The response of mean and extreme precipitation to anthropogenic global warming stems both from warming of the troposphere and dynamical changes in the large-scale circulation, especially upward motions. The interaction between these two components complicates future projections and makes the attribution of extreme precipitation events challenging, both using conditional (e.g., analog-based) and unconditional (e.g., extreme value theory-based) methods. In this study we reflect upon this problem and propose some possible solutions to tackle it starting from the case study of Storm Boris, that led to major floods over central Europe in mid-September 2024. The first step is the identification of key circulation features associated with the event, whose representation is deemed crucial to obtain realistic analogs: the presence of a slow-moving, upper-level potential vorticity (PV) cutoff, the peculiar track of the surface cyclone associated with Boris, and the presence of anomalously strong forcing for ascent. Circulation analogs of Boris are then identified in a large ensemble of present-day and future-climate simulations with the CESM1 model, to understand how "Boris-like" storms will change in an end-of-the-century high warming scenario. We find that the combined use of upper-level PV and of a surface cyclone identification algorithm substantially improves the quality of the analogs, both in terms of the large-scale flow pattern and the precipitation associated with the cyclone. Analogs of Boris restricted to the same season in a warmer climate feature on average less precipitation, due to an overall weakening of upper-level-driven ascent over Europe. However, analogs of Boris not restricted to the same season show a seasonality shift: they become less frequent at the end of the warm season and more frequent in the shoulder seasons - when the dynamical and thermal conditions of September in the present-day climate can be recovered again -, and exhibit an increase in mean precipitation in the warmer climate. The results obtained from the analog-based approach are then compared with an unconditional, statistics-based approach focusing only on the seasonal and yearly maxima of precipitation: the latter approach allows to recover the expected intensification of extreme precipitation in a warmer climate - at the price, however, of considering events that do not necessarily have the same dynamics as Storm Boris. The sensitivity of attribution outcomes with respect to implicit and explicit methodological choices is discussed in detail. The systematic comparison of different approaches, the two-step methodology to obtain more reliable analogs of heavy precipitation events, and the focus on process understanding are key ingredients of this study, with general implications for investigating the role of climate change for specific weather extremes.

Reliable fine-scale hail projections are needed for robust risk assessments, yet future trends remain uncertain and contradictory across studies. Using 11-year pan-European convection-permitting climate simulations for the present and a + 3 °C pseudo-global-warming climate, we compare an online hail-growth diagnostic (HAILCAST) with an offline machine learning model (XGBoost) trained on ERA5 hail environments. Under present conditions, both approaches provide a plausible European hail climatology. However, XGBoost predicts widespread hail suppression in a warmer climate, mainly driven by increasing freezing-level heights outside the training distribution. HAILCAST instead simulates how enhanced storm updrafts can sustain hail growth despite a warmer atmosphere, projecting increases over central-eastern Europe and larger hailstones. The results show that data-driven approaches alone should be used with caution under global warming, because the relationships learned under present-day conditions are not climate-change invariant, underscoring the need for physically based hail representations in convection-permitting models to robustly assess future convective hazards.

Precipitation from mixed-phase clouds at high-latitudes is difficult to represent correctly in numerical weather prediction models. Paired water vapour and precipitation isotope measurements provide a constraint on the integrated effect of evaporation and condensation processes, but have rarely been collected in a way that allows to use these for model validation and improvement. Here we present a collection of spatially distributed measurements of water isotopes in the different phases at high time resolution during the ISLAS2021 field campaign over the period 15 to 30 March 2021. The main observational site of this campaign was Andenes, Norway (69.3144 degrees N, 16.1194 degrees E). Isotopic measurements were conducted simultaneously at sea level and a mountain observatory, as well as additional coastal sites at distances of 120 km (Troms & oslash;, Norway) and 1100 km (Bergen, Norway), enabling the assessment of spatial representativeness of vapour isotope measurements. Precipitation samples for water isotope analysis were collected on site at sub-event time resolution, and along a transect across the Vester & aring;len archipelago. These measurements were complemented by a suite of aerosol measurements, including ice-nucleating particles, and additional in situ and remote sensing observations of meteorological variables. During the two weeks of the ISLAS2021 field campaign, frequent alternations between mid-latitude and arctic weather systems were encountered, providing a range of different cases for more detailed process studies. The dataset is available at 10.1594/PANGAEA.984616 , and can serve as a test bed for assessing the spatial representativeness and sampling strategies for water isotope measurements on meteorological time scales. Furthermore, we anticipate our data to be useful in various aspects related to cloud microphysics, for example the quantification of riming processes in convective clouds, the role of ice nucleating particles in marine cold-air outbreaks, and on the condensation efficiency of mid-latitude storms.

