More than 100 forecasters and scientists attended this year’s Using ECMWF’s Forecasts (UEF) meeting, which took place at the Centre from 12 to 16 June 2017. They heard about ECMWF’s plans for future products, services and research and were able to network and share experiences with participants from other countries. Activities ranged from oral presentations, posters and demonstrations to hands-on sessions. The demonstrations provided an opportunity to show and try out software or services developed at institutions based in ECMWF’s Member and Co-operating States.
UEF is one of the channels for ECMWF data users to provide feedback on products and services and to request new products. For the last three years this has been done successfully through the ‘User Voice Corner’. At the end of the meeting, user voices are collated and they subsequently feed into ECMWF’s future plans.
The theme of this year’s UEF was ‘storms’. Severe storms, whether they happen in winter or summer, have considerable impact on people’s lives and may lead to significant disruption to services and commercial activities. They can come with not just strong winds and heavy rain but also hail, lightning, blizzards, floods and storm surges. These can lead to damage and destruction of infrastructure, injury and death. Predicting the onset, intensity and track of severe storms with enough lead time is therefore essential for readiness and damage limitation. Moreover, some applications may require information on weather regime changes in the extended-range forecast, such as the prospect of a stormy period two weeks ahead, or seasonal outlooks.
Workshop: Storm naming – does it work?
Gerald Flemming (Met Éireann) and Will Lang (Met Office) ran this workshop to share the experience of a collaborative project between the two meteorological services initiated in 2015. The project started as a pilot initiative to investigate the effect of ‘naming’ large-scale mid-latitude windstorms on the reach and influence of severe weather advice for the UK and Ireland. Gerald and Will started the workshop by asking the participants how they would describe themselves if they could not say their names. It was a rather difficult exercise, which highlighted the importance of a name as an identifier for a person. Similarly, a named storm will allow the audience to focus more on its potential impact on their lives. The workshop also aimed to collect feedback on how the scheme could help end users to benefit from improved NWP capabilities and how it could enhance the authoritative voice of European national meteorological services.
Numerical weather prediction (NWP) models support meteorological services with forecasts of when a storm will form, where it will strike and how severe it will be, and with an indication of the degree of confidence we can have in the forecast. ECMWF has been at the forefront of NWP development for many years and our Strategy includes the provision of high-quality severe weather forecast products.
The meeting focused on three thematic areas:
- Processing of model outputs to support the forecasting of severe storms and associated weather phenomena.
- Diagnostics involving tools or studies that highlight strengths and weaknesses of ECMWF's Integrated Forecasting System (IFS) in predicting storms.
- Impact of storms on sectoral applications.
“Providing model output which is useful and supports the work of our Member and Co-operating States is at the heart of ECMWF's Strategy,” said Director of Forecasts Florian Pappenberger in the opening lecture of the UEF. He also provided a timeline of IFS upgrades with particular emphasis on IFS Cycle 43r3 implemented on 11 July 2017. A number of products aimed at supporting forecasting activities were also shown. They included the probability of precipitation type, point rainfall, distributions for the monthly forecast, moisture flux and regime transitions.
Ensemble forecasting was presented in many talks: Helen Titley from the UK Met Office gave an invited talk on processing ensemble information. She showed a variety of tailored applications available to Met Office operational meteorologists based on both ECMWF and Met Office ensemble forecast data to help forecast severe extratropical cyclones.
Workshop: Hazard impact modelling for storms
“Participants were encouraged to think about opportunities for impact prediction and risk-based warnings and services,” Ken Mylne from the UK Met Office explains. During the workshop many different weather-related impacts were discussed. Some of the key ideas were that impact predictions are powerful tools in supporting effective communication and helping decision-makers to understand the situation, but that they may still require expert interpretation.
Historical impacts of weather of any type can be related to particular weather regimes, giving probabilities of impact depending on regime occurrence. Global ensembles are ideal for predicting the probabilities of regime occurrence, enabling the prediction of heightened risks of impacts without the need for high-resolution ensembles.
Storms have an impact on transport and power distribution networks. Vulnerabilities in these sectors vary greatly. It is therefore important to use ensemble forecasting systems at higher resolution. Enhanced diagnostics of precipitation type like those developed at ECMWF will help with downscaling in conjunction with high-resolution mapping of transport networks and power distribution networks.
“Predicting high-impact weather events is a crucial task for forecasting centres,” said ECMWF scientist Linus Magnusson. He presented examples of evaluation and diagnostics that can be used to understand the predictability of severe events and stressed the importance of identifying key features in the development of severe weather that can be verified to improve the IFS. During his contribution, ECMWF's review of headlines scores to assess forecast performance was mentioned. The review ties in with ECMWF's Strategy to 2025, which calls for ensemble forecasts at 5 km resolution.
Lara Gunn from MetDesk showed how ECMWF data are used to provide tailored forecasts for the transport and energy sectors. She reminded the audience that heavy rain in the Alps has an impact on Alpine reservoirs and the upper Danube, which in turn affects energy production across France and Eastern Europe. Being able to predict these kinds of event with high accuracy is thus of great importance for fragile European markets.
A presentation from the European Severe Storms Laboratory (ESSL), which organises the annual ESSL Testbed event, showed how the laboratory is used to train forecasters in predicting severe convective storms using state-of-the-art forecasting tools. ESSL can also facilitate the evaluation of experimental and/or new forecast products based on numerical weather prediction models as well as radar and satellites. ECMWF is a member of ESSL.
Paul Knightley (MeteoGroup) presented some fascinating videos and photos of storms mainly from his own storm chasing trips. “Storm chasing is really exciting,” said Paul, “but you have to be really patient. You have to stay put in a place and wait for the storm to form.” While seeing a storm form and following it is really rewarding, it can also be a terrifying experience as the course of the storm may deviate from the expected path and threaten properties and lives.
This year EUMETSAT contributed with a plenary talk and a workshop on satellite data and products as storm monitoring tools and as input into data assimilation schemes. Satellite and conventional data are critical to the development and improvement of NWP. ECMWF and EUMETSAT have worked closely together to demonstrate the value of new satellite observations and to ensure maximum benefit for Member and Co-operating States from investments in the satellite programme.
EUMETSAT has provided its user community with more than three decades’ worth of satellite data. Its current programmes include the procurement and future operation of Meteosat Third Generation (MTG) satellites and EUMETSAT Polar System Second Generation (EPS-SG) satellites. The Infrared Sounder on MTG will provide information with the potential to improve forecasts of severe weather events. In preliminary studies the sounder has been shown to be able to detect the initiation of atmospheric rivers, which are long, narrow and transient corridors of strong horizontal water vapour transport. Such rivers are usually associated with a low-level jet stream ahead of a cold front or an extratropical cyclone. They produce heavy precipitation where they are forced upwards.
Visualising ensemble forecasts
During the UEF 2016 we launched ‘The Challenge’: ECMWF data users could submit proposals to improve the way probabilistic information and in particular the popular ENS Meteogram is displayed. The Challenge was won by Dave MacLeod, Hannah Christensen, Stephen Juriche and Aneesh Subramanian with an entry based on the philosophy of progressive disclosure of information. We would like to thank all those who submitted a proposal and congratulate the winners.