Crapart et al. (2026) Global projections of aridity index for mid and long-term future based on CMIP6 scenarios

Identification

• Journal: Hydrology and earth system sciences
• Year: 2026
• Date: 2026-01-13
• Authors: Camille Crapart, Juliette Blanchet, Arona Diedhiou
• DOI: 10.5194/hess-30-163-2026

Research Groups

Univ. Grenoble Alpes, IRD, CNRS, INRAE, Grenoble INP, IGE, 38000 Grenoble, France

Short Summary

This study evaluates and projects global aridity index (AI) and dryland distribution using a multimodel ensemble of 13 CMIP6 models, finding that most regions will experience decreasing AI values, signifying drier conditions and an expansion of drylands globally by the end of the century, with varying temporal patterns across different socio-economic pathways.

Objective

• To evaluate and project global aridity index (AI) and dryland distribution using the FAO Aridity Index based on Penman-Monteith potential evapotranspiration.
• To compute the global aridity index based on the Penman-Monteith equation for mid-term (2030–2059) and long-term (2070–2099) future periods using CMIP6 models, identifying areas prone to aridification and providing maps of aridity categories for three shared socio-economic pathways (SSP 2–4.5, SSP 3–7.0, and SSP 5–8.5).

Study Configuration

• Spatial Scale: Global, with a horizontal resolution of 100 km.
• Temporal Scale: Reference period (1970–1999), mid-term future (2030–2059), and long-term future (2070–2099).

Methodology and Data

• Models used: A multimodel ensemble of 13 CMIP6 models: CAS-ESM2-0, CESM2-WACCM, CMCC-CM2-SR5, CMCC-ESM2, CNRM-CM6-1, EC-Earth3, FGOALS-f3-L, GFDL-ESM4, INM-CM4-8, INM-CM5-0, MPI-ESM1-2, MRI-ESM2-0, NorESM2-MM.
• Data sources: WorldClim database (observations and reanalysis, 1970–1999), ERA5 monthly aggregated reanalysis (1970–1999), and CMIP6 model outputs for historical (1850–2014) and future scenarios (SSP 2–4.5, SSP 3–7.0, SSP 5–8.5 for 2015–2100).

Main Results

• The CMIP6 multimodel average for the reference period (1970–1999) generally reproduces observed aridity patterns but shows a slightly wetter world than ERA5 and WorldClim, with regional biases (e.g., North-Eastern Brazil simulated as humid instead of semi-arid).
• Most regions are projected to maintain their current climate classification but will experience decreasing Aridity Index (AI) values, indicating drier conditions.
• Under SSP 2–4.5 and SSP 5–8.5, significant drying is projected for the mid-term (2030–2059), with continued but slower changes by the century’s end (2070–2099). Regions particularly affected include North and Central America, the Mediterranean Basin, and areas adjacent to present-day deserts.
• SSP 3–7.0 shows limited drying or localized wetting in the mid-term, followed by extensive drying in the long-term, leading to higher adaptation costs due to drastic shifts in climatic conditions.
• Globally, the extent of drylands (hyperarid, arid, semi-arid, and dry subhumid areas) is projected to increase by 3% in SSP 2–4.5, 3.9% in SSP 3–7.0, and 5.1% in SSP 5–8.5 by the end of the century, compared to the 1970–1999 reference period (31.8% drylands).
• The Mediterranean basin and Central America are projected to experience the largest decreases in AI, with significant drying also in South America, Europe, and Oceania.

Contributions

• Provides the first global, mid-term and long-term future estimations of the aridity index calculated using the Penman-Monteith reference potential evapotranspiration based on CMIP6 models.
• Offers a comprehensive evaluation of the performance of a 13-model CMIP6 ensemble in simulating historical aridity index and dryland distribution against WorldClim and ERA5 datasets.
• Quantifies and maps the projected expansion of drylands and shifts in aridity categories under different Shared Socioeconomic Pathways (SSP 2–4.5, SSP 3–7.0, SSP 5–8.5), highlighting regional vulnerabilities and the distinct temporal evolution of aridity in SSP 3–7.0.

Funding

• IRD (Institut de Recherche pour le Développement; France) grant number “UMR IGE Imputation 252RA5”.
• RNER-CC (AFD-C2D) project implemented in the CNCCI (Côte d’Ivoire National Center of High Performance Computing).

Citation

@article{Crapart2026Global,
  author = {Crapart, Camille and Anquetin, Sandrine and Blanchet, Juliette and Diedhiou, Arona},
  title = {Global projections of aridity index for mid and long-term future based on CMIP6 scenarios},
  journal = {Hydrology and earth system sciences},
  year = {2026},
  doi = {10.5194/hess-30-163-2026},
  url = {https://doi.org/10.5194/hess-30-163-2026}
}