climate Projections: General Circulation Models (GCMs) from CMIP6 ScenarioMIP for future rainfall characteristics.
  • Attribution Analysis: Random Forest (RF) and SHAP (SHapley Additive exPlanations).
• Data sources:
  • Digital Elevation Model (DEM): 8 m × 8 m resolution for Guangzhou urban area, 30 m × 30 m for historical land use.
  • Historical Land Use Data: Guangdong Province for 1980, 1990, 2000, 2005, 2010, and 2015.
  • Pipe Network Data: Provided by the Water Affairs Bureau of Guangzhou Municipality.
  • Satellite Map: 4.8 m × 4.8 m resolution from AMap (for 2025 land use interpretation).
  • Soil Data: Harmonized World Soil Database.
  • Rainfall Data: Guangzhou central urban data from the Greater Bay Area hydrological database (for historical calibration), and CMIP6 GCMs (for future rainfall scenarios).
  • Future Land Use Data: 2050 land-use grid data under SSP2–4.5 and SSP5–8.5 scenarios (from Chen and Wang, 2020).

Main Results

• The Chebei River Basin underwent significant urbanization, with urban areas expanding nearly threefold from 1980 to 2025, primarily converting farmland. By 2050, urban areas are projected to cover over 70% of the basin, with future expansion predominantly at the expense of forests.
• Flood inundation areas expanded in sync with urbanization phases, showing rapid growth from 1990 to 2010 (6.02–12.48 hectares per year) and projected continued growth (3.40–9.11 hectares per year) by 2050, indicating persistent flood hazard amplification even at high impervious surface ratios.
• During 1980–2025, the relative increase in flood-prone areas was most pronounced for 10-year (73.33%) and 20-year (71.18%) return period rainfall events. For 2050 under SSP scenarios, 100-year extreme events are projected to show larger relative increases (21.59% and 27.85%) compared to 20-year events (13.37% and 17.63%).
• Interpretable AI (Random Forest and SHAP) identified urban area expansion (importance: 0.378) and forest degradation (importance: 0.307) as the two most dominant land-use drivers exacerbating flood hazard, followed by rainfall intensity (importance: 0.161).
• A critical non-linear urbanization threshold of approximately 65% was uncovered. Below this threshold, the flood-aggravating effect of urbanization is more pronounced under moderately intense rainfall. Once exceeding 65%, the system shifts to a runoff-dominated regime where extreme rainfall strongly interacts with urban expansion, significantly amplifying flood hazard.

Contributions

• Developed an innovative analytical framework by integrating a 1D-2D coupled hydrodynamic model (TSWM) with an interpretable AI framework (Random Forest and SHAP analysis) for long-term, multi-scenario flood hazard assessment.
• Quantitatively disentangled and attributed the independent and combined effects of urbanization and climate change on flood hazards, moving beyond phenomenological descriptions to mechanistic insights.
• Discovered and quantitatively defined a critical non-linear urbanization threshold (approximately 65%) that marks a fundamental regime shift in the watershed's hydrological response, transitioning from a "buffer-dominated" to a "runoff-dominated" system.
• Provided data-driven, "tipping-point"-aware scientific guidance for differentiated urban planning and flood mitigation strategies tailored to a city's specific urbanization stage.

Funding

• National Natural Science Foundation of China (52379010, 52209019)
• Natural Science Foundation of Guangdong Province (2023B1515020087, 2022A1515240071, 2022A1515010019)

Citation

@article{Xu2025Assessing,
  author = {Xu, Qipeng and He, Sijing and Jiang, Jie and Zeng, Zhaoyang and Wang, Zhaoli and Duan, Huanfeng},
  title = {Assessing the compound impacts of urbanization and climate change on flood hazards in rapid urbanized basin: A case study of Chebei River Basin, South China},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.102942},
  url = {https://doi.org/10.1016/j.ejrh.2025.102942}
}