European Severe Storms Laboratory

Convective hazards pose a significant threat to society and are among the most challenging phenomena for forecasting. The increasing spatial and temporal resolution of (European Centre for Medium-range Weather Forecasts) ECMWF Forecasting system makes it increasingly more suitable for applications targeting severe convection. The European Severe Storms Laboratory (ESSL) has developed additive logistic regression models (AR-CHaMo) for hail and severe convective wind gusts using the ERA5 re-analysis, observations of lightning and severe convection from around the world. ECMWF and ESSL collaborate in a project to apply AR-CHaMo to the ECMWF ensemble (ENS) to provide global predictions of convective hazards in the medium range. Probabilistic products for large hail and severe wind gusts from ENS are undergoing thorough testing, evaluation and skill assessment. Latest results suggest that AR-CHaMo forecasts on ECMWF ENS provide skilful predictions of large hail and severe wind gust risk many days in advance. ESSL and ECMWF continue to refine and develop these probabilistic products by testing more predictors, extending training datasets and including more convective hazards such as tornadoes. These products are already available via the ESSL’s Weather Displayer and are undergoing testing during the regular annual Testbeds. Evaluation of AR-CHaMo forecasts has also been performed internally at ECMWF on a number of severe convection events. The use of the ECMWF’s Integrated Forecasting System (IFS) model levels is capable of providing more accurate predictors and hence better probabilistic AR-CHaMo forecast.

ESSL and EUMETSAT have organised several expert workshops on using and visualizing the Lightning Imager (LI) data from the Meteosat Third Generation satellite. Also based on the recommendations gained from the workshop, ESSL developed the LI product visualizations presented to the forecasters at the ESSL-EUMETSAT forecaster training Testbeds in 2024 and 2025. These involved point- and grid-based data, such as the size of observed flashes, group density calculated over small grids, or the spatial extent of the observed flashes. Participants of each of the Testbed weeks were asked to evaluate the products and fill out questionnaires. We present the workshops' main results and the summary of the evaluation provided by the Testbed participants. We concentrate on two issues: how useful the LI data is for nowcasting severe convective storms and which products or visualizations were preferred by the forecasters.Besides that, the ESSL staff has been inspecting the LI data over both convective seasons. Using the experience from 2024, ESSL has produced a guide on using the LI data alongside a detailed description of the LI behaviour for 12 cases of severe and non-severe convective storms. Patterns in LI data were compared with storm hazards on the ground reported in the ESWD. We present some insights gained from the data usage during these two years. This includes some interesting patterns or artifacts, such as a few cases of severe storms showing a sudden dip in the lightning, the occurrence of a lightning ring surrounding the updraft area, or a large fraction of flashes detected from the side of the storm as a reflection from the low clouds.

Abstract. Highly accurate manually-generated labels in aerial and satellite images are used for the training of deep learning-based segmentation algorithms and should be available in large numbers and cover many different scenarios to increase the accuracy and generalization capability of the underlying models. Existing labels can be efficiently reused by photogrammetric projections onto additional overlapping aerial or satellite images, enabling great variability in the appearance of the scenes based on differences in viewing angles and environmental conditions. In this work, we investigate whether the additionally generated training data can effectively lead to an increase in prediction accuracy. To this end, we collected aerial images overlapping with the already annotated Traffic Infrastructure and Surroundings (TIAS) dataset, taken from a large-scale historical database spanning 2011 to 2024, and generated new training data by means of photogrammetric projections of existing labels onto these additional images. Training a Dense-U-Net model on the whole TIAS dataset or a part therefore, with and without additional projected labels, showed that this technique could be beneficial to improve the performance of a model if only a small amount of annotations is available comparatively to a large amount of overlapping aerial images.

