Very Large Hail in a Warming Climate: Climatology, Trends and Losses

Very Large Hail in a Warming Climate: Climatology, Trends and Losses | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Very Large Hail in a Warming Climate: Climatology, Trends and Losses Francesco Battaglioli, Mateusz Taszarek, Pieter Groenemeijer, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6196143/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 29 Dec, 2025 Read the published version in Nature Geoscience → Version 1 posted You are reading this latest preprint version Abstract We statistically constructed a long-term (1950–2023) global climatology of very large hail (VLH; \(\:\ge\:\) 5 cm) and investigated the associated trends in the context of a warming climate. Northern Argentina has the highest VLH occurrence, followed by Uruguay, Paraguay, and southern Brazil. The USA Great Plains and South Africa also exhibit high VLH frequency while Europe, Australia, and especially Asia experience lower activity. VLH occurrence increased the most across Europe, being strongly correlated with rising temperatures and primarily driven by an increase in low-level moisture and instability. Significant decreases in VLH frequency are limited to the Southern Hemisphere and occurred most strongly across South America as a result of decreasing mid-level humidity and instability. Hail-related losses have risen in the USA, Australia and Europe. The more frequent occurrence of VLH contributes to this increase in Europe, while in the USA and Australia, socio-economic factors are mainly responsible for the rise. Earth and environmental sciences/Natural hazards Earth and environmental sciences/Climate sciences/Climate change Figures Figure 1 Figure 2 Figure 3 Figure 4 Main Hail is a primary contributor to insurance losses from severe convective storms (SCSs), with the USA accounting for 50–80% of insured SCS losses (Gallagher Re 2024), resulting in annual damages between USD 8 and 14 billion (Podlaha et al. 2020). These losses are comparable to those caused by landfalling hurricanes (Gunturi and Tippett 2017 ). Single hail events exceeding USD 1 billion in damage have been recorded in the USA, Australia (Schuster et al. 2005 ), and Europe (Kunz et al. 2018 ), with northern Italy experiencing a record USD 6 billion loss in 2023 (Panosetti and Tomassetti 2024 ). Small hail (< 2 cm) can cause damage to agriculture (Rana et al. 2022 ), while damage to infrastructure, which contributes most to economic losses, increases sharply as hail becomes larger than 5 cm (Púčik et al. 2019 ). Despite the higher socioeconomic impacts of very large hail (VLH), research on global climatologies has focused on hail of smaller size (Prein and Holland 2018 , Bang and Cecil 2019 ). These studies were also constrained by the coarse resolution of satellite and reanalysis datasets (ERA-Interim; Dee et al. 2011) available at the time. The ERA5 reanalysis (Hersbach et al. 2020) offers higher spatial and temporal resolution, enhancing the representation of hail-favorable environmental conditions, although it does not simulate hail explicitly. To overcome this limitation, we apply a statistical model (AR-CHaMo; Rädler et al. 2018 , Battaglioli et al. 2023a ) to generate a global climatology of VLH based on atmospheric variables such as instability, vertical wind shear, and cloud base height. This model has proven effective in reconstructing the spatial distribution of VLH and the associated trends, even in regions lacking extensive ground-based observations (Battaglioli et al. 2023a ). Regional trend analyses have been conducted in the USA (Tang et al. 2019 , Gensini et al. 2024 ), Europe (Punge et al. 2017 , Rädler et al. 2018 ) , Australia (Raupach et al. 2023 ) and China (Zhang et al. 2008 ) however, a comprehensive global analysis remains absent. This work aims to address the existing knowledge gap by evaluating globally the long-term trends of VLH and the associated economic losses. The global climatology of very large hail The AR-CHaMo models were applied to the ERA5 reanalysis, assigning a probability of VLH to every grid point (0.25° × 0.25° spatial resolution) for each 3-hourly period of the past 74 years, from 1950 to 2023, In total, more than 80 billion atmospheric profiles were used to construct the global climatology of VLH (Fig. 1 ). The global maximum in VLH frequency is found in northern Argentina, where the mean annual number of VLH events is around 0.50, the equivalent of a VLH event every two years per grid box. This aligns with previous studies (Mezher et al. 2012 ), which report the highest frequency in the northern lee of the Argentina Andes. VLH frequency peaks from November to February although it has decreased in recent decades, particularly in December and January. Other areas of high VLH frequency include the tri-border region of Uruguay, Paraguay, and Brazil. Smaller but non-zero VLH frequency extends also into the Central Andes region of Bolivia and southern Peru. Across South America, the necessary ingredients for hailstorms frequently overlap. The Andes foothills provide a focus for lift (Rasmussen and Houze 2011), while the South American low-level jet advects warm and moist air from the Amazon. This, combined with high deep vertical wind shear, provides an ideal environment for hailstorms (Bechis et al. 2022). In the USA, the Great Plains exhibit a high VLH frequency with a maximum located between Kansas, Nebraska, and