Alex Priestley (EUMETNET)
Title: Opportunistic weather observations in EUMETNET: past, present and future.
Abstract: Opportunistic observations form an important part of the networks operated by EUMETNET. Some of these observations are now mature operational networks, giving important benefits to weather forecasting models, such as GNSS-derived water vapour measurements and derived aircraft observations. Other sources of information, such as personal weather station data, are used widely on a national basis, but unlocking their wider benefits is currently held back by the ability to share data across Europe. In this talk I will cover some of EUMETNET’s past work to help exploit opportunistic observations across Europe, and some plans for future opportunistic data use including webcams and smartphone data.
Biography: Alex Priestley is the EUCOS Network Manager for the EUMETNET Observations area, based at the UK Met Office in Edinburgh. Following an undergraduate degree in Geophysics and Meteorology, he worked in observations data analysis and quality control, and then as a senior weather forecaster in Scotland. He gained his PhD in snow hydrology modelling and geophysics in 2022, and has since worked as a lecturer at the Met Office college, before taking his current role. His expertise is in meteorological observation systems and data, operational forecasting, snow modelling and international project coordination.
Stefanie Hollborn (DWD)
Title: Enhancing Numerical Weather Prediction with Opportunistic and Crowdsourced Observations: Insights from NetAtmo and Microwave Link Assimilation
Abstract: The integration of opportunistic and crowd sourced observations into numerical weather prediction (NWP) presents a promising path toward enhancing high-resolution, short-term forecasting.
This talk explores the potential of crowd-sourced weather data-specifically from NetAtmo personal weather stations and Commercial Microwave Links (CMLs)-to improve NWP skill. We will present findings from two recent research projects that evaluate the assimilation of these novel data sources using state-of-the-art ensemble data assimilation systems.
First, results from a study on NetAtmo near surface temperature and humidity observations demonstrate the added value of high-density data in regional NWP through assimilation into the convection-permitting ensemble system ICON-D2 with the KENDA data assimiliation framework. Assimilating NetAtmo with proper quality control and bias corrections shows a positive impact, underlining the feasibility and benefit of integrating citizen weather observations into operational systems.
Second, we report on the RealPEP project, which focuses on assimilating CML attenuation data as a proxy for path-averaged rain rates. Using the ICON-D2 model and EMVORADO radar forward operator, CML-specific attenuation is assimilated via the Localized Ensemble Transform Kalman Filter (LETKF). Case studies and statistical evaluation reveal that CML assimilation effectively enhances convective initiation and short-term precipitation forecast skill. Improvements in fractional skill scores (FSS) for radar reflectivity forecasts further quantify these benefits.
Together, these results underscore the growing importance of opportunistic sensing-both crowd-sourced and infrastructure-derived-in modern meteorology, providing valuable complementary information in data-sparse regions and enhancing the resilience of forecasting systems.
Biography: Stefanie Hollborn studied mathematics and philosophy at the universities of Mainz and Toronto. She obtained her doctorate in numerical mathematics, focusing on electrical impedance tomography, and transitioned to the research and development department of the German Weather Service (DWD) in Offenbach in 2016. Since 2022, she has been leading the department for observation modeling and verification. Her focus includes, in addition to maintaining and overseeing the operational data assimilation routines of the DWD and the continuous quality control of the numerical weather prediction system, the further development of the data assimilation methods used at DWD, the integration of new observations, and the development of observation operators. Currently, part of her work is centered on the development of AI-based methods for weather and climate services.
Li-Pen Wang (National Taiwan University )
Title: AI in hydrometeorological applications: From nowcasting to climate modelling
Abstract: In recent years, Artificial Intelligence (AI) has led to significant breakthroughs across many research fields, including hydrometeorology. However, its broader application remains constrained by its “black box” nature. In this talk, Dr Li-Pen Wang will present his team’s latest work on developing data-driven approaches to tackle hydrometeorological challenges. He will highlight how AI can enhance radar rainfall estimation and nowcasting, and how it can help bridge climate dynamics with local rainfall modelling. Dr Wang will also explore how OpenSense project can open new opportunities with help from AI for future research.
Biography: Dr. Li-Pen Wang is a civil engineer specialising in hydrometeorology, applied statistics and software development. He currently leads a research team at National Taiwan University (NTU), focusing on data-driven solutions to hydrometeorological challenges. He is also a co-founder of a UK-based start-up that provides consulting services to water companies, national meteorological agencies and (re)insurance companies. Recently, he was invited to join the Expert Roster of the International Monetary Fund (IMF), where he advises on modelling the impacts of climate change on flooding and the corresponding socio-economic losses.