Convective storms are an important weather hazard in Europe as shown by the high number of severe wind gusts, large hail, tornadoes, and flash floods recorded each year in the European Severe Weather Database. A recent innovation to the ESWD was the introduction of the new tornado International Fujita scale for rating tornado and wind intensity from damage. In 2023, no fewer than 62182 new reports were entered, and reinsurer Munich Re estimated severe thunderstorms to account for the majority of weather-related losses in Europe in 2023 with a total damage of € 10 billion.At the core of ESSL’s mission is conducting and facilitating research on severe weather at a European level. Over the years, the organisation has grown with support from its members which include most of Europe’s weather services and commercial sector partners. In addition to research ESSL is active in the area of forecaster training and the evaluation of novel forecasting and nowcasting applications at the ESSL Testbed.The recorded multi-year trends of severe weather apparent in the ESWD are often dominated by non-meteorological factors, but for large hail indications are strong that its frequency is changing, illustrated by the new hailstone size record of 19 cm diameter in northern Italy in July 2023. ESSL’s recent models of large hail climatology across Europe and the world support these trends. A key challenge for the research community is to develop methods to estimate trends from ever higher-resolution reanalyses and climate models. This is not straightforward as even the highest resolution models do not resolve tornadoes or microbursts, let alone hailstones, and already show biases at coarse scales.The mentioned work modelling severe weather has given new insights into which environmental characteristics are important to severe weather occurrence. For hail, we additionally studied the conditions under which individual hailstorms in 2021, 2022, and 2023 that were particularly severe. We show the importance of the vertical distribution of buoyancy and wind in a storm-centred reference framework, defined using radar-observed storm motion.High vertical wind shear above the boundary layer and high CAPE above the -10 °C isotherm for hail, and a combination of vertical vorticity and strong streamwise vorticity for tornadoes. ESSL is collaborating with ECMWF to develop forecast tools based on these concepts. That said, many questions remain, for example regarding the pre-convective environment and mountain ranges, and with the developing storms. For instance, an important concentration of severe weather is evident surrounding the Alps. To address related questions, ESSL has taken the initiative for a multi-year multi-national field campaign in central Europe called TIM (Thunderstorm Intensification from Mountains to plains), in which it will collaborate with a large number of research institutes.

A derecho is a widespread, long-lived, straight-line windstorm that is associated with a fast-moving group of severe thunderstorms known as a mesoscale convective system.During 18 August 2022, a highly intense and organized convective storm, classified as a derecho, developed over the western Mediterranean Sea affecting Corsica, northern Italy and Austria, with wind gusts up to 62 m/s and giant hail (~11 cm). There were 12 fatalities and 106 people injured. This event received much attention in the media for its extraordinary impact and the rareness over the Mediterranean Sea. The derecho developed over an extreme marine heatwave that persisted during the whole summer. Therefore, the hypothesis of a relationship between the extreme atmospheric event and the extreme marine heatwave rapidly arose, and thus, a possible link with anthropogenic climate change.This convective event can be considered as extreme from the affected locations point of view (in terms of winds) but also is between one of the most powerful derechos ever recorded in the USA and Europe. Also, the event developed over an extreme marine heatwave that was mainly affecting the western Mediterranean Sea during summer 2022.Here, by performing model simulations with both the NCAR Model for Prediction Across Scales and the Météo-France nonhydrostatic operational AROME model and using an storyline approach, we find a relationship between the marine heatwave, the actual anthropogenic climate change conditions, and the development of this extremely rare and severe convective event. We also find a future worrying increase in intensity, size and duration of such an event with future climate change conditions.

This study compares the convective parameters between two 8-year periods of hail reporting in Poland: 1948–1955 (period A) and 2015–2022 (period B). The data on hail incidents for period A was derived from hail annuals of the Institute of Meteorology and Water Management, while for period B from the European Severe Weather Database. Period A contains more hail reports (18375) compared to the recent period B (3677). ERA5 reanalysis and thundeR package were used to calculate convective parameters for both periods. Convective environments indicated a more favourable environment for the hail occurrence in period B with larger instability, vertical wind shear and storm-relative helicity. The most widespread event with the largest hailstone ever documented in Poland (14 cm) also occurred in period B, on 24 June 2021. However, the mean number of hail reports per day was higher in period A due to a more effective method of hail reporting. In both periods convective environments were also clearly more favourable for hail in the days with hail exceeding 5 cm in diameter compared to the days with smaller hail and in days with the large quantities of hail reports. This study showed that ERA5 was able to capture convective environments reliably well for events as old as the 1950s.

<p>In this work, we use 8 years of OPERA radar data, ESWD severe weather reports, and ATDnet lightning detection data to create a climatology of quasi-linear convective systems (QLCS) across Europe. In the first step, 15-minute composite animations of radar reflectivity and lightning data were used to manually identify 2201 QLCS polygons in the period of 2014 to 2021. Severe weather reports, lightning data, and morphological properties (such as the presence of bowing segments) allowed classifying QLCSs according to their intensity into 1844 marginal, 304 moderate, and 53 derecho cases. The manual evaluation also allowed us to identify: the basic archetype of the system, the location of stratiform precipitation relative to the active convective part, propagation with respect to the mean wind, areal coverage, width, length, accompanying hazards, and social impacts associated with each QLCS. Results indicate that QLCSs are the most frequent during summer in Central Europe, while in southern Europe the season of their occurrence is extended to late autumn. QLCSs are the least frequent during winter when they appear mostly in northwestern Europe in the form of narrow cold frontal rainbands (NCFR). In spring they are most common across western Europe. The vast majority of systems move from southwest and west. 22% of the systems were associated with a bow echo signature and 6% produced a mesoscale convective vortex. Among precipitation modes, trailing (53%) and embedded (45%) stratiform types were the most common. The longest classified QLCS had a length of 2200 km (9 Aug 2018), while the widest reached 1635 km (10 Jan 2015). The most frequent hazard accompanying QLCSs was lightning (produced during 95% of the total QLCSs lifetime), followed by severe winds gusts (7.7%), excessive precipitation (5.3%), large hail (2.7%), and tornadoes (0.4%). Derechos had the largest coverage of severe wind reports with respect to their path area (49%), while back-building QLCS had the largest coverage of excessive precipitation events (12%). Large hail was most common with bow-echo complexes (6.5%).</p>

