Dabar et al. (2026) Evaluation of precipitation uncertainty in CORDEX models over North-Eastern Africa and the Arabian Peninsula: contributions of domain, driving GCM, and RCM configuration

Identification

• Journal: Atmospheric Research
• Year: 2026
• Date: 2026-09-21
• Authors: Omar Assowe Dabar, Benjamin Pohl, Pierre Camberlin, Moussa Mohamed Waberi, Moussa Mahdi Ahmed, Mohamed Osman Awaleh
• DOI: 10.1016/j.atmosres.2026.109352

Research Groups

• Observatoire Régional de Recherche sur l'Environnement et le Climat (ORREC), Centre d'Etudes et de Recherche de Djibouti (CERD), Djibouti-ville, Djibouti
• Centre de Recherches de Climatologie, UMR 6282 Biogéosciences, CNRS / Université Bourgogne Europe, Dijon, France
• Institut des Sciences de la Terre (IST), Centre d'Etudes et de Recherche de Djibouti (CERD), Djibouti-ville, Djibouti

Short Summary

This study quantifies the relative contributions of domain configuration, driving Global Climate Models (GCMs), and Regional Climate Models (RCMs) to precipitation uncertainty in CORDEX simulations over North-Eastern Africa and the Arabian Peninsula. It finds that GCMs are the dominant source of precipitation variability (48.1%), followed by RCMs (26.4%) and domain contributions (25.5%), with this behavior persisting in future projections.

Objective

• To quantify the relative contributions of domain configuration, driving Global Climate Models (GCMs), and Regional Climate Models (RCMs) to precipitation uncertainty in CORDEX simulations over North-Eastern Africa and the Arabian Peninsula, for both historical (1981–2005) and future (2070–2100 under RCP8.5) periods.

Study Configuration

• Spatial Scale: North-Eastern Africa and the Arabian Peninsula, specifically within the CORDEX Africa (AFR) and Middle East and North Africa (MENA) domains.
• Temporal Scale: Historical period (1981–2005) and future projections (2070–2100) under the Representative Concentration Pathway 8.5 (RCP8.5) scenario.

Methodology and Data

• Models used: Regional Climate Models (RCMs) RegCM4 and RCA4. Five GCM–RCM combinations common to both CORDEX AFR and MENA domains were analyzed.
• Data sources: CORDEX simulations (AFR and MENA domains). Observational data implicitly used for validation of seasonal rainfall patterns.

Main Results

• Both RCMs and GCMs consistently reproduce the observed large-scale seasonal rainfall patterns, including summertime maxima over Sudan/western-central Ethiopia and bimodal rainy seasons across the Horn of Africa/Arabian Peninsula.
• RCMs generally reduce major seasonal biases present in the GCMs but show uneven improvement in representing topographically driven precipitation.
• Orographic effects strongly influence RCM biases; RCA4 exhibits a pronounced increase in overestimation with elevation, while RegCM4 shows weaker altitude dependence.
• Uncertainty analysis reveals that Global Climate Models (GCMs) dominate precipitation variability (48.1%), followed by Regional Climate Models (RCMs) (26.4%) and domain contributions (25.5%).
• Strong regional contrasts in uncertainty are observed, linked to topography, coastal influences, and lateral boundary conditions.
• Future projections (2070–2100 under RCP8.5) confirm the persistence of this behavior, with GCM forcing governing the large-scale response, while RCM and domain effects introduce secondary but spatially structured modifications consistent with the historical period.

Contributions

• Provides a quantitative assessment of the relative contributions of domain configuration, driving GCMs, and RCMs to precipitation uncertainty within the CORDEX framework for North-Eastern Africa and the Arabian Peninsula.
• Highlights the dominant role of GCMs in precipitation variability and the spatially structured, secondary modifications introduced by RCMs and domain effects, both historically and in future projections.
• Emphasizes the implications of high regional precipitation uncertainty for climate change adaptation decisions in this vulnerable region.

Funding

• The paper text does not explicitly list specific funding projects, programs, or reference codes.

Citation

@article{Dabar2026Evaluation,
  author = {Dabar, Omar Assowe and Pohl, Benjamin and Camberlin, Pierre and Waberi, Moussa Mohamed and Ahmed, Moussa Mahdi and Awaleh, Mohamed  Osman},
  title = {Evaluation of precipitation uncertainty in CORDEX models over North-Eastern Africa and the Arabian Peninsula: contributions of domain, driving GCM, and RCM configuration},
  journal = {Atmospheric Research},
  year = {2026},
  doi = {10.1016/j.atmosres.2026.109352},
  url = {https://doi.org/10.1016/j.atmosres.2026.109352}
}