ECMWF Newsletter #185

A new index to provide early alerts for extreme precipitation

Jessica Keune
Christopher Barnard
Fredrik Wetterhall
Francesca Di Giuseppe

 

In the framework of the EU-funded project CENTAUR, which aims to bring innovation to several Copernicus services, ECMWF has developed an innovative Extreme Precipitation Index (EPIX). This new index is designed to support the Copernicus Emergency Management Service (CEMS) in guiding satellite mapping during natural disasters, providing location-specific, unbiased evaluations of event rarity and severity using return periods. The proposed extreme precipitation forecasts were recently among the finalists for the prestigious Harry Otten Prize for innovative ideas in meteorology at the European Meteorological Society annual meeting in Ljubljana, Slovenia.

The challenge of forecasting extreme precipitation

Predicting extreme precipitation remains a significant challenge in weather forecasting. Physical numerical models are typically calibrated to represent mean conditions, making rare and intense events difficult to capture. Higher-resolution simulations, such as those developed under the EU Destination Earth initiative, can improve forecasts of intense precipitation but are costly to run. Next-generation AI-based forecasting systems could also enhance predictions if trained on extreme events, but the scarcity of such events poses a challenge due to their lack of representation in training datasets.

EPIX offers a practical solution for pre-alert purposes. Instead of predicting exact rainfall amounts – which are essential for downstream applications such as hydrology, fire, or drought monitoring – the index identifies locations where extreme precipitation is likely to occur, without specifying precise quantities.

EPIX forecast for extreme precipitation in Valencia in October 2024
EPIX forecast for extreme precipitation in Valencia in October 2024. On 29 October 2024, Valencia experienced over 600 mm of precipitation in a single day, prompting activation of a CEMS mapping acquisition. The area of interest was selected based on ground reports and numerical weather prediction. EPIX identified the event with a Level 3 warning at least three days in advance and provided sufficient lead time to pre-position the required SAR satellites used for urban flood detection. The Valencia event, with estimated damages in the hundreds of billions of euros, represents the costliest extreme precipitation event in Europe on record. (https://www.ecmwf.int/en/newsletter/183/news/extreme-precipitation-spains-valencia-region)

How EPIX works

The index is built on three pillars:

  1. Transforming precipitation into return periods to assess rarity.
  2. Evaluating spatial coherence of return periods to determine the event's extent.
  3. Measuring temporal persistence to gauge severity.

By condensing this information into a single index with three warning levels, EPIX enables authorities to prioritise regions and pre-task satellites for rapid mapping, providing geospatial intelligence for rescue operations on the ground. Its simplicity and compatibility with existing forecasting systems, without significant computational overhead, have generated widespread interest amongst the community.

From innovation to uptake

Although the team did not win the Harry Otten Prize, being amongst the three finalists highlights the innovation and relevance of EPIX in operational meteorology.

Next steps for EPIX include full automation for multiple cities within the CENTAUR project network as part of the demonstration phase. The index is also being tested by the European Response Coordination Centre as a potential new product, supporting the European Crisis Management Laboratory of the Joint Research Centre in efficiently running the CEMS mapping acquisition system.

With its combination of practicality, scientific rigour and operational readiness, EPIX is set to become a valuable tool for emergency management, helping authorities respond more effectively to extreme precipitation events across Europe.

Acknowledgment

The European Union's Horizon Europe research and innovation programme, under grant agreement No. 101082720 – CENTAUR, funded this research. Views and opinions expressed are however those of the authors only and do not necessarily reflect those of the European Union or the Commission. Neither the European Union nor the granting authority can be held responsible for them.