Severe convective storms were the costliest natural hazard globally in 2023, with hail as a major driver of economic losses. Single hail events regularly cause damage exceeding 4 billion U.S. dollars in Europe (e.g., France 2022, Italy 2023). The substantial risk of hail prompted the research initiative scClim in Switzerland, which unites expertise from multiple disciplines to advance the understanding of hail risk and its impacts in a changing climate across central Europe. Our approach combines a unique set of hail observations from high-resolution polarimetric radars, automated surface-based sensors, drones, and crowdsourced reports with European-wide convection-permitting climate simulations featuring an online hail diagnostic (HAILCAST). We further developed an open-source, seamless hail impact modeling platform together with stakeholders. The platform provides hail event hindcasts, forecasts, and impact assessments for vehicles, buildings, and crops, using the CLIMADA risk modeling framework. Our climate simulations, generating 11-yr hail climatologies for both the present climate and a +3 degrees C warming scenario, show increased hail frequencies in northeastern Europe and decreased frequencies in southwestern Europe. Hailstorm track analyses reveal larger maximum hail sizes, more extensive hail swaths, and intensified precipitation and wind for cells producing large hail. Consequently, the future hail damage potential to buildings increases, while agricultural impacts present a more complex picture: Earlier growing seasons reduce crop exposure to hail, but regional increases in hail frequency amplify overall risk. These findings provide novel insights for developing adaptation strategies in sectors vulnerable to hail damage in a warming climate.

Ice nucleating particles (INPs) catalyze primary ice formation in Arctic low-level mixed-phase clouds, influencing their persistence and radiative properties. Knowledge of the abundance, sources, and nature of INPs over the remote Arctic Ocean is scarce, particularly in the Eurasian Arctic. In this work, we present summertime measurements of INP concentrations (NINP) in immersion mode from the ship-based Arctic Century Expedition exploring the Barents, Kara, and Laptev Seas and the adjacent high Arctic islands and archipelagos during August to September 2021. Atmospheric NINP were found to be lower than in continental high-latitude sites, particularly at temperatures below −15 °C, suggesting a lower abundance of mineral dust INPs. The geographical NINP variability in the Eurasian Arctic shows that the highest NINP are observed when the ship was in the ice-free ocean, marginal ice zones (MIZ), and in the vicinity of land. Very low NINP were measured within the ice pack. The peak NINP was observed north of Novaya Zemlya where backward trajectories indicate air parcels arriving from the western Siberian coast. Overall, we find that INP sources are local to regional, with little evidence for long-range transport to the investigated area of the Eurasian Arctic in summer months.

Abstract Extreme precipitation events can have severe impacts on society and the environment. Understanding what causes these events is a vital step toward better prediction and improved disaster preparedness. One research direction is to answer the question: Where did the moisture that rained here come from? The moisture sources for precipitation (i.e., where the moisture originally evaporated) cannot be measured directly and, therefore, a variety of different moisture-tracking methods have been developed and evolved over time. To better understand the uncertainty of these methods, we unite the community to advance common understanding and guidelines. As the first step, in this study, we quantify moisture sources of three extreme precipitation events using methods obtained from 14 different research groups. These three events cover different meteorological conditions: monsoon precipitation in Pakistan, convective precipitation in Australia, and atmospheric river-associated precipitation over Scotland. We find that for the three cases, the different moisture-tracking methods qualitatively agree in moisture source patterns, but there are regional and quantitative differences. For example, for the Pakistan case, the recycling ratio shows a multimethod spread of 2%–20%. We also find similar behavior across methods for the three different events, where methods consistently show either more recycling or more sources further away from the precipitation region. This coordinated model intercomparison facilitates the explanation and quantification of uncertainty, acting as a point of reference and inspiration for future work and literature on moisture tracking. Significance Statement Extreme precipitation events have severe impacts on society and the environment. Understanding what causes these events is a vital step toward better prediction and improved disaster preparedness. One research direction, to gain insights into the driving processes for moisture transport leading to extreme precipitation events, is to answer the question: Where did the moisture that rained here come from? As there are no direct observations available to answer this question, a variety of computational methods have been developed over time to determine the sources of precipitation. To better understand the uncertainty of these methods, we quantify the spread in moisture sources across 14 methods for three extreme precipitation events. This work is a point of reference and inspiration for future moisture-tracking studies.

