Zhang et al. (2025) Flood risk assessment in data-scarce South Sudan using a flood modeling framework

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Date: 2025-08-27
• Authors: Tianyan Zhang, Zengchuan Dong, Wenzhuo Wang, Zhiqin Qian, Ji‐Tao Zhang, Pengfei Lu, Yalei Han, Zhuozheng Li, Jianyou Ji, Yong Hou, Huacong Li
• DOI: 10.1016/j.ejrh.2025.102743

Research Groups

• College of Hydrology and Water Resources, Hohai University, Nanjing, China
• Business School of Hohai University, Nanjing, China
• College of Agricultural Science and Engineering, Nanjing, China
• Huadong Engineering Corporation Limited, Hangzhou, China
• China National Petroleum Corporation, China
• China Jiangsu International Juba Company, China

Short Summary

This study develops a national flood modeling framework for data-scarce South Sudan using integrated ground observations and satellite data with a coupled hydrological-hydrodynamic model. It reveals that the 2021–2023 flood caused unprecedented Nile River backflow into the Ghazal basin at Lake NOE, and that high-return-period floods (≥50-year) induce significant Nile discharge into the Ghazal basin via Nerboar, challenging previous hydrological assumptions.

Objective

• To develop a national flood modeling framework for data-scarce South Sudan by integrating sparse ground observations with satellite-derived datasets.
• To simulate the 2021–2023 flood event and various design flood scenarios to assess flood extent, flow dynamics, and associated risks.
• To establish a basis for national flood control planning and investigate the potential of integrating observed and satellite data for this purpose.

Study Configuration

• Spatial Scale: National scale, focusing on the wetlands within the Ghazal River and Nile River basins in South Sudan. The hydrodynamic model covers a total area of 140,339.7 km².
• Temporal Scale:
  • Historical Data: Hydrological and meteorological records prior to 1983, field data collected since 1985, and rainfall data for design scenarios spanning 1950–2022.
  • Simulation Period: The 2021–2023 flood event.
  • Model Calibration: January 2021 to June 2022.
  • Model Validation: July 2022 to April 2023.

Methodology and Data

• Models used:
  • Coupled hydrological-hydrodynamic model (general framework)
  • Mike 21 (two-dimensional hydrodynamic simulation software)
  • Soil and Water Assessment Tool (SWAT) model (for runoff simulation in subbasins with observed data)
  • Long Short-Term Memory (LSTM) model (for runoff estimation in data-scarce regions)
  • Generalized extreme value distributions (for design rainfall values)
• Data sources:
  • Observed Data (Fieldwork & Local Sources):
    • Streamflow and water level measurements from key stations (Nimule, Juba, Mangala, Malakal, Nyamlell, Wau, Tonj, northern Bentiu) provided by the Ministry of Water Resources and Irrigation of South Sudan.
    • Monthly precipitation records for Wau and Malakal from the South Sudan Civil Aviation Authority.
    • Elevation data collected in 2023 via Unmanned Aerial Vehicle (UAV) surveys near Lake Noe and field surveys of lakebed elevations.
    • Insights from local communities, South Sudanese government officials, and personnel from local oil companies.
  • Open-Source Data (Satellite, Reanalysis & Global Databases):
    • Meteorological: Climatic Research Unit gridded Time Series (CRU TS) dataset, Global Land Evaporation Amsterdam Model (GLEAM) dataset, Fifth generation of ECMWF atmospheric reanalysis for climate (ERA5) reanalysis dataset, Integrated Multi-satellite Retrievals for GPM (IMERG) product.
    • Elevation: NASA’s Shuttle Radar Topography Mission (SRTM) dataset.
    • Soil: Harmonized World Soil Database (HWSD) database v1.2.
    • Land Use: MODIS Terra/Aqua Surface Reflectance 8-Day L3 Global 500 m (MOD09A1) product.
    • Water Level: Global Reservoirs and Lakes Monitor (G-REALM) dataset.
    • Imagery: Google Earth imagery (2021–2023) for delineating river boundaries.

Main Results

• The 2021–2023 flood event was the largest since 1903, inundating 30,000 km², damaging over 100,000 homes, and severely impacting 1.24 million individuals, with economic losses estimated at a minimum of 671 million USD.
• During the 2021–2023 flood (February 2021 to August 2022), an unprecedented prolonged reversal of Nile River flow towards the Ghazal River at Lake NOE was observed, challenging previous assumptions of negligible water exchange.
• Model simulations for design flood scenarios (10, 20, 50, 100 years) indicate that Nile River backflow into the Ghazal basin near Lake Noe is unlikely under natural conditions.
• In flood scenarios exceeding a 50-year return period (100-year and 50-year), the Nile River discharges significant floodwater from Nerboar into the Ghazal basin, a process previously overlooked by Sudd wetland models.
• The 2020–2023 flood event was primarily driven by exceptionally high and sustained discharge from the Nile River, linked to human intervention in Lake Victoria's water level management, rather than local precipitation in South Sudan.
• Maximum inundation depths during the 2021–2023 flood exceeded 0.6 m across most affected regions, with certain areas in the northern Sudd Wetland experiencing depths greater than 1.6 m, and near Lake Noe, depths ranged from 2.2 m to 3.8 m.
• Flow velocities reached a maximum of 2 m/s (average 0.74 m/s) in river channels, significantly decreasing to 0.007–0.015 m/s in wetland and floodplain areas dominated by tall sedge vegetation.
• In the 100-year flood scenario, the White Nile breaches its eastern riverbank, flooding the cities of Leer, Nyala, and Ganyiel, and inundates Bentiu via Lake Noe.
• The calibrated MIKE model achieved Nash-Sutcliffe Efficiency (NSE) coefficients of 0.86 during calibration and 0.82 during validation, effectively capturing flood evolution.

Contributions

• Developed the first national flood modeling framework for data-scarce South Sudan, integrating sparse ground observations with diverse satellite and reanalysis datasets.
• Acquired and compiled critical long-term hydrological data (post-1984) through extensive fieldwork, addressing a significant data deficiency in the region.
• Challenged long-standing hydrological assumptions by identifying unprecedented Nile River backflow into the Ghazal basin at Lake NOE during the 2021–2023 flood and significant discharge from Nerboar in high-return-period floods.
• Provided methodologies for quantifying water volume in Ghazal River inflow zones and determining the maximum computational cell area for hydrodynamic models.
• Delivered the developed model files and associated calculations to the government of South Sudan, establishing a foundational tool for national flood management and production restoration initiatives.
• Elucidated complex wetland water exchange processes, offering valuable insights for Nile basin water allocation strategies.
• Proposed a scalable, open-source framework adaptable for flood risk assessment and management in other data-scarce regions globally.

Funding

• Financial support was provided by Hohai University.

Citation

@article{Zhang2025Flood,
  author = {Zhang, Tianyan and Dong, Zengchuan and Wang, Wenzhuo and Qian, Zhiqin and Zhang, Ji‐Tao and Lu, Pengfei and Han, Yalei and Li, Zhuozheng and Ji, Jianyou and Hou, Yong and Li, Huacong},
  title = {Flood risk assessment in data-scarce South Sudan using a flood modeling framework},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.102743},
  url = {https://doi.org/10.1016/j.ejrh.2025.102743}
}