Çelik (2026) Shifting aridity patterns in Türkiye: a comparative assessment of De Martonne, UNEP, Erinç and Budyko Dryness indexes

Identification

• Journal: Theoretical and Applied Climatology
• Year: 2026
• Date: 2026-02-27
• Authors: Mehmet Ali Çelik
• DOI: 10.1007/s00704-026-06118-w

Research Groups

• Geography Department, Faculty of Arts and Sciences, Iğdır University, Iğdır, Türkiye
• Department of Geography, Nakhchivan State University, Nakhchivan, Azerbaijan

Short Summary

This study comparatively assessed shifting aridity patterns in Türkiye using four dryness indexes and high-resolution climate data for two climate normal periods. It found a consistent expansion of semi-arid areas, primarily driven by increasing potential evapotranspiration rather than precipitation variability, with aridity changes concentrated in semi-arid transition zones.

Objective

• To produce and compare aridity maps for Türkiye for two climate normal periods (1958–1990 and 1991–2020) using the De Martonne, UNEP, Erinç, and Budyko Dryness indexes with high-resolution climate data, identifying spatial patterns and changes in aridity to evaluate recent hydroclimatic impacts of climate change.

Study Configuration

• Spatial Scale: Türkiye, with approximately 4 kilometers (km) spatial resolution (pixel-based/grid-cell based analysis). Focus on 30 agricultural basins.
• Temporal Scale: Two consecutive 30-year climate normal periods: 1958–1990 (reference) and 1991–2020 (current).

Methodology and Data

• Models used:
  • De Martonne Aridity Index (DAI = P / (T + 10))
  • UNEP Aridity Index (UNEPAI = P / PET)
  • Erinç Aridity Index (EAI = P / T)
  • Budyko Dryness Index (DIB = PET / P) (Where P = total annual precipitation in millimeters (mm), T = average annual air temperature in degrees Celsius (°C), PET = annual potential evapotranspiration in millimeters (mm)).
• Data sources:
  • Primary: TerraClimate dataset (global, monthly climate and water balance data, ~4 km spatial resolution, 1958 onwards), combining CRU Ts4.0, WorldClim, and PRISM with climate model outputs.
  • Validation: General Directorate of Meteorology of Türkiye (MGM) station data, ERA5-Land reanalysis dataset (~9 km resolution).
  • Platform: Google Earth Engine (GEE) for pixel-based spatialization and analysis.

Main Results

• TerraClimate-based De Martonne and Erinç indices showed strong spatial agreement with MGM station data (Pearson correlation coefficient, r = 0.82 and r = 0.84, respectively).
• TerraClimate-based Budyko and UNEP indices showed good spatial validation with ERA5-Land (r = 0.78 and r = 0.76, respectively).
• Inter-index correlation for 1958–1990 showed high correlation between De Martonne, UNEP, and Erinç (r = 0.79–0.85), while Budyko showed medium-low correlation (r = 0.39–0.49).
• For 1991–2020, the correlation between De Martonne and UNEP remained high (r = 0.83), but Erinç's correlation with other indexes significantly weakened (e.g., r = 0.29 with De Martonne, r = 0.06 with Budyko), indicating its inability to fully capture recent PET dynamics.
• Delta analyses (1991–2020 vs. 1958–1990) revealed a consistent expansion of semi-arid areas across all indices, ranging from +0.36% (Erinç) to +0.87% (De Martonne).
• Arid areas showed limited changes, mostly slight decreases or minor increases (e.g., Budyko: -0.06%, De Martonne: -0.34%, Erinç: +0.04%, UNEP: -0.03%).
• Water–energy balance-based indices (UNEP, Budyko) captured the recent drying signal more strongly and consistently, indicating increasing potential evapotranspiration (PET) as a dominant driver of drying processes over precipitation variability.
• Aridity increase is concentrated in semi-arid transition zones and critical agricultural basins (e.g., Central Anatolia, Konya Closed Basin, Middle Kızılırmak, Great Ağrı basins).
• The Mann–Whitney U test showed no statistically significant difference (p > 0.05) between the two periods for any index, suggesting that observed changes, while spatially discernible, remain largely within the limits of natural climate variability at the national or basin scale.

Contributions

• First comprehensive comparative assessment of four widely used aridity indexes (De Martonne, UNEP, Erinç, Budyko) across Türkiye using a consistent high-resolution dataset and time periods.
• Reduced uncertainties in aridity assessments by employing a multi-index approach, revealing distinct hydroclimatic dimensions of aridity dynamics.
• Highlighted the increasing dominance of atmospheric water demand (PET) over precipitation variability as a driver of aridification in Türkiye, particularly through the water-energy balance perspective of the Budyko index.
• Provided high-spatial detail aridity maps and delta analysis for Türkiye's agricultural basins, offering a robust scientific foundation for targeted water resource management and climate adaptation strategies.
• Demonstrated that aridity changes are concentrated in semi-arid transition zones and key agricultural production regions, emphasizing their vulnerability to climate change.

Funding

The author received no financial support from public, commercial, or not-for-profit funding agencies to conduct this research.

Citation

@article{Çelik2026Shifting,
  author = {Çelik, Mehmet Ali},
  title = {Shifting aridity patterns in Türkiye: a comparative assessment of De Martonne, UNEP, Erinç and Budyko Dryness indexes},
  journal = {Theoretical and Applied Climatology},
  year = {2026},
  doi = {10.1007/s00704-026-06118-w},
  url = {https://doi.org/10.1007/s00704-026-06118-w}
}