Moist diabatic processes – such as air-sea fluxes, turbulent mixing, cloud microphysics – are key drivers of midlatitude high-impact weather. These processes affect the atmospheric temperature distribution and stability, thereby directly modifying mesoscale circulation patterns. Mesoscale structures, in turn, tend to be the most hazardous features within midlatitude weather systems and are closely linked to forecast uncertainties. We refer to these features as mesoscale moisture-cycling structures (MOCs): anomalies in moisture and wind fields on scales of approximately 1-50 km, embedded within midlatitude weather systems such as extratropical cyclones, their fronts and airstreams. It remains a major challenge to correctly represent moist diabatic processes and their impact on MOCs in numerical weather models.Recent airborne field campaigns in tropical and polar regions have demonstrated the power of water isotope observations to quantify and disentangle the role of different diabatic processes. Building on this approach, NAWDICiso, i.e. the isotopic component of the North Atlantic Waveguide, Dry Intrusion, and Downstream Impact Campaign (NAWDIC, January – March 2026) aimed at conducting multi-platform observations of water vapour isotopes on two aircrafts (French ATR-42 operated by Safire and German Cessna F406 D-ILAB operated by TU Braunschweig) and at ground-based stations in Brittany (operated at the KITcube together with KIT), Ireland as well as within a European-wide precipitation sampling network to survey the downstream impact of North Atlantic cyclones. This intensive measurement period enables us to capture the imprint of diabatic processes on MOCs through simultaneous observations of stable water isotopes in water vapour and precipitation. Here, we present a first overview of the collected data and selected case studies from the NAWDICiso observation network. These measurements, combined with km-scale resolution isotope and tagging-enabled numerical model simulations, provide the basis for identifying and characterising moist diabatic processes within MOCs. Ultimately, these observations deliver unprecedented three-dimensional insights into MOCs in midlatitude weather systems, which are essential for improving forecasts of the development, intensification, and surface impacts of these weather systems.

Hailstorms have shown rising severity and frequency in recent years, posing a growing threat to crops and presenting significant challenges for the agricultural and insurance sectors in the face of climate change. As part of an interdisciplinary project (scCLIM, Seamless coupling of kilometer-resolution weather predictions and climate simulations with hail impact assessments for multiple sectors), this study focuses on assessing the impact of future hail occurrence on wheat across Europe.We utilize results from high-resolution climate simulations with a grid spacing of 2.2 km, which were conducted using the COSMO regional climate model for both current and future climate. The future climate simulation, targeting a 3°C global warming scenario, was performed using the pseudo-global warming approach. Hail activity was simulated using the hail growth model HAILCAST, which was embedded within COSMO. A model of wheat phenology was used to estimate the wheat harvest dates based on COSMO outputs, enabling an assessment of the present and future exposure of wheat to hail. By integrating high-resolution climate simulations with a crop phenology model, this approach bridges the gap between agricultural production and climate risks associated with extreme events. In this contribution, we examine the temporal and spatial alignment between hail events and crop development, with a particular focus on assessing the sensitivity of future risk of hail damage to wheat with respect to the interplay between changes in hail occurrence and earlier harvest dates. The results reveal regional variations in hail impacts on wheat across Europe, offering valuable insights into crop management, climate change adaptation strategies, and risk assessment within the insurance sector.