Colorado, aligning with radar-based climatologies (Murillo et al. 2021, Wendt and Jirak 2021 ). The overlap of moisture from the Gulf of Mexico (Molina et al. 2016 ), Elevated Mixed Layers generated over the high terrain of the southwest (Andrews et al. 2024 ), and cyclonic activity in spring and early summer (Li et al. 2021) provide a combination of high instability and wind shear which makes the region a hail hotspot. A pronounced east-west gradient is observed across northern Texas, where VLH frequency peaks between March and June. The VLH seasonal cycle in Dallas corresponds with the dates of major hail events: 5 April 2003 (Marshall et al. 2004 ), 24 May 2011 (Brown et al. 2015 ), and 11 June 2023 (Gallagher RE 2023 ). In Canada, the frequency is lower, with the most affected areas being southern Saskatchewan and Alberta, in line with previous studies (Brunet and Brimelow 2024 ). South Africa is another VLH hotspot, with a frequency exceeding 0.30 events per year in the northeast and along the Eastern Cape, where the Agulhas current provides moisture and the eastern Escarpment Mountains a focus for lift (Blamey et al. 2016). In Durban, over the last two decades, the peak month of occurrence has shifted from November to October. VLH is also present in equatorial Africa, the western Sahel, and northern Africa however a lack of ground-based reports and uncertainties in the accuracy of satellite-based estimates (Ferraro et al. 2015 ) complicate comparisons with observations in these regions. VLH is less frequent in Europe, with the occurrence maximized in the lee of mountain ranges. High VLH frequency is found across northeastern Spain, southwestern France, and northern Italy which stands out as the European hotspot, in agreement with previous studies (Punge et al. 2017 ). The seasonal cycle in Milan peaks in summer, although there is a tendency towards an earlier onset of VLH occurrence in spring. Other local maxima are observed in Switzerland, southern Germany, and southeastern Austria, with VLH occurrence extending into southeast Europe. Further east, VLH frequency peaks locally in Turkey, Iran, and the Arabian Peninsula, though climatological studies to confirm these patterns are lacking. Asia experiences the lowest VLH activity, with northern Pakistan and Bangladesh having the highest frequency at around 0.12 events per year. VLH also peaks locally in northeastern China during spring and summer, while lower but non-zero VLH frequency is found over the Tibetan Plateau and southern China, consistent with ground-based report climatologies (Ni et al. 2020 ). In Australia, VLH is most frequent in the southeast, as shown by Brook et al. ( 2024 ). A second maximum is present in the southwest interior, though reports in this area are rare due to low population density, and the limited radar coverage prevents verification of these findings. The global climate trends of very large hail Trends in VLH frequency (1950–2023) are not globally homogenous but exhibit a strong zonal dependence: positive and significant trends are almost exclusively present in the Northern Hemisphere, while negative trends are limited to the Southern Hemisphere (Fig. 2 ). Europe is the only continent to exhibit a widespread and statistically significant increase in VLH frequency, extending into northern Africa and the Middle East. These findings align with pan-European (Rädler et al. 2018 , Púčik et al. 2019 , Taszarek et al. 2021a ) and regional analyses from single countries (e.g., Switzerland; Wilhelm et al. 2024 ). Northern Italy shows the largest global increase in VLH frequency (up to + 0.03 events per decade). In Milan, VLH frequency has consistently exceeded the 1950–2023 average over the last two decades, with 2021–2023 marking the most active years. This period corresponds with a record number of hail reports in Italy, including a 19-cm hailstone in July 2023 (ESWD 2024). Positive and significant trends are not limited to Italy but extend across most of mainland Europe. North America displays positive VLH trends, particularly in southern Canada, where southern Alberta has the largest increase (+ 0.01 events per decade). An increase of similar magnitude is also seen in the mountainous regions of Mexico. Trends are mostly not significant in the USA except for very localized increases in the Great Plains. Compared to previous work focusing on hail-prone environments (Tang et al. 2019 ), the increase is more localized, although a tendency for very large hailstones to become more frequent has been shown in recent studies (Gensini et al. 2024 ). In Dallas, VLH frequency has increased by 43.4% since the 1950s, and 2023 was the most active year, with a + 105% departure from the 1950–2023 average. This aligns with the record losses reported across Texas in 2023 (Munich RE 2024 ). In northern Argentina, VLH frequency has decreased the most. The decrease is steady in Mendoza, where every year since 2006, except for 2015 and 2016, has experienced below-average VLH frequency (mean = -29.8%). This decrease is supported by a reduction in hail reports (Beal et al. 2020 ). Localized positive trends are found in Bolivia and Peru. South Africa has the second-largest VLH decrease, especially in the northeast, where the decadal trend reaches − 0.02 events per decade. This trend is also supported by a general reduction in severe thunderstorm environments (Taszarek et al. 2021b ). Negative trends are observed locally in Equatorial Africa, although