Varotsos et al. (2025) CLIMADAT-GRid: a high-resolution daily gridded precipitation and temperature dataset for Greece

Identification

• Journal: Earth system science data
• Year: 2025
• Date: 2025-09-10
• Authors: Konstantinos V. Varotsos, George Katavoutas, Gianna Kitsara, Anna Karali, Giannis Lemesios, Πλάτων Πατλάκας, Maria Hatzaki, Vassilis Tenentes, Athanasios Sarantopoulos, Basil Psiloglou, Aristeidis Koutroulis, Manolis Grillakis, Christos Giannakopoulos
• DOI: 10.5194/essd-17-4455-2025

Research Groups

• Institute for Environmental Research and Sustainable Development, National Observatory of Athens, Athens, Greece
• Department of Physics, National and Kapodistrian University of Athens, Athens, Greece
• Laboratory of Climatology and Atmospheric Environment, Section of Geography and Climatology, Department of Geology and Geoenvironment, National and Kapodistrian University of Athens, Athens, Greece
• Hellenic National Meteorological Service, Athens, Greece
• School of Chemical and Environmental Engineering, Technical University of Crete, Chania, Greece

Short Summary

This study introduces CLIMADAT-GRid, the first publicly available high-resolution (1 km × 1 km) daily gridded dataset for air temperature and precipitation across Greece for 1981–2019, demonstrating superior spatial performance and closer agreement with observational data compared to global products like CHELSA-W5E5.

Objective

• To develop and evaluate CLIMADAT-GRid, the first publicly available high-resolution (1 km × 1 km) daily gridded air temperature and precipitation dataset for Greece, covering the period 1981–2019, and benchmark its performance against existing global climate products.

Study Configuration

• Spatial Scale: Greece, 1 km × 1 km grid resolution.
• Temporal Scale: Daily, covering the period 1981–2019.

Methodology and Data

• Models used:
  • Fixed Rank Kriging (FRK) - selected for CLIMADAT-GRid construction.
  • Generalized Additive Models (GAMs)
  • k-Nearest Neighbors (KNN)
  • Support Vector Machines (SVM)
  • Weather Research and Forecasting (WRF) model (version 4.1.3) - used for blending with temperature observations.
• Data sources:
  • Daily observational data from 122 meteorological stations (temperature) and 312 stations (precipitation) across Greece (National Observatory of Athens Automatic Network (NOAAN), Hellenic National Meteorological Service (HNMS), National Observatory of Athens in Thissio, General Secretariat for Natural Environment and Water of the Ministry of Environment and Energy).
  • ERA5-Land reanalysis dataset (for gap filling and homogenization of temperature data).
  • ERA5 hourly data (for WRF initial and boundary conditions).
  • ASTER Global Digital Elevation Map (GDEM) (30 m resolution, for terrain elevation).
  • Coordination of Information on the Environment (CORINE) (250 m resolution, for land use).
  • CHELSA-W5E5 v1 (for benchmarking).

Main Results

• Fixed Rank Kriging (FRK) was identified as the most reliable interpolation method for CLIMADAT-GRid, consistently performing well across variables and timescales, and effectively capturing spatial patterns over Greece's complex terrain.
• Temperature: CLIMADAT-GRid showed comparable spatial average temperatures to CHELSA-W5E5 but consistently matched station observations better, with near-zero bias and minimal error metrics (MAE, RMSE) for daily maximum (TX), mean (TG), and minimum (TN) temperatures. CHELSA-W5E5 systematically underestimated TX (biases from -0.49 °C to -0.69 °C) and TG (biases from -0.29 °C to -0.79 °C). CLIMADAT-GRid, blended with WRF, more effectively captured orographic temperature gradients and urban heat island effects. For extreme temperature indices (SU, SU35, TR), CLIMADAT-GRid showed stronger agreement with observations, particularly for SU (TX > 25 °C), with a lower bias (1 °C) compared to CHELSA-W5E5 (-9.34 °C).
• Precipitation: CLIMADAT-GRid indicated generally higher precipitation values than CHELSA-W5E5, especially during the rainy season, and showed minimal biases against observations (annual BIAS: -1.56 %). CHELSA-W5E5 significantly underestimated precipitation (annual BIAS: -19 %, seasonal biases up to -38.27 % in JJA). Both datasets overestimated the number of wet days (RR1), with CLIMADAT-GRid showing a positive bias of approximately 49 days per year and a more pronounced orographic pattern. For heavy precipitation days (RR10, PR ≥ 10 mm), CLIMADAT-GRid performed better quantitatively with a lower bias (-3 days per year) and higher KGE (0.86) compared to CHELSA-W5E5 (-7 days per year, KGE 0.73).

Contributions

• Development of CLIMADAT-GRid, the first publicly available daily gridded air temperature and precipitation dataset for Greece at a high resolution of 1 km × 1 km, covering the period 1981–2019.
• Comprehensive evaluation of multiple interpolation techniques for gridded climate dataset construction, identifying Fixed Rank Kriging as the most robust for complex terrain.
• Integration of high-resolution atmospheric model (WRF) simulations with observational data to enhance temperature gridding, addressing limitations of sparse observational networks.
• Benchmarking against a global high-resolution product (CHELSA-W5E5), demonstrating CLIMADAT-GRid's superior spatial performance and closer agreement with observations for both mean and extreme values over Greece.

Funding

• "High resolution gridded CLIMAte change DATasets for Greece, CLIMADAT-hub" project (project ID: 15478).
• National Recovery and Resilience Plan Greece 2.0, funded by the European Union – NextGenerationEU (implementation body: HFRI).

Citation

@article{Varotsos2025CLIMADATGRid,
  author = {Varotsos, Konstantinos V. and Katavoutas, George and Kitsara, Gianna and Karali, Anna and Lemesios, Giannis and Πατλάκας, Πλάτων and Hatzaki, Maria and Tenentes, Vassilis and Sarantopoulos, Athanasios and Psiloglou, Basil and Koutroulis, Aristeidis and Grillakis, Manolis and Giannakopoulos, Christos},
  title = {CLIMADAT-GRid: a high-resolution daily gridded precipitation and temperature dataset for Greece},
  journal = {Earth system science data},
  year = {2025},
  doi = {10.5194/essd-17-4455-2025},
  url = {https://doi.org/10.5194/essd-17-4455-2025}
}