Zemzami et al. (2026) Integrating climate models to confront the illusion of certainty in water planning: evidence from Morocco

Identification

• Journal: Climatic Change
• Year: 2026
• Date: 2026-01-01
• Authors: Mahmoud Zemzami, Hajar Habib, Souad Haïda
• DOI: 10.1007/s10584-026-04108-5

Research Groups

• Higher School of Education and Training (ESEF), Department of Earth Sciences, University of Mohammed First, Oujda, Morocco
• Department of Environmental Sciences and Management, College of Science, University of Liège, Liège, Belgium
• College of Science, Department of Geology, University of Ibn Tofail, Kenitra, Morocco
• LCM2E Lab Geo-Environment and Health, Poly-Disciplinary College, University of Mohammed First, Nador, Morocco

Short Summary

This study examines the implications of hydroclimatic non-stationarity on water planning in a semi-arid Moroccan basin, revealing that traditional stationary assumptions significantly overestimate water availability and conceal critical system vulnerabilities, necessitating climate-informed adaptive strategies combining reservoirs and desalination.

Objective

• To examine the implications of hydroclimatic non-stationarity on water planning in a semi-arid Moroccan basin.
• To illustrate the challenges of water planning in Morocco under the stationary paradigm and demonstrate the advantages of incorporating General Circulation Model (GCM) projections for climate-informed, adaptive water planning.

Study Configuration

• Spatial Scale: Rhiss and Nekkor basins in northern Morocco, with a total area of approximately 1370 square kilometers. GCM spatial resolution ranges from approximately 100 to 180 kilometers.
• Temporal Scale: Observed hydroclimatic data from 1977 to 2012 (36 years). Future climate projections from 1950 to 2100, with water planning scenarios focusing on the period 2024–2060.

Methodology and Data

• Models used:
  • Climate Models: Ensemble of 10 CMIP6 GCMs (CanESM5, CNRM-CM6, FGOALS-f3-L, FIO-ESM-2–0, GISS-E2-1-G, HadGEM3, INM-CM4-8, MIROC-ES2L, MRI-ESM2-0, NorESM2-LM) under SSP4.5 and SSP8.5 scenarios.
  • Statistical Downscaling: Power Transform (PT), Linear Scaling (LS), and Quantile-Mapping (Qmap).
  • Hydrological Model: Multilayer Feedforward Neural Network (MFFNN) model, trained with the Levenberg-Marquardt algorithm.
  • Water System Model: Scenario-based river-reservoir simulation framework based on water balance calculations.
• Data sources:
  • Observed Hydroclimatic Data: Long-term monthly time series of precipitation, temperature, and streamflow observations from the Directorate of Water Resources Planning (DRPE 2024). Missing values reconstructed using machine learning techniques.
  • Reservoir System Data: Dam characteristics (e.g., crest elevation, outlet capacity, intake levels), reservoir level-surface area-storage capacity relationships, water demands, operation rules, and evaporation rates.
  • Water Demands: Official projections for municipal water demand (MWD) and irrigation demand from the National Office of Water and Electricity (ONEE) and the Ministry of Agriculture.
  • GCM Projections: Precipitation and temperature projections from the Coupled Model Intercomparison Project phase 6 (CMIP6) retrieved from the World Climate Research Programme-WCRP portal.

Main Results

• Stationary planning assumptions, based on historical data (1977–2012), systematically overestimate water availability, suggesting the Rhiss reservoir could meet municipal water demand (15 Mm³/yr) and most irrigation demand (8 Mm³/yr), with a 3.5 Mm³/yr desalination plant being sufficient.
• Under CMIP6 SSP4.5 and SSP8.5 scenarios, future inflows to the Rhiss reservoir generally decline, with significant inter-model divergence. SSP4.5 projects inflow reductions from -6.4% to -36.1%, while SSP8.5 shows a broader range from -46.1% to +26.1% relative to the baseline.
• The hydrological model accurately captured non-linear rainfall/runoff relationships (RMSE = 2.39, Nash-Sutcliffe Coefficient = 0.87, Persistence Index = 0.98).
• Under non-stationary climate futures, reservoir-based supply alone is insufficient to reliably meet future municipal and irrigation demands.
• The integration of a 7.3 Mm³/yr seawater desalination plant significantly improves municipal water supply reliability, allowing the reservoir to meet reduced MWD requirements (7.7 Mm³/yr) in most scenarios.
• However, irrigation deficits remain severe across most plausible scenarios, ranging from 49% to 86% shortage under SSP4.5 (with desalination) and 88% to complete failure under SSP8.5 (with desalination).
• The study identifies a "governance trap" where the illusion of certainty from deterministic planning leads to inadequate translation of uncertainty into action, resulting in reactive crisis management rather than anticipated planning.
• To secure municipal water supply and sustain irrigation under anticipated drier and warmer conditions, desalination capacity would need to be expanded to 16.6 Mm³/yr (for 80% irrigation demand) or 18.2 Mm³/yr (for 100% irrigation demand).

Contributions

• Critically demonstrates the limitations of traditional stationary water planning in a semi-arid, climate-vulnerable region (Morocco) by contrasting it with a robust, non-stationary, and climate-informed approach.
• Provides a comprehensive methodological framework integrating CMIP6 climate projections, statistical downscaling, a calibrated hydrological model, and a scenario-based water system simulation to assess future water security under deep uncertainty.
• Quantifies the extent to which stationary planning assumptions lead to overestimation of water availability and mask critical system vulnerabilities, particularly for irrigation.
• Evaluates the effectiveness of combining conventional (reservoir) and non-conventional (desalination) water sources under various future climate and allocation scenarios, offering insights for optimizing infrastructure design capacity.
• Highlights the institutional and governance challenges ("governance trap") that hinder the adoption of adaptive water management strategies, emphasizing the urgent need for a paradigm shift in water planning in water-stressed regions.
• Offers a transferable methodological framework and insights applicable to other water-stressed regions facing similar climatic and institutional challenges.

Funding

Not applicable.

Citation

@article{Zemzami2026Integrating,
  author = {Zemzami, Mahmoud and Habib, Hajar and Haïda, Souad},
  title = {Integrating climate models to confront the illusion of certainty in water planning: evidence from Morocco},
  journal = {Climatic Change},
  year = {2026},
  doi = {10.1007/s10584-026-04108-5},
  url = {https://doi.org/10.1007/s10584-026-04108-5}
}