Hail severely impacts humans, crops, and infrastructure. Quantifying future hail trends is extremely challenging due to the complex dynamic, thermodynamic, and microphysical processes behind severe convective storms. Here, we combine a km-scale convection-permitting regional climate model and an online hail diagnostic to quantitatively assess changes in hail frequency in Europe imposed by a C global warming level. Results show increases in summer hail frequency in northeastern Europe and decreases to the southwest for intense and severe hail days, related to changes in low-tropospheric water vapor content, convective available potential energy and convective inhibition. Small hail days generally decline across continental Europe, due to increased melting of hailstones with higher melting level height. The physical-based simulation approach captures convection and hail processes consistently, providing a solid basis for assessing the socioeconomic implications of hail and its trends with global warming.

The isotopic composition of water vapor can be used to track atmospheric hydrological processes and to evaluate numerical models simulating the water cycle. Accurate model-observation comparisons require understanding the spatial and temporal variability of tropospheric water vapor isotopes. The challenging task of obtaining highly resolved water vapor isotopic observations is typically addressed through airborne measurements performed aboard conventional aircraft, but these offer limited microscale insights. This study uses ultralight aircraft observations to investigate water vapor isotopic composition in the lower troposphere over southern France in late summer 2021. Combining observations with models, we identify key drivers of isotopic variability and detect short-lived, small-scale processes. The key findings of this study are that (i) at hourly and sub-daily scales, vertical mixing is the primary driver of isotopic variability in the lowermost troposphere above the study site; (ii) evapotranspiration significantly impacts the boundary layer water vapor isotopic signature, as revealed by the delta 18O-delta D relationship; and (iii) while water vapor isotopes generally follow large-scale humidity patterns, with separation distances that might range up to 100-300 km, they also reveal distinct small-scale structures (approximately hundreds of meters) that are not fully explained by humidity variations alone, highlighting sensitivity of water vapor isotopic composition to additional fine-scale processes. The latter are particularly evident for delta D, which also exhibit the largest differences in horizontal and vertical gradients. Combined with other airborne datasets, our results support a simple model driven by surface observations to simulate tropospheric delta D vertical profiles, improving surface-satellite comparisons.

Supercell thunderstorms are the most hazardous thunderstorm category and particularly impactful to society. Their monitoring is challenging and often confined to the radar networks of single countries. By exploiting kilometer-scale climate simulations, we have derived a previously unknown characterization of supercell occurrence in Europe for the current and a warmer climate. The current climate shows several hundred supercells per convective season. Occurrence peaks are colocated with complex topography, e.g., the Alps. The absolute frequency maximum lies along the southern Alps and minima over the oceans and flat areas. Comparing a current-climate simulation with a pseudo-global warming +3°C global warming scenario, the future climate simulation shows an average increase of supercell occurrence by 11%. However, there is a spatial dipole of change with strong increases in supercell frequencies in central and eastern Europe and a decrease in frequency over the Iberian Peninsula and southwestern France.

Hail and lightning, associated with severe convective storms, can cause extensive damage to infrastructure, agriculture, and ecosystems. Because of the small scale of these storms and the complexity of the involved processes, observing and modeling convective storms is challenging. The potential of online diagnostics in convection-permitting models to simulate hail and lightning, especially over climatic time scales and extended regions, has not yet been fully exploited. To address this gap, we present a European-wide hail and lightning climatology (2011-2021) using the Consortium for Small Scale Modeling (COSMO) regional climate model with a horizontal grid spacing of 2.2 km, coupled with a hail growth model (HAILCAST) and the lightning potential index (LPI) diagnostics. We further developed a new European-wide hail product based on the Operational Program for the Exchange of Weather Radar Information (OPERA) composite. Model validation against observations demonstrates an overall good performance in simulating hail and lightning on spatial, seasonal, and diurnal scales. The highest hail frequencies occur during summer along the slopes of high mountain ridges, such as the Alps, Pyrenees, and the Carpathians, aligning with observed lightning hotspots in Europe. In autumn, hail and lightning occur predominantly over the Mediterranean and along the Adriatic coast. Severe hail events with a maximum hail diameter larger than 20 mm mainly occur in the Po Valley, western Spain, and Eastern Europe. This 11-year simulation provides a European-wide data set of severe convective storms and their properties, serving as a basis for further studies of convective events and their impacts.