they remain unverified due to a lack of direct observations. In contrast, northern Africa shows a positive VLH frequency trend comparable in magnitude to that of mainland Europe, as noted by Rädler et al. ( 2018 ). In Asia, VLH frequency trends are generally small or non-existent. However, positive and significant trends are present across Pakistan, the Tibetan Plateau, and northeastern China. In Beijing, above-average VLH activity occurred in the 1980s, followed by a decreasing trend in hail days from 1980 to 2005 (Zhang et al. 2008 ). Negative and significant trends in the region are limited to the eastern coast of India. Finally, in Australia, trends are largely negative in continental areas of the north while coastal regions show constant VLH frequency. In Brisbane, a strong year-to-year variability is present, and trends are not significant, as also shown by Raupach et al. ( 2023 ). Hail trends attribution and relationship with global warming In Fig. 3 , we display the grid-based correlation between the annual average VLH frequency and the 2-meter temperature yearly mean (1950–2023). The relationship between temperature change and VLH trends is not uniform globally. Europe is the only region where an increase in 2-meter temperature significantly correlates with higher VLH frequency, especially in southern Europe ( r ≈ 0.7), the Mediterranean, Turkey, and the Middle East. This positive correlation reflects an increase in instability and thunderstorm intensity (Taszarek et al. 2021b , Battaglioli et al. 2023a ). In the USA, the correlation is weaker and spatially heterogeneous, with positive values in the east and negative ones in the west, where drying in the mid-levels reduces storm potential (Taszarek et al. 2021b , Andrews et al. 2024 ). In South America, a bimodal pattern appears along the lee of the Andes, with positive correlations ( r > 0.5) in Peru and Bolivia but negative correlations in Argentina. Similar patterns occur in southern and central Brazil however correlations remain small (-0.3 < r < + 0.3), suggesting other factors might better explain VLH frequency trends. In Africa, correlations are generally small and non-significant, except in the Sahel region, where r < -0.5. This area coincides with a marked decrease in convective precipitation (Taszarek et al. 2021b ) Correlations vary by region but remain small in Australia, ranging from − 0.3 inland to + 0.3 on the coast. The correlation mirrors moisture patterns (Supplementary Fig. 1), with increasing coastal moisture and inland drying, which reduces storm potential (Raupach et al. 2023 ). In conclusion, although temperature as a stand-alone parameter does not represent an ingredient for storm or hail formation, changes in its values can correlate with changes in storm ingredients, influencing VLH frequency. Northern Italy and Northwestern Argentina: Quantitative Trend Attribution Observing that upward 2-meter temperature trends coincide with both increasing and decreasing trends in VLH frequency prompts us to investigate the drivers of these changes. We focus on two regions with the most pronounced and opposite trends, northern Italy and northwestern Argentina, and analyze the contributions of the various predictor parameters of the VLH model (Supplementary Table 1) to the trends depicted in Fig. 2 . In northern Italy, the sum of all predictors’ contributions is + 73%. While thunderstorms have become slightly less frequent (-6%), their severity has substantially increased. More specifically, the probability that a storm will be capable of producing VLH has risen by 79%. This increase is mainly due to rising atmospheric instability in the cold portion of the storm, which results from enhanced latent heat release and higher water vapor content. This aligns with previous findings (Rädler et al. 2018 , Pilguj et al. 2022 ) suggesting that increased instability leads to stronger storms in the mid-latitudes. In northwestern Argentina, the sum of all predictors’ contributions is -25%. Thunderstorms have both become less frequent (-6%) and less prone to producing VLH (-16%). The reduction in thunderstorm likelihood is consistent with patterns of increasing drought (Feron et al. 2024) and is primarily attributed to declining mid-level atmospheric relative humidity and instability. A detailed breakdown of parameter contributions for both northern Italy and northwestern Argentina is provided in Supplementary Table 2. Trends in insured losses across Europe, the United States, and Australia Understanding the relationship between changes in VLH frequency and hail losses is crucial, given the significant economic impacts of hail (Púčik et al. 2019 ). We analyzed trends in hail loss events in regions with high insurance penetration: Germany, Austria, Benelux countries (Europe), the eastern two-thirds of the USA, and populated coastal regions of Australia. To compare long-term trends in hail loss events with VLH frequencies, we aggregated VLH probabilities over the selected domains from 1993 to 2023. Figure 4 shows the comparison between the number of hail loss events (see Methods) and the accumulated VLH probabilities in the three regions. Over the past thirty years, hail loss events have increased in Europe (Germany, Austria, Benelux), the USA, and Australia, driven by a combination of social, economic, and environmental factors (Kron et al. 2012 ). The modeled VLH frequency generally reproduces yearly variations in annual loss events