Ducco et al. (2026) The role of precipitation and irrigation on groundwater droughts in the Piemonte Plain, Italy

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-03-14
• Authors: Edoardo Ducco, Ilaria Butera, Stefania Tamea
• DOI: 10.1016/j.ejrh.2026.103315

Research Groups

Politecnico di Torino – Department of Environment, Land and Infrastructure Engineering, Turin, Italy

Short Summary

This study investigates the relationship between meteorological and groundwater droughts in the shallow aquifers of the Piemonte Plain, Italy, focusing on the impact of widespread irrigation. It finds that irrigation significantly weakens the correlation between precipitation and groundwater levels, delays groundwater response, and mitigates the propagation of meteorological drought into severe groundwater drought, particularly in rice-cultivated areas.

Objective

• To investigate the relationship between meteorological and groundwater droughts in shallow aquifers of the Piemonte Plain, Italy, considering the influence of widespread irrigation.
• To analyze groundwater-level trends and seasonal behaviors over the period 2000–2023.
• To study anomalies in precipitation and groundwater levels using the Standardized Precipitation Index (SPI) and Standardized Groundwater Index (SGI) across multiple time-windows and lags.
• To assess SPI–SGI response times during and outside the irrigation period using a newly introduced correlation-weighted lag.
• To perform a conditional frequency analysis to understand the propagation of meteorological drought into groundwater drought.

Study Configuration

• Spatial Scale: Piemonte Plain (upper Po plain), northwestern Italy, spanning 6890 square kilometers.
• Temporal Scale: Groundwater level trends and anomalies analyzed for the period 2000–2023. Precipitation data used for analysis were limited to the same period.

Methodology and Data

• Models used:
  • Mann-Kendall (MK) test with Sen’s slope estimation for trend analysis.
  • Trend-Free Pre-Whitening Mann-Kendall with Yue’s modification (TFPW-Y) for autocorrelated data.
  • Standardized Precipitation Index (SPI) using piece-wise Gamma probability distribution for precipitation anomalies.
  • Standardized Groundwater Index (SGI) using Kernel Density Estimation (KDE) for groundwater level anomalies.
  • Pearson correlation coefficient for assessing linear relationships between SPI and SGI.
  • Correlation-weighted lag (lagw) to characterize groundwater response time.
  • Conditional frequency analysis for drought propagation assessment.
• Data sources:
  • Daily precipitation data from the North Western Italy Optimal Interpolation (NWIOI) dataset (ARPA Piemonte, 2023a), with a resolution of 0.125 degrees.
  • Spatial data regarding groundwater bodies and piezometer positions from ARPA Piemonte’s “Portale delle Acque” (ARPA Piemonte, 2023b).
  • Daily groundwater level time series from 106 selected piezometers (out of 125) from ARPA Piemonte (2023b) and Regione Piemonte (2018).

Main Results

• The shallow aquifer system in the Piemonte Plain exhibits a widespread decline in groundwater levels over the 2000–2023 period, with an increased prevalence of statistically significant negative trends compared to previous studies (2000–2017).
• Seasonal analysis reveals "Irrigation Pattern" (statistically significant negative trends during April-September) predominantly in northern groundwater bodies and rice-cultivated areas, suggesting a year-over-year decline in seasonal groundwater levels potentially linked to changes in irrigation practices.
• SPI-SGI correlations are consistently higher during the non-irrigation period (October-March) compared to the irrigation period (April-September), indicating that irrigation interferes with the natural recharge from precipitation and weakens the connection between precipitation and groundwater.
• Rice-cultivated areas show the lowest SPI-SGI correlation values, where flooding irrigation significantly disrupts the precipitation-groundwater relationship.
• The newly introduced correlation-weighted lag (lagw) indicates average delays in SGI response to SPI of approximately 1.89 months for SPI1-SGI1, 1.54 months for SPI3-SGI1, and 1.34 months for SPI6-SGI1. Lagw values are generally larger during the irrigation period (e.g., 1.9 months for SPI1-SGI1 vs. 1.7 months in non-irrigation), implying a more gradual SGI reaction to precipitation anomalies during this time.
• Conditional frequency analysis demonstrates that during the irrigation period, strong negative SPI values are less frequently followed by strong negative SGI outcomes, highlighting irrigation's role in mitigating the propagation of meteorological drought into severe groundwater drought.

Contributions

• Quantifies the spatial and temporal variability of groundwater drought lag and attenuation under significant anthropogenic influence (irrigation).
• Introduces a novel correlation-weighted lag (lagw) method to more accurately characterize groundwater response time to precipitation, addressing limitations of single-lag maximum correlation approaches.
• Provides a detailed, period-specific assessment of how irrigation impacts the propagation of meteorological drought into groundwater drought, distinguishing between irrigation and non-irrigation periods.
• Emphasizes the critical need for drought monitoring and early-warning systems in irrigated areas to integrate both meteorological and groundwater indicators, considering the distinct behaviors during irrigation and non-irrigation seasons.
• Offers a transferable methodology for analyzing groundwater-climate linkages in managed hydrological systems, beyond the specific study region.

Funding

• European Union’s Horizon Europe programme under Grant Agreement No. 101112876, MountResilience.

Citation

@article{Ducco2026role,
  author = {Ducco, Edoardo and Butera, Ilaria and Tamea, Stefania},
  title = {The role of precipitation and irrigation on groundwater droughts in the Piemonte Plain, Italy},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2026.103315},
  url = {https://doi.org/10.1016/j.ejrh.2026.103315}
}