Abstract The last comprehensive statistics of tornadoes in northern Eurasia (NE) were published more than 30 years ago. This paper introduces a new database of tornadoes in NE that spans from the tenth century to 2016. The database, compiled using various sources, contains 2879 tornado cases over land and water and includes tornado characteristics. Tornadoes are common for most regions of NE, with a density reaching four cases per 104 km2 in 1900–2016 in some regions. Tornadoes over land have distinct annual and diurnal cycles: they form mostly in May–August, with a maximum in June, and during daytime, with a maximum at 1700–1800 local time. Waterspouts form in all months with a maximum in late summer and mostly at 0900–1300 local time. Most tornadoes are weak and short lived. The Fujita-scale intensity is ≤F1 for 80% and ≥F3 for 3% out of all rated tornadoes. Half last less than 10 min. The average annual number of all tornadoes over land is around 150, including 10 and 2 tornadoes with ≥F2 and ≥F3 intensity, respectively. Annually, 1–2 tornadoes lead to casualties and result in 2.9 fatalities and 36.3 injuries. Despite the incompleteness of the dataset, our results show that tornadoes in NE, although being rare, are not as extremely rare as has been thought before. The results illustrate the substantial underestimation of tornado threat by the general public, researchers, and meteorologists, and unambiguously indicate the need for systematic assessments and forecasting of tornadoes by national weather services.

Abstract The processes leading to the development of hail and the distribution of these events worldwide are reviewed. Microphysical and physical characteristics of hail development are described to provide context of the notable gaps in our understanding of what drives hail to grow large, or what determines how it falls to the ground. Distributional characteristics of hail are explored, utilizing both surface observations of hailstones and remotely sensed observational data sets to identify opportunities and needs for new observations. These observational deficiencies contribute to our limited capacity to both forecast hail or its expected size and reduce the effectiveness of using favorable conditions for hail development as a proxy to frequency where observations are unavailable. Given the substantive influences of both climate variability and the changing Earth system on hail, the latest understanding of their contributions to risk are addressed. Applying this understanding of the distribution and physical characteristics of hail, the damage by hail to agriculture and insured property is assessed. Much remains unknown about the processes leading to hail growth and environmental controls on hail occurrence, size, and magnitude, particularly outside of the United States and Europe. A better understanding of the global occurrence of hail is also needed to better anticipate the hazard and associated impacts.

AbstractsThe history of severe thunderstorm research and forecasting over the past century has been a remarkable story involving interactions between technological development of observational and modeling capabilities, research into physical processes, and the forecasting of phenomena with the goal of reducing loss of life and property. Perhaps more so than any other field of meteorology, the relationship between researchers and forecasters has been particularly close in the severe thunderstorm domain, with both groups depending on improved observational capabilities.The advances that have been made have depended on observing systems that did not exist 100 years ago, particularly radar and upper-air systems. They have allowed scientists to observe storm behavior and structure and the environmental setting in which storms occur. This has led to improved understanding of processes, which in turn has allowed forecasters to use those same observational systems to improve forecasts. Because of the relatively rare and small-scale nature of many severe thunderstorm events, severe thunderstorm researchers have developed mobile instrumentation capabilities that have allowed them to collect high-quality observations in the vicinity of storms.Since much of the world is subject to severe thunderstorm hazards, research has taken place around the world, with the local emphasis dependent on what threats are perceived in that area, subject to the availability of resources to study the threat. Frequently, the topics of interest depend upon a single event, or a small number of events, of a particular kind that aroused public or economic interests in that area. International cooperation has been an important contributor to collecting and disseminating knowledge.As the AMS turns 100, the range of research relating to severe thunderstorms is expanding. The time scale of forecasting or projecting is increasing, with work going on to study forecasts on the seasonal to subseasonal time scales, as well as addressing how climate change may influence severe thunderstorms. With its roots in studying weather that impacts the public, severe thunderstorm research now includes significant work from the social science community, some as standalone research and some in active collaborative efforts with physical scientists.In addition, the traditional emphases of the field continue to grow. Improved radar and numerical modeling capabilities allow meteorologists to see and model details that were unobservable and not understood a half century ago. The long tradition of collecting observations in the field has led to improved quality and quantity of observations, as well as the capability to collect them in locations that were previously inaccessible. Much of that work has been driven by the gaps in understanding identified by theoretical and operational practice.