Ghezali et al. (2025) Combined CA-ANN, CMIP6 GM and SCS-CN modeling of future impacts of climate change and urbanization on potential natural groundwater recharge at city scale

Identification

• Journal: Environmental Monitoring and Assessment
• Year: 2025
• Date: 2025-10-31
• Authors: Sana Ghezali, Mohamed Amine Boukhemacha
• DOI: 10.1007/s10661-025-14743-7

Research Groups

• Ecole Nationale Polytechnique, Laboratories LMGCE/LRS-Eau, Algiers, Algeria
• Ecole Nationale Polytechnique, Laboratory LMGCE, Algiers, Algeria

Short Summary

This study modeled the long-term impacts of climate change and urbanization on potential natural groundwater recharge in Algiers (1986-2100) under SSP2-4.5 and SSP5-8.5 scenarios. Findings project a significant decline in groundwater recharge, primarily driven by climate change, with urbanization having a lesser but still notable impact.

Objective

• To investigate the long-term city-scale future impacts, both combined and isolated, of climate change and urbanization on potential natural groundwater recharge from precipitation (PNGWRP) in Algiers, Algeria, over the period 1986–2100, fully accounting for the spatial and temporal variability of urbanization.

Study Configuration

• Spatial Scale: City scale (Algiers, Algeria), covering an area of 774 square kilometers.
• Temporal Scale: Long-term projections from 1986 to 2100, with a yearly time step.

Methodology and Data

• Models used:
  • Cellular Automata-Artificial Neural Network (CA-ANN) for continuous spatiotemporal land use-land cover (LULC) mapping.
  • CMIP6 Global Models (GM), specifically the Max Planck Institute Earth System Model version 1.2—High Resolution (MPI-ESM1-2-HR), for climate change projections under SSP2-4.5 and SSP5-8.5 scenarios.
  • Soil Conservation Service-Curve Number (SCS-CN) method for hydrological modeling and PNGWRP assessment.
• Data sources:
  • Remote sensing historical LULC data: European Space Agency (ESA) Climate Change Initiative Land Cover (CCI-LC) maps (1992–2022) at 300 meter spatial resolution.
  • Digital Elevation Model (DEM) and derived terrain slope/aspect maps.
  • Observed historical daily precipitation data (1986–2016) from the Dar El-Beidha rain gauge (National Water Resources Agency).
  • Future projected daily precipitation data (2017–2100) from NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP-CMIP6) datasets at 0.25 degree × 0.25 degree spatial resolution.
  • Soil map for deriving hydrological soil groups (HSG).

Main Results

• Urbanization: Urban areas in Algiers are projected to increase from 18.2% of the total area in 1986 to 86.5% by 2100, primarily at the expense of agricultural land (declining from 58.8% to 8.85%), bare land (from 11.78% to 0.61%), and forest cover (from 8.66% to 3.19%).
• Climate Change Indicators (2017-2100 vs. 1986-2016 baseline):
  • Annual precipitation anomaly shows a decreasing trend: under SSP5-8.5, average precipitation is projected to decrease by approximately 13.8% (near future), 26.5% (mid-term), and 36% (far future) relative to the 587.7 millimeters per year baseline. SSP2-4.5 shows similar reductions.
  • Frequency and average intensity of heavy precipitation days decline progressively. For SSP2-4.5, frequency drops from a baseline of 0.72% to 0.05% (far future), and average intensity from 37.97 millimeters per day to 4.65 millimeters per day.
  • A long-term shift towards drier soil conditions is indicated by an increase in the frequency of Antecedent Moisture Condition 1 (AMC1) and a decrease in AMC3.
• Potential Natural Groundwater Recharge from Precipitation (PNGWRP):
  • A significant declining trend in PNGWRP is projected for 2017-2100 under both SSP scenarios. Future multiannual averages are reduced to 50.51 millimeters per year (49.2% of baseline) for SSP2-4.5 and 52.51 millimeters per year (51.3% of baseline) for SSP5-8.5, compared to the 102.8 millimeters per year baseline (1986-2016).
  • A strong linear correlation exists between precipitation and PNGWRP: 27.5% of precipitation contributes to potential groundwater recharge under SSP2-4.5 (R^2 = 0.8199), and 24.7% under SSP5-8.5 (R^2 = 0.7867).
  • PNGWRP is found to be more sensitive to changes in climatic conditions (precipitation patterns) than to urbanization.
  • Achieving a moderate emission pathway (SSP2-4.5) could increase potential groundwater recharge by 2.8% compared to a high emission pathway (SSP5-8.5).
  • The close alignment between historical (1936-2016) and future SSP5-8.5 projections suggests that the region's hydrological response is already following a high-emission trajectory, indicating an ongoing decline in recharge.

Contributions

• Provides a comprehensive, long-term (1986-2100) assessment of combined and isolated impacts of climate change and urbanization on potential natural groundwater recharge at a city scale.
• Introduces a novel approach that fully accounts for the continuous spatiotemporal evolution of land use-land cover (LULC) using a CA-ANN model, addressing a limitation of previous studies.
• Quantifies the relative sensitivity of groundwater recharge to climate change versus urbanization, demonstrating the dominant role of climatic factors.
• Highlights the potential benefits (2.8% increase in recharge) of adhering to a moderate emission pathway (SSP2-4.5) compared to a high emission pathway (SSP5-8.5).
• Suggests that the hydrological response in the study area is already trending towards a high-emission scenario, emphasizing the urgency of mitigation and adaptation.
• Underscores the critical need for integrated urban planning and water resource management strategies to ensure long-term groundwater sustainability in urban environments.

Funding

• Algerian Ministry of Higher Education and Scientific Research (MESRS-PRFU project WatSedMan, grant contract Nr. A17N01ES160220220001).

Citation

@article{Ghezali2025Combined,
  author = {Ghezali, Sana and Boukhemacha, Mohamed Amine},
  title = {Combined CA-ANN, CMIP6 GM and SCS-CN modeling of future impacts of climate change and urbanization on potential natural groundwater recharge at city scale},
  journal = {Environmental Monitoring and Assessment},
  year = {2025},
  doi = {10.1007/s10661-025-14743-7},
  url = {https://doi.org/10.1007/s10661-025-14743-7}
}