EWC User Workshop 2026 Agenda available!

8:30 - 09:00

Arrival and registration

9:00 - 09:05

Opening and welcome

9:05 - 9:30

Introduction & overview of the current status of the EWC

An introduction to the service with philosophy and key benefits. 

9:30 - 09:50

EWC Technical Basics

Basic service offering and practicalities including instructions to login via IAM and Horizon, using OpenStack Horizon and EWC CLI.

10:20 - 11:05

Service updates

Secret management

Kamil Rajtar (EUMETSAT)

Claudiu Vasile Nagy (EUMETSAT)

Marcos Hermida Mera (ECMWF)

11:05 - 11:30

Community Hub 

Francesco Murdaca (EUMETSAT)

11:30- 11:50

Building the MLCast data ecosystem on the foundations of the EWC

In this presentation I will outline my experience with using the European Weather Cloud (EWC) to serve as the data hosting service for the MLCast Deep Learining Nowcasting Community (https://mlcast.org), https://github.com/mlcast-community/mlcast-datasets. I will walk through how mlcast-datasets (https://github.com/mlcast-community/mlcast-datasets) was built, how it works and how we make use of EWC to enable the research and development we do on nowcasting using deep learning.

You have heard it before: in the age of deep learning and big data we need well-defined and accessible datasets (or ARCO, Analysis-Read Cloud Optimised). In this talk I will demonstrate how we achieved this in the MLCast community and show how you can do the same in your situation.

11:50 - 12:10

KAL (Kernel Atmospheric Learner): A Hybrid MSG SEVIRI and Machine Learning Framework for Quantitative Precipitation Estimation in Radar-Absent Regions

This presentation will introduce the (Kernel Atmospheric Learner), an hybrid machine learning framework, developed and trained on the European Weather Cloud (EWC), for high-resolution daily Quantitative Precipitation Estimation (QPE) in topographically complex regions that lack ground-based weather radar.

12:10 - 12:30

Visualisation platform for AI Models

RMI runs a lightweight forecast viewer on EWC that turns AI-LAM weather model output into an interactive map, letting users browse variables and lead times through a simple web interface.
Forecasts are also made available directly via an S3 bucket for users who want programmatic access to the raw data.
Together, the viewer and the data bucket show how a relatively simple EWC setup can support both visual exploration and direct data access with minimal overhead.

13:30 - 13:50

Studying the evolution of heat waves in Italy over the present century.

The study evaluates a suite of climate indicators related to summer heat waves in Italy, aiming to construct a comprehensive, quantitative profile of their evolving characteristics, specifically focusing on frequency, duration and severity over the present century. The base data layer consists of statistically downscaled and bias-corrected projections for Italy, developed by CMCC and Agenzia ItaliaMeteo. The temporal scope of the study is restricted to an extended summer season spanning May to September, capturing the period when heat waves are most acute and pose the greatest risk to human health. Starting from the daily TMAX projections, we calculated the indicators year by year, up to 2100. The reference period, 1985-2014, is used both as the representative distribution to calculate the percentiles and as the time frame to calculate climate normals. Consequentially, we computed the metrics both as absolute values and as anomalies with respect to the 1985-2014 normals. We studied the evolution of those distributions across the chosen periods, in terms of their mean value and standard deviation. Both these two parameters show a projected evolution towards higher values, meaning the occurrence of extreme events from the tail of the distribution are expected to become more frequent. We also studied some specific areas in more detail.
The study was conducted through a tenancy of the European Weather Cloud dedicated to the Italian initiative for C3S NCP.

13:50 - 14:10

Sentinel 5 - Processing chain for H2O, OClO and BrO at the European Weather Cloud implemented with DLR's BPMN-based Workflow Management Processor (WMP)

The presentation will cover the BPMN-based workflow management and its deployment and testing at the EWC.

14:10 - 14:30

Deploying and running satproc on the EWC for processing KNMI ACSAF workloads

OpenGeoWeb - Web based meteorological workstation now on EWC Community Hub

This session presents the OpenGeoWeb meteorological workstation hub item now available on the EWC Community Hub, showing how this open‑source GeoWeb‑based platform enables interactive visualisation and analysis of weather and climate data directly in the cloud. It highlights how this item can be deployed on Kubernetes and what can be customized for the users. GeoWeb is developed by KNMI, FMI and MET Norway.

FLEXPART‑IFS on the EWC Community Hub: A Practical Guide from Prototype to Published Hub Item

16:00 - 16:25

Discussion and question for feedback

Hybrid discussion, favouring the in-person discussion. Opportunity for feedback

Xavier Abellan (ECMWF)  & Roope Tervo (EUMETSAT)