Marine cold-air outbreaks (mCAOs) are a characteristic type of high-impact weather in the European Arctic and are characterized by an intense water cycle where polar cloud processes play an important role. Model simulations and weather forecasts of mCAO events are challenging and associated with poor predictability. One reason is that processes related to the water cycle interact with one another on a wide range of scales. In regional models, some of these processes are resolved and others are fully or partly parameterised. To test and improve numerical weather prediction models, additional observations and novel types of measurements of water vapour are highly demanded. Stable water isotopes are an increasingly available measurement, allowing to trace sub-grid scale processes, and providing the potential to constrain the mass budget of the atmospheric water cycle during mCAO events. During the ISLAS2022 field experiment (21 March to 10 April 2022), the stable isotope composition of water vapour and liquid samples, cloud structures, and other meteorological parameters were collected between Svalbard and Northern Scandinavia on various measurement platforms. Airborne survey flights to Svalbard provided the ocean evaporation signature and subsequent processing of water vapour during mCAO conditions. During a number of flights, mCAO airmasses were repeatedly sampled over a course of hours to days, allowing to characterize their thermodynamic evolution as clouds were first forming, then glaciating and precipitating. In addition, vapour isotope and sea water isotope measurements were taken continuously onboard R/V Helmer Hanssen between Tromsø and the Greenland west coast. Finally, coordinated land-based measurement activity over Northern Norway and Sweden allowed collection of precipitation samples, thus closing the mass budget of the mCAO events. Furthermore, using buoyancy-controlled meteorological balloons launched from Ny Ålesund, we additionally obtained continuous in-situ measurements of the boundary-layer evolution during the mCAO. We provide an overview over the airborne and ground-based measurement activities during the campaign and provide several examples to highlight the potential of the stable water isotope measurements to constrain the water budget of mCAOs in conjunction with traditional meteorological observations.

Historical and future hail trends over Europe generally point to an increased hail threat. However, these trends often diverge at the regional scale. In particular, Western and Southern Europe show conflicting signals with some models and observations indicating more frequent hail events, while others suggest a decline in a warmer climate. Most existing future projections of hail occurrence rely on hail proxies, estimates based on environmental conditions indicative of hail formation, derived from global and regional climate models that use parameterized convection schemes. Recent developments have introduced more advanced statistical hail models and refined proxies. Despite these advancements, systematic comparisons of different hail proxies - especially when derived from a common dataset - remain limited. In this study, we utilize high-resolution (2 km), convection-permitting regional COSMO climate simulations with the embedded online hail diagnostic HAILCAST to assess the present day and future hail occurrence in a 3°C pseudo global warming scenario. We compare hail frequencies and hail frequency changes derived from (i) established hail proxies based on environmental thresholds and statistical models and (ii) the HAILCAST online diagnostic. Our goal is to evaluate how spatial and temporal patterns of hail occurrence differ between methods and to assess the associated uncertainties in hail trend projections across Europe.

In September 2023, Storm Daniel formed in the central Mediterranean Sea, causing significant socioeconomic impacts in Greece, including fatalities and severe damage to agricultural infrastructure. Within a few days, it evolved into a tropical-like storm (medicane) that made landfall in Libya, likely becoming, to our knowledge, the most catastrophic and lethal weather event ever documented in the region. This study places Storm Daniel as a centerpiece of the disasters in Greece and Libya. We conducted a comprehensive analysis that links a cyclone system with hazardous weather conditions relevant to extreme precipitation, floods and significant sea wave activity. In addition, we examine Daniel's predictability in different development stages and draw connections with previous case studies. Given the climatologically extreme precipitation produced by Daniel, we examine the capacity of numerical weather prediction models to capture such extremes, and we finally investigate potential links to climate change. Daniel initially developed like any other intense Mediterranean cyclone, including medicanes: due to upper-tropospheric forcing followed by Rossby wave breaking. At this stage, it produced significant socioeconomic impacts in Greece. As it intensified and attained tropical-like characteristics, it developed markedly just prior to landfall, reaching peak intensity over land. Considering the short lead times (around 4 d), the cyclone formation exhibited low predictability, whilst landfall in Libya was more predictable. Our analysis of impacts highlights that numerical weather prediction models can capture the extreme character of precipitation and flooding in both Greece and Libya, providing crucial information on the expected severity of imminent flood events. We also examine moisture sources contributing to extreme precipitation. Our findings indicate that large-scale atmospheric circulation was the primary driver, drawing substantial water vapor from the eastern Mediterranean, the Black Sea and continental Europe. The intensification of Storm Daniel was likely driven by anomalously warm SST in the Mediterranean and Black Sea, enhancing evaporation and contributing to the extreme precipitation along the Libyan coast. Finally, our analysis supports the interpretation of its impacts as characteristic of human-driven climate change but also highlights the exceptionality of this cyclone, especially in its medicane phase, which complicates the comparison with other cyclones.