well, especially in Europe between 1998 and 2007. The correlation between the trends in VLH frequency and those in hail loss events suggests that increasing VLH occurrence strongly contributes to the rise in hail loss events in Europe. In contrast, the USA and Australia display different patterns. While VLH frequency increased only slightly in the USA and decreased in Australia, hail loss events continued to rise in both regions. This discrepancy suggests that socio-economic factors, such as changes in exposure and vulnerability, are more important drivers of the increase in hail loss events than meteorological factors. These findings are consistent with previous research on tornadoes in the USA (Strader et al. 2017 , Bouwer 2019 ). Summary and Discussion We reconstructed the global occurrence of VLH over a period of 74 years (1950–2023). using an additive logistic regression model (AR-CHaMo) trained with lightning observations, hail reports, and atmospheric predictors from the ERA5 reanalysis across Europe, the USA, and Australia. VLH is most frequent across northern Argentina and the border regions of Uruguay, Paraguay, and Brazil. High VLH frequency is also modeled in the USA Great Plains and parts of South Africa. While VLH is less frequent in other regions (Europe, Africa, Oceania, and Asia), the reconstruction aligns with regional climatologies in both the simulated spatial distribution and the seasonal patterns. Additionally, the recent extreme hail activity in the USA and Europe is well represented by the model, with 2023 standing out as the year with the highest VLH frequency. Although no universally valid relationship between VLH trends and global warming has been identified, some significant regional correlations exist. In Europe, VLH frequency has increased the most, correlating with rising temperatures. Trends are especially pronounced in northern Italy, driven by an increase in low-level moisture and instability in the cold part of storms. Other regions with positive significant trends include the Middle East, southern Canada, parts of Mexico, and localized areas in the USA. Increasing temperatures do not always correlate with an increase in VLH frequency: in the western USA and continental Australia, drying has reduced the storm potential. Negative trends in VLH frequency are observed mainly in the Southern Hemisphere, with the strongest declines in South Africa and northern Argentina, where they are primarily driven by a decrease in mid-level humidity and instability. Finally, we examined the relationship between changes in VLH frequency and hail losses. Hail loss events have increased in parts of Europe, the USA, and Australia. In Europe, rising VLH occurrence contributes to the increase in hail losses, alongside growing exposure and vulnerability. In the USA and Australia, changes in meteorological conditions do not correlate with the increase in hail loss events; instead, exposure and vulnerability drive the increase. To confirm this hypothesis, future studies should dive deeper into the topic of loss normalization or investigate trends in exposure and vulnerability. Some limitations must be considered when interpreting these results. The VLH model, trained on three mid-latitude regions, was applied globally therefore, results should be interpreted with caution in areas where regional climatologies do not allow for verification or known biases in ERA5 are present (e.g., the tropics; Taszarek et al. 2021c ). Future research should incorporate newly available databases from regions like South America, Canada, and China to better capture hail environments on a global scale. Another area of future research concerns the use of reanalysis data at a higher resolution to explicitly simulate storm development, which was not possible in ERA5. Despite the limitations, this is the first analysis of the climatological occurrence of VLH, the associated long-term trends in the context of a warming climate, and the relationship with losses on a global scale. Data & Methods AR-CHaMo models The Additive Regressive Convective Hazard Model (AR-CHaMo; Rädler et al. 2018 , Battaglioli et al. 2023a ) is a logistic regression model that yields the probability of VLH as a function of environmental predictors from the atmospheric reanalysis (listed in Supplementary Table 1) that were calculated using the ThundeR rawinsonde package (Taszarek et al. 2023). AR-CHaMo calculates this probability by combining the likelihood of a thunderstorm ( \(\:{P}_{lightning}\) ) and the conditional probability of VLH given a thunderstorm ( \(\:{P}_{VLH|lightning}\) ). \(\:{P}_{lightning}\) was trained simultaneously using lightning observations (over a billion) from the Arrival Time Difference network (ATDnet; Enno et al. 2020 ) across Europe (34.5°–63.5°N, 9.0°W–46.0°E), the National Lightning Detection Data (NLDN; Koehler 2020 ) from the USA (29.0°–41.5°N, 109.0°–79.0°W) and the Global Position and Tracking Systems (GPATS; Dowdy and Kuleshov 2014 ) from Australia (10.5°– 40.0°S, 119.0°–153.5°E). \(\:{P}_{VLH|lightning}\) was trained simultaneously using hail reports (over 4500) from the European Severe Weather Database (ESWD; 45.0°–54.0°N, 5.0°–22.0°E), the Storm Prediction Centre (SPC) Storm Archive (30.5°–41.5°N, 105.0°–82.0°W) and the Australian Bureau of Meteorology (BOM; 22.5°– 37.5°S, 149.5°–153.5°E). \(\:{P}_{lightning}\) was improved by using a grid-based ratio between lightning observations from the Earth