Buttafuocò et al. (2025) Mapping average annual precipitation accounting for location-dependent variations

Identification

• Journal: Theoretical and Applied Climatology
• Year: 2025
• Date: 2025-11-25
• Authors: Gabriele Buttafuocò, Massimo Conforti, Tommaso Caloiero
• DOI: 10.1007/s00704-025-05909-x

Research Groups

• National Research Council of Italy - Institute for Agricultural and Forest Systems in the Mediterranean (ISAFOM), Rende, CS, Italy
• National Research Council of Italy - Research Institute for Geo-hydrological Protection (IRPI), Rende, CS, Italy

Short Summary

This study demonstrates that a local geostatistics approach with varying variogram model parameters significantly improves the modeling of average annual precipitation in a mountainous Mediterranean region compared to a global variogram, leading to more accurate predictions, especially in orographically complex areas.

Objective

• To determine whether the local geostatistics approach with varying variogram model parameters improves the modelling of average annual precipitation compared to the use of a global variogram with constant parameters for the entire study area, incorporating elevation as an auxiliary variable.

Study Configuration

• Spatial Scale: Calabria region, southern Italy, covering approximately 15,080 square kilometers.
• Temporal Scale: Annual average precipitation over the period 1951–2022.

Methodology and Data

• Models used: Geostatistical methods including experimental and theoretical variogram fitting (cubic model), Gaussian anamorphosis, block kriging with external drift (KED) using elevation as a covariate, and a local variogram optimization algorithm. Software: Isatis.neo, release 2024.12.
• Data sources:
  • Precipitation data: Long-term records (1951–2022) from 84 precipitation stations in Calabria, managed by the Multi-Risk Functional Center of the Regional Agency for Environmental Protection of Calabria (Arpacal).
  • Elevation data: Digital Elevation Model (DEM) with 80 m × 80 m cell size.

Main Results

• The local geostatistics approach significantly improved precipitation predictions, as evidenced by all accuracy measures: Mean Absolute Error (MAE) of 121.27 mm (compared to 157.84 mm for global), Root Mean Squared Error of Prediction (RMSEP) of 159.40 mm (compared to 203.85 mm for global), and Mean Relative Error (MRE) of 0.12 (compared to 0.15 for global).
• The goodness-of-prediction (G) for the local model was 22.27%, substantially higher than the 0.59% for the global model, indicating greater effectiveness.
• The coefficient of determination (R²) for predicted versus measured values was 0.74 for the local approach, outperforming the global approach's R² of 0.58.
• Both approaches produced generally similar spatial distributions of average annual precipitation, reflecting the region's orography with a distinct west-east gradient.
• Significant local differences were observed, with the greatest improvements from the local approach concentrated along mountain ranges and in more orographically complex areas. While most differences were within ±25 mm, larger differences (up to 150–200 mm) occurred in small, specific regions.

Contributions

• Demonstrated the superior performance of a local geostatistics approach with varying variogram parameters for modeling average annual precipitation in complex topographical regions with sparse data.
• Provided a robust methodology for enhancing precipitation estimation accuracy, particularly relevant for water resource management and climate change adaptation strategies in climate hotspot areas like the Mediterranean.
• Emphasized the necessity of adapting geostatistical algorithms to specific data conditions and complex terrains, challenging the assumption of a single, global variogram model.

Funding

• Open access funding provided by Consiglio Nazionale Delle Ricerche (CNR) within the CRUI-CARE Agreement.
• No external funding was received for this research.

Citation

@article{Buttafuocò2025Mapping,
  author = {Buttafuocò, Gabriele and Conforti, Massimo and Caloiero, Tommaso},
  title = {Mapping average annual precipitation accounting for location-dependent variations},
  journal = {Theoretical and Applied Climatology},
  year = {2025},
  doi = {10.1007/s00704-025-05909-x},
  url = {https://doi.org/10.1007/s00704-025-05909-x}
}