Ranzi et al. (2025) A Multi‐Century Meteo‐Hydrological Analysis in the Italian Alps: Daily Streamflow (1862–2022) at Different Time Scales

Identification

• Journal: Hydrological Processes
• Year: 2025
• Date: 2025-10-29
• Authors: Roberto Ranzi, Emanuele Eccel, Paolo Colosio, Eleni Maria Michailidi
• DOI: 10.1002/hyp.70305

Research Groups

• Department of Civil, Environmental, Architectural Engineering and Mathematics, Università degli Studi di Brescia, Brescia, Italy
• Fondazione Edmund Mach, San Michele all'Adige, Italy
• Innovation and Development Office, GAIA S.P.A., Pietrasanta, Lucca, Italy

Short Summary

This paper reconstructs and analyzes a 161-year daily streamflow time series for the Adige river in the Italian Alps to identify and quantify the impact of natural and anthropic factors on hydrological changes. The study reveals a significant decline in annual streamflow and runoff coefficient, primarily driven by human activities such as water withdrawals, enhanced evapotranspiration due to temperature increase, and afforestation, rather than precipitation changes.

Objective

• To reconstruct and analyze a complete long-term daily streamflow time series of the Adige river from 1862 to 2022.
• To identify, attribute, and quantify the impact of natural climatic and anthropic factors on changes in streamflow in a mountain region with marked orographic and climatic gradients.
• To quantify observed changes in the long term at annual, monthly, and daily scales, and to assess the magnitude of their likely causes.

Study Configuration

• Spatial Scale: Adige river basin, Italian Alps, gauged at the Trento hydrometric station. The basin area is 9763 km², including 130.5 km² in Switzerland. Altitude ranges from 186 m to 3899 m above sea level. Land use is primarily woods (61%), meadows and pastures (20%), agricultural crops (4%), bare rocks (11%), and urban areas (2%).
• Temporal Scale: 161 years (January 1862 – December 2022) for daily hydrometric levels and streamflow. Meteorological data also cover multi-century periods. Analyses were conducted for the entire period, and sub-periods before (1862–1945) and after (1946–2022) major reservoir construction.

Methodology and Data

• Models used:
  • Chiu (1988) model (simplified by Moramarco et al. 2004) for mean velocity estimation.
  • Chezy stage-discharge equation for streamflow reconstruction (1862–1922).
  • Thornthwaite and Mather (1955, 1957) water balance algorithm (implemented in R package "ClimClass").
  • Theil-Sen slopes for trend estimation.
  • Mann-Kendall test for statistical significance of trends.
  • Pettitt test for change-point detection.
  • Moving Average and Running Trend Analysis (MARTA).
  • Wavelet transform with Morlet-6 basis function for spectral analysis and wavelet coherence spectra.
• Data sources:
  • Hydrometric: Manual water level readings from Trento Municipal Historical Archives (1862–1927), official daily water level data from Austrian Hydrographic Service (1893–1912), and daily water level/streamflow data from Italian Water Magistrate of Venice (1919–1974), Trento Province Hydrographic Service (1975–1988), and Office of Hydraulic Works of the Autonomous Province of Trento (1989–2022). Historical surface velocity profiles and discharge measurements (1882–1907). Streamflow data from Boara Pisani and Serravalle gauging stations for gap filling (1944–1950).
  • Meteorological: Homogenized series from the HISTALP database, original series from Meteorological Services of Trento and Bolzano provinces and Fondazione E. Mach, and daily precipitation series from Cavalese monastery records.
  • Teleconnection Indices: Atlantic Multidecadal Oscillation (AMO), North Atlantic Oscillation (NAO), and Western Mediterranean Oscillation (WeMO).
  • Ancillary Data: River cross-section geometry from historical drawings and reports, CORINE land cover classification (2018), reservoir characteristics (PAT 1978), and water balance reports (2000/60/EC Water Framework Directive).

Main Results

• Annual streamflow declined at a statistically significant rate of -1.01 mm per year, representing a -1.4% decrease per decade from 1862 to 2022.
• Annual precipitation showed a slight, statistically non-significant decrease of -0.19 mm per year.
• The runoff coefficient significantly decreased at a rate of -0.010 per decade.
• Hydrological losses (precipitation minus runoff) significantly increased at a rate of 0.86 mm per year, totaling an 86 mm increase over the last century.
• A statistically significant change point in annual runoff was detected just before the 1940s dry period, without a corresponding change in precipitation.
• Monthly runoff distribution shifted, with a statistically significant increase in winter (DJFM) and a decrease in summer (JJA) after 1945, attributed to reservoir operations and enhanced evapotranspiration.
• Wavelet spectral analysis showed increased energy at the weekly time scale after 1945, peaking in the late 1990s and then declining, reflecting hydropower regulation and subsequent environmental legislation.
• Only the Atlantic Multidecadal Oscillation (AMO) exhibited significant coherence with hydrometeorological variables at 11-30 year time scales, but this coherence vanished after the 1950s, likely due to reservoir effects.
• The 86 mm per century increase in hydrological losses is attributed to: water withdrawals (38 mm per century, mainly for irrigation), enhanced evapotranspiration due to temperature increase (30 mm per century), and afforestation (17 mm per century). The impact of increased artificial lake surface was negligible (approximately 1 mm per century).
• Daily streamflow declined across all durations, including annual maxima, with a significant decrease of -1.27 m³ s⁻¹ per year, influenced by historical floods and reservoir flood control. Low-duration flows (e.g., 355-day duration) increased after 2000 due to environmental flow regulations.

Contributions

• Reconstruction and analysis of the second longest (161-year) daily streamflow time series for an Italian river, providing a unique and previously unpublished dataset for long-term hydrological studies.
• Comprehensive quantification and attribution of multi-century streamflow changes in the Italian Alps to a combination of natural climatic factors (e.g., AMO, temperature increase) and significant anthropic influences (reservoirs, agricultural/civil/industrial water withdrawals, land-use changes like afforestation).
• Detailed investigation of changes in monthly runoff regimes and spectral properties, clearly demonstrating the impact of hydropower regulation and subsequent environmental legislation on streamflow variability.
• Robust verification of reconstructed data through comparison with nearby river basins and official records, enhancing the reliability of long-term hydrological analyses in a data-scarce region.
• Advances attribution science by disentangling the complex interplay of climate change and human activities in shaping hydrological cycles in mountain environments, providing crucial insights for water resource management under non-stationary conditions.

Funding

• Po River District Basin Authority (Grant Number 8536/2020)
• University of Brescia

Citation

@article{Ranzi2025MultiCentury,
  author = {Ranzi, Roberto and Eccel, Emanuele and Colosio, Paolo and Michailidi, Eleni Maria},
  title = {A Multi‐Century Meteo‐Hydrological Analysis in the Italian Alps: Daily Streamflow (1862–2022) at Different Time Scales},
  journal = {Hydrological Processes},
  year = {2025},
  doi = {10.1002/hyp.70305},
  url = {https://doi.org/10.1002/hyp.70305}
}