Networks Global Lightning Detection Network (ENGLN; DiGangi et al. 2022 ) and modeled lightning occurrence, improving the lightning distribution, especially across (sub-)tropical regions where AR-CHaMo was not trained. The mean annual expected number of VLH events was calculated as the sum of all three-hourly probabilities in a year multiplied by three. To do so, we assumed three-hourly probabilities to be independent of each other, although this may not strictly be the case. Insurance loss data Data on hail losses were obtained from the NatCatSERVICE database of Munich RE (Kron et al. 2012 ). Within the database, each loss entry includes its type, location, date, loss, and description. Using the description, we filtered only for events where the loss was at least partly caused by hail, ensuring that events caused exclusively by other convective hazards were not taken into consideration. Single events often covered multiple days, meaning that those entries included multiple severe convective storms. In these cases, loss location coordinates indicate the loss center. To account for inflation, urban growth, and regional wealth differences, the estimated overall losses were normalized to the 2019 levels of destructible wealth. For 2020 to 2023, we used the inflation-adjusted values of the losses. Normalization was done by considering the Gross Domestic Product discretized on a 1° × 1° grid. The accuracy of the loss estimates depends on the insurance penetration in a given country for a specific hazard (Kron et al. 2012 ). In Europe, we selected Germany, the Benelux countries, and Austria. Australia was also considered, given the fact that hail causes substantial losses (Schuster et al. 2005 ), similar to the USA (Munich RE 2023). For the USA, we applied a threshold of USD 1 billion to capture the most impactful events and to limit the chances that “minor” events were missed in the early times. Declarations Acknowledgments F.B., P.G., and T.P. contributions were supported by ESSL. M.T. was funded by a grant from the Polish National Science Centre (grant no. 2020/39/D/ST10/00768). A.R. contribution was supported by Munich RE. The ERA5 reanalysis computations were performed in the Poznań Supercomputing and Networking Center (grant no. 648). Author information Author Affiliations European Severe Storms Laboratory e.V., Wessling, Germany Francesco Battaglioli, Pieter Groenemeijer European Severe Storms Laboratory - Science & Training, Wiener Neustadt, Austria Pieter Groenemeijer, Tomas Púčik Department of Meteorology and Climatology, Adam Mickiewicz University, Poznan, Poland Mateusz Taszarek Munich Re, Munich, Germany Anja Rädler Contributions F.B. conceived the idea and designed this study. F.B. performed the analysis, wrote the manuscript, and produced all figures apart from Fig. 4 (A.R). M.T. calculated environmental predictors from the ERA5 reanalysis and contributed to the interpretation of the results. P.G. and T.P. contributed to the scientific analysis of the results together with A.R., who contributed especially to the section “Trends in insured losses across Europe, the United States, and Australia”. Corresponding author Correspondence to Francesco Battaglioli ( [email protected] ). Ethics declaration Competing interests The authors declare no competing interests. Data availability ERA5 hourly data on single levels can be obtained from the Copernicus Climate Data Store at ​​ https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview . Convective parameters from ERA5 data are calculated using the open-source ThundeR package that is available at https://bczernecki.github.io/thundeR/ . USA hail reports are available from the Storm Prediction Centre Storm Archive ( https://www.spc.noaa.gov/exper/archive/events/ ), Australian hail reports are available from the Bureau of Meteorology ( http://www.bom.gov.au/australia/stormarchive/ ) and European ones can be obtained from the European Severe Weather Database ( https://eswd.eu/ ). ENGLN global lightning observations are publicly available at http://thunderhours.earthnetworks.com while regional ones were obtained for this study and are property of the Met Office (ATDnet), Vaisala (NLDN), and the Bureau of Meteorology (GPATS). The NatCatSERVICE dataset is property of Munich RE and was made available exclusively for this study. The AR-CHaMo datasets can be made available upon request to the corresponding author, Francesco Battaglioli ( [email protected] ). Code availability Predictor parameters were calculated using the ThundeR rawinsonde package, which is openly accessible at http://www.rawinsonde.com/thunder_app/ . Code for the model calculation can be made available upon request to the corresponding author, Francesco Battaglioli ( [email protected] ). References Andrews, M. S., V. A. Gensini, A. M. Haberlie, W. S. Ashley, A. C. Michaelis, & Taszarek M. Climatology of the Elevated Mixed Layer over the Contiguous United States and Northern Mexico Using ERA5: 1979–2021. J. Climate, 37, 1833–1851 (2024) https://doi.org/10.1175/JCLI-D-23-0517.1 . Bang, S., & Cecil D. Constructing a Multifrequency Passive Microwave Hail Retrieval and Climatology in the GPM Domain. J. Appl. Meteor. Climatol. 