The Southern Ocean is a key component of the climate system, where clouds especially matter. Therefore, it is important to correctly simulate clouds in climate models. Even though there has been substantial improvement, climate models still struggle in their representation of cloud microphysical properties. In this study, based on data from the Antarctic Circumnavigation Expedition in 2026/17, we explore environmental factors, such as stable water isotopes in atmospheric water vapor, cyclones and boundary layer stability, that influence the abundance of aerosols and their size distribution, the most important variables for particles to act as cloud condensation nuclei (CCN), along a latitudinal gradient from 35°S to 75°S. Moreover, we use a cloud parcel model to estimate the cloud droplet number concentration and cloud maximum supersaturation (SS) based on the particles’ size distribution, hygroscopicity and measured updraft velocities. Based on the latitudinal gradient of observed CCN, which features a distinct minimum around 60°S, and the carbon monoxide mixing ratios, which reach background levels south of 60°S indicating absence of anthropogenic influence, we compare aerosol properties north and south of this latitude. The northern aerosol population features two distinct Aitken modes, a nucleation mode and a mode with a Hoppel minimum around 60 nm. The presence of cyclones reduces the particle number concentrations over all diameters. We also observe a stronger Aitken mode presence in unstable boundary layer conditions, where downward mixing of freshly formed particles in the outflow of clouds in the free troposphere can occur. The southern population features only three modes, a nucleation mode and two distinct bimodal distributions with Hoppel minima around 70 nm. Only in stable boundary layer conditions an Aitken mode emerges in the 75th percentile that is larger in particle number than the accumulation mode, pointing towards a potential source of condensable vapors from the ocean surface that grow the Aitken mode, leading to observably higher kappa values. The Aitken mode is further associated with air masses with relatively less depletion in d18O, pointing towards a marine source further north. The cloud droplet number concentration simulations feature the same latitudinal pattern as the measured CCN with the “dip” around 60°S. This is consistent with droplet observations from satellites. Interestingly, the simulated cloud maximum SS tends to increase with latitude, from roughly 0.27% at 40°S towards 0.43% at 75°S. To estimate the sensitivity of clouds towards available aerosol particles, we form the ratio of the particle number concentration larger than the observed Hoppel minimum over the simulated cloud droplet number concentrations. We find that clouds north and south of 60°S experience elevated sensitivity (ratio < 1) to aerosol concentrations in 23 % and 27 % of the time, respectively. This demonstrates that the Southern Ocean cloud regime is indeed sensitive to aerosol number and size distributions, which in turn are influenced by synoptic features (e.g., cyclones) and marine boundary layer stability. On the other hand, frequent occurrence of low SS, demonstrates that cloud formation is also often updraft limited.

