Klaus et al. (2025) Brief communication: How extreme was the thunderstorm rain in Vienna on 17 August 2024? A temporal and spatial analysis

Identification

• Journal: Natural hazards and earth system sciences
• Year: 2025
• Date: 2025-12-04
• Authors: Vinzent Klaus, Johannes Laimighofer, Fabian Lehner
• DOI: 10.5194/nhess-25-4807-2025

Research Groups

• GeoSphere Austria, Vienna, Austria
• Institute of Statistics, BOKU University, Vienna, Austria
• Institute of Meteorology and Climatology, BOKU University, Vienna, Austria

Short Summary

This study quantifies the exceptional nature of the 17 August 2024 thunderstorm in Vienna, which delivered 107 mm of rain in two hours with an estimated 700-year return period, demonstrating that such extreme events are often missed by gauge networks and their frequency is likely increased by climate change.

Objective

• To quantify the exceptional nature of the 2-hour precipitation event in Vienna on 17 August 2024.
• To analyze the long-term sub-daily precipitation record at Hohe Warte to determine the event's extremity.
• To compare the event with observations from other stations in the Vienna basin and identify comparable events.
• To examine the spatial characteristics of the event using the radar-based INCA data set.
• To estimate 100-year return levels of 2-hourly precipitation for the INCA data set and assess their alignment with station data.

Study Configuration

• Spatial Scale: Vienna and its surrounding stations, covering approximately 1700 square kilometers.
• Temporal Scale:
  • Hohe Warte weather station: Hourly precipitation records from 1941 to 2024 (84 years).
  • Other rain gauges: Complete available hourly records from their onset through the end of 2023.
  • INCA data set: 15-minute data aggregated to hourly totals from 2004 to 2023 (20 years).
  • ERA5 data: Total precipitable water (PWAT) from 1941 to 2024; average mid-tropospheric temperature (500-700 hPa) from 1980s onwards.
  • Event duration analyzed: 2-hour precipitation sums.

Methodology and Data

• Models used:
  • Extreme Value Analysis: Peaks Over Threshold (POT) approach.
  • Return Period Estimation: Regional Frequency Analysis (RFA) using L-moments and Generalized Pareto Distribution (GPD).
  • Trend Analysis: Block bootstrapped Mann–Kendall trend test.
  • Climate Change Impact Assessment: Distributional Bayesian regression to account for precipitation extremes conditional on cloud temperature.
  • Statistical software: R (packages lmomco, lmomRFA, bamlss).
• Data sources:
  • Rain gauge data: Hourly precipitation from GeoSphere Austria's semi-automatic weather stations (TAWES), including the Hohe Warte station.
  • Radar-based gridded dataset: Integrated Nowcasting through Comprehensive Analysis (INCA) (1 km x 1 km spatial resolution, 15-minute data).
  • Reanalysis data: ERA5 (for total precipitable water and average temperature between 500 and 700 hPa).
  • Satellite images: High-Resolution Visible (HRV) from MSG satellites.
  • Weather radar data: Reflectivity and Doppler radial velocity from the Austrian radar network operated by Austro Control.

Main Results

• The 17 August 2024 thunderstorm produced 107 mm of rainfall within two hours at the Hohe Warte weather station in Vienna.
• This event has an estimated return period of approximately 700 years based on the 84-year Hohe Warte record and regional frequency analysis (interquartile range: 507 to 1383 years).
• Conditioning the return period on average mid-tropospheric temperature reduces the estimated return period to approximately 300 years, suggesting an increased frequency of such events due to climate change.
• The 107 mm event is unprecedented in the Vienna basin; prior to 2024, only two events at Hohe Warte exceeded 50 mm in two hours (62 mm in 2014, 58 mm in 2021), and no other station recorded an event exceeding 62 mm.
• The 100-year return periods for 2-hour precipitation were estimated to be 63 to 77 mm for rain gauge data and 60 to 91 mm for the radar-based INCA data.
• The INCA data set (2004–2023) identified 55 events exceeding 50 mm in two hours, significantly more than the 6 events recorded by the station network, highlighting that gauges often miss extreme, spatially confined events.
• For the 2024 event, INCA data showed that an area of 17 square kilometers was hit by more than 75 mm of rainfall, and 4 square kilometers by more than 100 mm.
• No significant statistical trend was found in the extreme precipitation time series at Hohe Warte, but mean summer cloud temperature (700 to 500 hPa) has been increasing since the 1980s.

Contributions

• Provides a comprehensive quantification of an unprecedented extreme precipitation event in Vienna using one of the world's longest sub-daily precipitation time series (Hohe Warte, 84 years).
• Demonstrates the critical role of radar-based data (INCA) in capturing the spatial extent and frequency of extreme precipitation events that are often missed or underestimated by traditional rain gauge networks, even in densely instrumented urban areas.
• Offers evidence for the impact of climate change on the frequency of extreme sub-daily precipitation by showing a substantial reduction in the event's return period when conditioned on mid-tropospheric temperature.
• Emphasizes the importance of combining long-term station data with high-resolution radar data for robust extreme precipitation estimates crucial for water management and public engineering.

Funding

• Austrian Climate and Energy Fund under the program “ACRP13” (grant no. KR20AC0K17974).

Citation

@article{Klaus2025Brief,
  author = {Klaus, Vinzent and Laimighofer, Johannes and Lehner, Fabian},
  title = {Brief communication: How extreme was the thunderstorm rain in Vienna on 17 August 2024? A temporal and spatial analysis},
  journal = {Natural hazards and earth system sciences},
  year = {2025},
  doi = {10.5194/nhess-25-4807-2025},
  url = {https://doi.org/10.5194/nhess-25-4807-2025}
}