58, 1889–1904 (2019) doi: 10.1175/jamc-d-19-0042.1 . Battaglioli, F., P. Groenemeijer, T. Púčik, M. Taszarek, U. Ulbrich, & Rust H. Modeled Multidecadal Trends of Lightning and (Very) Large Hail in Europe and North America (1950–2021). J. Appl. Meteorol. Clim. 62(11), 1627–1653 (2023a) https://doi.org/10.1175/JAMC-D-22-0195.1 . Beal, A., R. Hallak, L. D. Martins, J. A. Martins, G. Biz, A. P. Rudke, & Tarley C. R. T. Climatology of hail in the triple border paraná, Santa Catarina (Brazil) and Argentina. Atmos. Res. 234, 104747 (2020) doi: 10.1016/j.atmosres.2019.104747 . Bechis, H., & Coauthors. A case study of a severe hailstorm in Mendoza, Argentina, during the RELAMPAGO-CACTI field campaign. Atmos Res. 271, 106127 (2022) https://doi.org/10.1016/j.atmosres.2022.106127 . Blamey, R. C., C. Middleton, C. Lennard, & Reason C. J. C. A climatology of potential severe convective environments across South Africa. Clim. Dyn. 49, 2161–2178 (2017), doi: 10.1007/s00382-016-3434-7 . Bouwer, L. M. Observed and Projected Impacts from Extreme Weather Events: Implications for Loss and Damage. Loss and Damage from Climate Change. Climate Risk Management, Policy and Governance, 63–82 (2019) https://doi.org/10.1007/978-3-319-72026-5_3 . Brunet, D. & Brimelow J. A Hail Climatology for Canada Using a Lightning Proxy. J. Appl. Meteor. Climatol. 63, 1227–1240 (2024) https://doi.org/10.1175/JAMC-D-23-0214.1 . Brook, J. P., J. S. Soderholm, A. Protat, H. McGowan, & Warren R. A. A radar- based hail climatology of Australia. Mon. Wea. Rev. 152(2), 607–628 (2024) doi: 10.1175/mwr-d-23- 0130.1 . Brown, T. M., W. H. Pogorzelski, & Giammanco I. M. Evaluating hail damage using property insurance claims data. Weather, Climate, and Society. 7(3), 197–210 (2015) doi: 10.1175/wcas-d-15-0011.1 DiGangi, E. A., M. Stock, & Lapierre J. Thunder Hours: How Old Methods Offer New Insights into Thunderstorm Climatology. Bull. Amer. Meteor. Soc. 103, E548–E569 (2022) https://doi.org/10.1175/BAMS-D-20-0198.1 . Dowdy, A. J., & Kuleshov Y. Climatology of lightning activity in Australia: spatial and seasonal variability. Australian Meteorological and Oceanographic J. 64(2), 103–108 (2014) https://doi.org/10.1071/ES14010 . Enno, S., J. Sugier, R. Alber, & Seltzer M. Lightning flash density in Europe based on 10 years of ATDnet data. Atmos. Res. 235, 104769 (2020) https://doi.org/10.1016/j.atmosres.2019.104769 . ESWD. European Severe Weather Database, accessed 6 April 2024, https://eswd.eu . Feron, S. & Coauthors. South America is becoming warmer, drier, and more flammable. Commun. Earth Environ. 5, 501 (2024) https://doi.org/10.1038/s43247-024-01654-7 . Ferraro, R., J. Beauchamp, D. Cecil, & Heymsfield G. A prototype hail detection algorithm and hail climatology developed with the Advanced Microwave Sounding Unit (AMSU). Atmos. Res. 163, 24–35 (2015) doi: 10.1016/j.atmosres.2014.08.010 . Gallagher RE. 2023: Natural Catastrophe and Climate Report: 2023, Available online: https://www.ajg.com/gallagherre/-/media/files/gallagher/gallagherre/news-and-insights/2024/january/natural-catastrophe-and-climate-report-2023.pdf (accessed on 3 Dec 2024). Gensini, V. A., W. Ashley, A. C. Michaelis, A. M. Haberlie, J. Goodin, & Wallace B. C. Hail size dichotomy in a warming climate. npj Climate and Atmos. Sci. 7, 185 (2024) https://doi.org/10.1038/s41612-024-00728-9 Gunturi, P., & Tippett, M. K. Managing severe thunderstorm risk: Impact of ENSO on U.S. tornado and hail frequencies (Tech. Rep.). Minneapolis, MN, USA: Willis Re (2017). Hersbach, H., & Coauthors. The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc. 146, 1999–2049 (2020) https://doi.org/10.1002/QJ.3802 . Li, F., D. R. Chavas, K. A. Need, N. Rosenbloom, & Dawson D. T. The role of Elevated Terrain and the Gulf of Mexico in the Production of Severe Local Storm Environments over North America. J Clim. 34(19), 7799–7819 https://doi.org/10.1175/JCLI-D-20-0607.1 Koehler, T. Cloud-to-Ground Lightning Flash Density and Thunderstorm Day Distributions over the Contiguous United States Derived from NLDN Measurements: 1993– 2018. Mon. Wea. Rev. 148(1), 313–332 (2020) https://doi.org/10.1175/MWR-D-19-0211.1 . Kron, W., M. Steuer, P. Löw, & Wirtz A. How to deal properly with a natural catastrophe database – analysis of flood losses Nat. Hazards Earth Syst. Sci. 12, 535–550 (2012) https://doi.org/10.5194/nhess-12-535-2012 . Kunz, M., U. Blahak, J. Handwerker, M. Schmidberger, H. J. Punge, S. Mohr, E. Fluck, & Bedka, K. M. The severe hailstorm in southwest Germany on 28 July 2013: Characteristics, impacts and meteorological conditions. Quart. J. Roy. Meteor. Soc. 144, 231–250 (2018) doi: 10.1002/qj.3197 . Marshall, T. P., R. F. Herzog, D. E. Mitchell, S. J. Morrison, and S. Rae, 2004: The Dallas- Ft. Worth, TX hailstorm: 5 April 2003. 22nd Conf. on Severe Local Storms , P9.1. doi: https://ams.confex.com/ams/11aram22sls/techprogram/paper_81089.htm . Molina, M. J., R. P. Timmer, & Allen, J. T. Importance of the Gulf of Mexico as climate driver for U.S. severe thunderstorm activity. Geophys. Res. Lett. 43(23), 12,295 – 12,304 (2016) https://doi.org/10.1002/2016GL071603 . Murillo, E., C. Homeyer, & Allen, J. A 23-Year Severe Hail Climatology Using GridRad MESH Observations. Mon. Wea. Rev. 149, 945–958 (2023) doi: 10.1175/mwr-d-20-0178.1 . Munich RE 2024: Natural disasters of 2023: Available online: https://www.munichre.com/en/company/media-relations/media-information-and-corporate-news/media-information/2024/natural-disaster-figures-2023.html (accessed on 24. Feb 2025) Mezher, R. N., M. Doyle, & Barros V. Climatology of hail in Argentina. Atmos. Res. 114–115, 70–82 (2012) doi: 10.1016/j.atmosres.2012.05.020 . Ni, X., A. Muehlbauer, J. T. Allen, Q. Zhang, & Fan, J. A Climatology