The atmosphere is an important reservoir for the essential elements selenium (Se) and sulfur (S) as well as for the toxic element arsenic (As). Atmospheric deposition is a source of these elements to terrestrial and marine environments, which can affect ecosystems and human health. The mobility and bioavailability of Se, S, and As in surface environments depend on their chemical forms (speciation). The factors that determine elemental speciation in atmospheric deposition are likely controlled by the speciation of these elements at the source (atmospheric emissions) and by their (bio)chemical transformations during transport. In addition, atmospheric transport of trace elements and their deposition patterns might be strongly linked to the atmospheric water cycle in particular cloud and precipitation formation, because wet deposition during precipitation is an important removal mechanism of trace elements from the atmosphere. To investigate the dynamical processes that govern the cycles of atmospheric water and trace elements in polar regions, including their sources, transport pathways, and sinks, we performed various chemical measurements (total element concentrations and speciation of Se, S and As) on atmospheric samples collected during the Arctic Century Expedition in the Kara and Laptev Seas (August-September 2021). Notably, trace element analyses were combined with a 4-week continuous time series of ship-based measurements of the isotopic composition of water vapour (i.e., δ2H and δ18O). Air parcel backward trajectories were used to identify atmospheric transport patterns of elemental and water isotope signatures, based on three-dimensional wind fields from the ERA5 atmospheric reanalysis dataset. Based on our chemical and meteorological observations and transport diagnostics, we present new insights into the variability of Se, S, and As concentration and speciation in atmospheric deposition and how they are linked to the atmospheric polar water cycle.

Thunderstorm-related severe weather, in particular hail, causes extensive damage to life and infrastructure in the Alpine region. However, changes in hail impact due to a warmer climate are still not fully understood. In the scClim project, convection-permitting regional climate simulations over Europe using the model COSMO with a ~2.2 km horizontal resolution have been conducted for present-day climate conditions (2011-2021) and a climate scenario with a 3°C global warming using a pseudo-global-warming approach. ERA5 reanalyses were used as boundary conditions and a CMIP6 simulation (MPI-ESM1-2-HR) to infer the large-scale climate-change signal. The integrated online diagnostic HAILCAST is used to calculate maximum hail size. The simulations provide total precipitation and maximum hail size estimates every 5 minutes, which allows for hail cell tracking in the climate simulations and the analysis of hail events in a warmer climate. Validation of the present-day simulation against observations of temperature, precipitation and hail shows an overall good model performance. For hail in particular, radar-based, station-based and crowd-sourced observations have been used to assess the model performance in simulating hail on spatial, diurnal and seasonal scales. The validation outcome encourages further study of the climate signal of hail as simulated with the pseudo-global-warming approach. We will show projected changes in the spatial distribution and seasonal cycle of hail over Europe as well as changes in lifetime, storm area and location of hail cells due to a 3°C global warming.

Hailstorms are among the most destructive weather events, posing significant threats to infrastructure, agriculture, and human life. This study applies hailstorm-tracking diagnostics to kilometer-scale, decade-long climate simulations over Europe using the COSMO v6 model driven by ERA5 reanalyses. Convection is treated explicitly, and hail is modeled online with the HAILCAST parameterization. Simulations represent current and future climate simulations, the latter corresponding to a +3 K global temperature increase implemented via a pseudo-global warming approach.We analyze high-frequency hail output at 5 min intervals, which enables tracking similar to 40000 hailstorms in Europe in current and future climate simulations separately. Storm track properties include length, duration, hail size, and spatial distribution, while three-dimensional environmental variables along these tracks yield storm-centered composites of hailstorm structure and allow for the examination of storm inflow environments. Our analysis reveals significant shifts in the characteristics of hailstorms under the future climate scenario. Notably, hail frequency trends vary across Europe, but the trends in hailstorm environments are comparatively uniform. The most striking results are as follows: (i) hail swath areas are projected to change in terms of both frequency and spatial extent, with a 2-fold increased frequency of storms producing similar to 50 mm and larger hail diameters. Per-storm hail swath areas generally expand by 15 %-30 %, with swath area increases being more important for smaller hail, while frequency changes dominate for larger hail. (ii) The effect of increased hail melting due to the higher elevation of the 0 degrees C level on the storm maximum hail diameters is found to be minor. (iii) Precipitation and wind hazards accompanying hailstorms are expected to increase on average by 20 % and 5 %, respectively, whereas extreme hail-precipitation compound events, i.e., hail with a diameter of at least 30 mm followed by 50 mm h-1 of rainfall, are projected to be twice as frequent in the future.