Luintel et al. (2026) Cross-comparison of national drought monitoring products in Central Europe using a new drought impact database

Identification

• Journal: Regional Environmental Change
• Year: 2026
• Date: 2026-02-27
• Authors: Nirajan Luintel, Emanuel Bueechi, Markéta Poděbradská, Mirek Tnka, Lukáš Dolák, Ksenija Cindrić Kalin, Lívia Labudová, Jan Řehoř, Ivan Lončar-Petrinjak, Gregor Gregorič, Katarzyna Żyłowska, Wouter Dorigo
• DOI: 10.1007/s10113-026-02536-8

Research Groups

• Department of Geodesy and Geoinformation, TU Wien, Vienna, Austria
• Global Change Research Institute, Czech Academy of Sciences, Brno, Czechia
• Institute of Agrosystems and Bioclimatology, Faculty of AgriSciences, Mendel University in Brno, Brno, Czechia
• Department of Geography, Faculty of Science, Masaryk University, Brno, Czechia
• Department of Geophysics, University of Zagreb, Zagreb, Croatia
• Slovak Hydrometeorological Institute, Bratislava, Slovakia
• Department of Climatology, Croatian Meteorological and Hydrological Service, Zagreb, Croatia
• Slovenian Environment Agency, Ljubljana, Slovenia
• Institute of Soil Science and Plant Cultivation-State Research Institute, Pulawy, Poland

Short Summary

This study evaluated six national drought monitoring products in Central Europe using a novel, high-resolution drought impact database derived from national newspaper reports (2000–2023). It found varying effectiveness in detecting impact occurrence and capturing impact severity across countries and indices, highlighting the need for multi-index approaches and improved impact data for operational drought management.

Objective

• To evaluate the effectiveness of six national drought monitoring products in Central Europe in detecting drought impact occurrence and capturing impact severity using a novel, NUTS3-level extreme event impact database.

Study Configuration

• Spatial Scale: Central Europe (Austria, Croatia, Czechia, Poland, Slovakia, Slovenia). Impact data and drought indices were analyzed at the NUTS3 administrative level, with results aggregated and presented at the NUTS2 level. Drought index data were either native raster (1 km or 0.1 degree) or point data interpolated to a 0.1-degree regular grid.
• Temporal Scale: Drought impact database and drought index analysis covered the period 2000–2023. Drought indices were standardized using a baseline period of 1981–2020. Aggregation periods for indices varied from 1 month to 24 months. Daily and monthly time series were used for analysis.

Methodology and Data

• Models used: SoilClim model (for deriving relative soil moisture, AWR, in Czechia).
• Data sources:
  • Drought Monitoring Products: Standardized Precipitation Index (SPI), Standardized Precipitation Evapotranspiration Index (SPEI), Standardized Relative Soil Moisture (AWR), Surface Water Balance (SB).
  • Input Data for Indices: Precipitation (P), Potential Evapotranspiration (PET), Reference Evapotranspiration (ET0), Temperature (T), Humidity (H), Wind speed (W), Shortwave solar radiation (R).
  • Data Sources for Inputs: Gridded in-situ observations (Austria), in-situ observations (Croatia, Slovakia, Slovenia), ERA5-Land reanalysis data (forcing SoilClim for Czechia/Central Europe).
  • Impact Data: A novel extreme event impact database (Dolák et al., in preparation) compiled from national newspaper reports (one national newspaper per country) for the period 2000–2023, with spatial detail down to the NUTS3 level.
  • Validation Metrics: Area Under Receiver-Operating Characteristic Curve (AUC) for detecting impact occurrence, and Spearman correlation coefficient (r) between drought index severity and the number of reported impacts for capturing impact severity.

Main Results

• Drought Detectability (AUC): AUC values showed a non-linear relationship with aggregation period, generally peaking at intermediate periods (3 to 6 months). Czechia, Croatia, and Slovenia exhibited consistently high AUC values (some exceeding 0.8), while Austria had lower values (mostly below 0.7). Soil moisture-based indices (AWR) generally peaked at shorter aggregation periods than precipitation-based indices (SPI, SPEI).
• Impact Severity (Spearman Correlation): Median Spearman correlation coefficients between drought severity and the number of reported impacts were generally low (0.1–0.4). Correlations tended to decrease with increasing aggregation period in Poland, Czechia, and Croatia (median values for short periods exceeding 0.45). In Slovakia, Slovenia, and Austria, correlations were mostly below 0.4 and less influenced by aggregation period. The highest correlations were found for AWR-based indices in Poland (up to 6-month aggregation) and for SPI-1 and SPEI-1 in Croatia.
• Regional Variability: The best-performing index for both detectability and severity varied regionally. AWR 0–40 cm depth indices often performed best in Czechia and Poland, while national products (e.g., SPI) were superior in Croatia.

Contributions

• This study provides the first cross-comparison and independent validation of operational national drought monitoring products in Central Europe.
• It introduces and utilizes a novel, high-resolution (NUTS3 level) drought impact database, compiled from national newspaper reports, for a comprehensive assessment.
• It offers quantitative insights into the strengths and weaknesses of various drought indices (SPI, SPEI, AWR, SB) and their optimal aggregation periods in capturing both drought impact occurrence and severity.
• The findings support national hydrometeorological services in enhancing their drought monitoring products and facilitate the development of a more effective, common drought monitoring framework for Central Europe.
• The research highlights the importance of considering multiple indices for drought monitoring and the potential benefits of incorporating sophisticated models and ancillary data.

Funding

• Interreg Central Europe
• European Union’s Clim4Cast project (grant number CE0100059)
• AdAgriF (grant number CZ.02.01.01/00/22_008/0004635)
• TU Wien Bibliothek (Open Access Funding Programme)

Citation

@article{Luintel2026Crosscomparison,
  author = {Luintel, Nirajan and Bueechi, Emanuel and Poděbradská, Markéta and Tnka, Mirek and Dolák, Lukáš and Kalin, Ksenija Cindrić and Labudová, Lívia and Řehoř, Jan and Lončar-Petrinjak, Ivan and Gregorič, Gregor and Żyłowska, Katarzyna and Dorigo, Wouter},
  title = {Cross-comparison of national drought monitoring products in Central Europe using a new drought impact database},
  journal = {Regional Environmental Change},
  year = {2026},
  doi = {10.1007/s10113-026-02536-8},
  url = {https://doi.org/10.1007/s10113-026-02536-8}
}