and Extreme Value Analysis of Large Hail in China. Mon. Wea. Rev. 148, 1431–1447 (2020) https://doi.org/10.1175/MWR-D-19-0276.1 . Panosetti, D., & Tomassetti, U. The July 2023 Northern Italy hailstorms from a climatological and (re)insurance market perspective. EGU General Assembly 2024, Vienna, Austria. Pilguj, N., M. Taszarek, M. Kryza, & Brooks H. Reconstruction of Violent Tornado Environments in Europe: High-Resolution Dynamical Downscaling of ERA5. Geophys. Res. Lett. 49(11), (2022) https://doi.org/10.1029/2022GL098242 . Podlaha, A., S. Bowen, & M. Lörinc. Weather, climate and catastrophe insight: 2019 annual report. Aon Annual Rep., 83 pp., (2020) http://thoughtleadership.aon.com/Documents/20200122-if-natcat2020.pdf . Prein, A. & Holland G. Global estimates of damaging hail hazard. Wea. and Climate Extremes. 22, 10–23 (2018) https://doi.org/10.1016/j.wace.2018.10.004 . Púčik, T., C. Castellano, P. Groenemeijer, T. Kühne, A. Rädler, B. Antonescu, & Faust E. Large Hail Incidence and Its Economic and Societal Impacts across Europe. Mon. Wea. Rev. 147, 3901–3916 (2019) doi: 10.1175/mwr-d-19-0204.1 . Punge, H., K. Bedka, M. Kunz, & Reinbold A. Hail frequency estimation across Europe based on a combination of overshooting top detections and the ERA-INTERIM reanalysis. Atmos. Res. 198, 34–43 (2017) https://doi.org/10.1016/j.atmosres.2017.07.025 . Rana, V. S., Sharma, S., Rana, N., Sharma, U., Patiyal, V., Banita, & Prasad H. Management of Hailstorms under a Changing Climate in Agriculture: A Review. Environ. Chem. Lett. 20, 3971–3991 (2022) https://doi.org/10.1007/s10311-022-01502-0 . Rädler, A., P. Groenemeijer, E. Faust, & Sausen R. Detecting Severe Weather Trends Using an Additive Regressive Convective Hazard Model (AR-CHaMo). J. Appl. Meteor. Climatol. 57(3), 569–587 (2018) https://doi.org/10.1175/JAMC-D-17-0132.1 . Raupach, T. H., J. S. Soderholm, R. A. Warren, & Sherwood S. C. Changes in hail hazard across Australia: 1979–2021. npj Climate and Atmos. Sci. 6(1) (2023) doi: 10.1038/s41612-023-00454-8 . Rasmusen, K. L. & Houze R. A. Orogenic Convection in Subtropical South America as seen by the TRMM Satellite. Mon Wea. Rev. 139(8), 2399–2420 (2011) https://doi.org/10.1175/MWR-D-10-05006.1 . Schuster, S., R. Blong, R. Leigh, & McAneney K. Characteristics of the 14 April 1999 Sydney hailstorm based on ground observations, weather radar, insurance data and emergency calls. Nat. Hazards Earth Syst. Sci. 5(5), 613–620 (2005) https://doi.org/10.5194/nhess- 5-613-2005 . Strader, S. M., W. S. Ashley, T. J. Pingel, & Krmenec A. J. Projected 21st century changes in tornado exposure, risk, and disaster potential. Climatic Change. 141, 301–313 (2017) https://doi.org/10.1007/s10584-017-1905-4 . Tang, B., V. Gensini, & Homeyer C. Trends in United States large hail environments and observations. npj Climate and Atmos. Sci. 2(1), 45 (2019) https://doi.org/10.1038/s41612-019-0103-7 . Taszarek, M., J. Allen, H. Brooks, N. Pilguj, and Czernecki B. Differing Trends in United States and European Severe Thunderstorm Environments in a Warming Climate. Bull. Amer. Meteor. Soc. 102, 296–322 (2021a) doi: 10.1175/bams-d-20-0004.1 . Taszarek, M., J. T. Allen, M. Marchio, & Brooks H. E. Global climatology and trends in convective environments from ERA5 and Rawinsonde Data. npj Climate and Atmos. Sci. 4(1). (2021b) doi: 10.1038/s41612-021-00190-x . Taszarek, M., N. Pilguj, J. Allen, V. Gensini, H. Brooks, & Szuster P. Comparison of convective parameters derived from ERA5 and MERRA2 with rawinsonde data over Europe and North America. J. Climate. 34(8), 1–55 (2021c) doi: 10.1175/jcli-d-20-0484.1 . Taszarek, M., B. Czernecki, & Szuster, P. thundeR - a rawinsonde package for processing convective parameters and visualizing atmospheric profiles, 11th European Conference on Severe Storms, Bucharest, Romania, 8–12 May 2023, ECSS2023-28, https://doi.org/10.5194/ecss2023-28 . Wendt, N., & Jirak I. An Hourly Climatology of Operational MRMS MESH- Diagnosed Severe and Significant Hail with Comparisons to Storm Data Hail Reports. Wea. Forecasting 36, 645–659 (2021) doi: 10.1175/waf-d-20-0158.1 . Wilhelm, L., C. Schwierz, K. Schröer, M. Taszarek, & Martius O. Reconstructing hail days in Switzerland with statistical models (1959–2022). Nat. Hazards Earth Syst. Sci. 24, 3869–3894 (2024) https://doi.org/10.5194/nhess-24-3869-2024 . Xie, B., Zhang, Q., & Wang, Y. Trends in hail in China during 1960–2005. Geophys. Res. Lett., 35, L13801 (2008) https://doi.org/10.1029/2008GL034067 Zhang, C., Q. Zhang, & Wang, Y. Climatology of hail in China: 1961–2005. J. Appl. Meteor. Climatol. 47(3), 795–804 (2008) doi: 10.1175/2007jamc1603.1 . Additional Declarations There is NO Competing Interest. Supplementary Files 20250310BattaglioliGlobalhailpapersupplementary.docx Cite Share Download PDF Status: Published Journal Publication published 29 Dec, 2025 Read the published version in Nature Geoscience → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6196143","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":449749460,"identity":"33209c5f-2535-489e-9376-42a72aa97c39","order_by":0,"name":"Francesco Battaglioli","email":"data:image/png;base64,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","orcid":"","institution":"European Severe Storms Laboratory e.V.","correspondingAuthor":true,"prefix":"","firstName":"Francesco","middleName":"","lastName":"Battaglioli","suffix":""},{"id":449749461,"identity":"5fda48dc-c03f-42db-a231-56d2d2e9cb39","order_by":1,"name":"Mateusz Taszarek","email":"","orcid":"https://orcid.org/0000-0001-9578-5872","institution":"Adam Mickiewicz University","correspondingAuthor":false,"prefix":"","firstName":"Mateusz","middleName":"","lastName":"Taszarek","suffix":""},{"id":449749462,"identity":"2e4ce5b9-5c7c-4293-be77-35563ce814c2","order_by":2,"name":"Pieter Groenemeijer","email":"","orcid":"","institution":"European Severe Storms Laboratory - Science \u0026 Training","correspondingAuthor":false,"prefix":"","firstName":"Pieter","middleName":"","lastName":"Groenemeijer","suffix":""},{"id":449749463,"identity":"014c6ec1-ac9b-4251-91d8-025c51a1d737","order_by":3,"name":"Tomas Pucik","email":"","orcid":"","institution":"European Severe Storms Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Tomas","middleName":"","lastName":"Pucik","suffix":""},{"id":449749464,"identity":"a21343e7-edb6-4387-9a3d-cd466c2a895e","order_by":4,"name":"Anja Rädler","email":"","orcid":"","institution":"Munich RE","correspondingAuthor":false,"prefix":"","firstName":"Anja","middleName":"","lastName":"Rädler","suffix":""}],"badges":[],"createdAt":"2025-03-10 13:52:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6196143/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6196143/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41561-025-01868-0","type":"published","date":"2025-12-29T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81676120,"identity":"9fae2bbc-3c11-4ecc-b162-b61d5f49bac6","added_by":"auto","created_at":"2025-04-30 07:36:12","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":516308,"visible":true,"origin":"","legend":"\u003cp\u003eMean annual number of VLH events between 1950 and 2023. The seasonal cycles for two 24-year periods (1950–1974, 1975–1999) and one 23-year period (2000–2023) are shown for six locations: Dallas (United States), Milan (Italy), Beijing (China), Mendoza (Argentina), Johannesburg (South Africa) and Brisbane (Australia). Maritime locations further than 100 km from the coast were excluded from the analysis since the impact of VLH events is much higher over land than in the oceans.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-6196143/v1/ad02ad33ec5d0f36092aeda6.png"},{"id":81675372,"identity":"8901d053-d94a-48df-8ae7-c9a49e32facb","added_by":"auto","created_at":"2025-04-30 07:20:12","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":463838,"visible":true,"origin":"","legend":"\u003cp\u003eMean trend in annual\u003cstrong\u003e \u003c/strong\u003enumber of VLH events per decade between 1950 and 2023. Trends significant at a 95 percent level are hatched. Hail stripes, as per Battaglioli et al. (2023a), displaying the yearly percentage departure from the long-term average, are plotted for the same locations as in Fig. 1.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-6196143/v1/48560b96c4924462b0cbea3d.png"},{"id":81675941,"identity":"7edeadfe-667f-4627-b327-d6a6be2f26d2","added_by":"auto","created_at":"2025-04-30 07:28:12","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":480970,"visible":true,"origin":"","legend":"\u003cp\u003eGrid-based Pearson correlation (\u003cem\u003er\u003c/em\u003e) between yearly VLH frequency and 2-meter temperature yearly mean for the period 1950–2023. Areas with a correlation significant at a 95 percent level (\u003cem\u003ep-value\u003c/em\u003e \u0026lt; 0.05) are hatched. To display only VLH-prone regions, values over areas with a low mean annual VLH frequency (\u0026lt; 0.06 events per year) are not shown.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-6196143/v1/35d6e74f82079d9236cc2b4c.png"},{"id":81675942,"identity":"11e7ad87-9521-441e-a8b8-2008893621df","added_by":"auto","created_at":"2025-04-30 07:28:12","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":770662,"visible":true,"origin":"","legend":"\u003cp\u003eTime series of the standardized mean annual sum of VLH probabilities (blue line) and hail loss events (green line) for Germany, Austria, and the Benelux countries (Europe), the eastern two-thirds of the United States, and coastal Australia. To consider a loss event in the United States, the normalized losses had to exceed USD 1 billion. Standardization was performed by comparing each yearly value to the time series average and adjusting it by the standard deviation. Linear trends are shown as dashed lines. The green dotted line in the USA and Australia graphs corresponds to zero hail loss events. For Europe, the line is below -2.5 as the smallest number of loss events was 3, e.g., in 1994.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-6196143/v1/7214d8b18a14e6f077571cec.png"},{"id":99211164,"identity":"ea6e5008-56fa-4b9d-8af8-68c7757f8e94","added_by":"auto","created_at":"2025-12-30 08:05:18","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2777796,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6196143/v1/44109e70-9443-4350-8719-75be6774de38.pdf"},{"id":81675377,"identity":"8aace25c-0eca-4060-9136-477aec4b9b4f","added_by":"auto","created_at":"2025-04-30 07:20:12","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":4309967,"visible":true,"origin":"","legend":"","description":"","filename":"20250310BattaglioliGlobalhailpapersupplementary.docx","url":"https://assets-eu.researchsquare.com/files/rs-6196143/v1/efbf4f49037d21cbf489c049.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Very Large Hail in a Warming Climate: Climatology, Trends and Losses","fulltext":[{"header":"Main","content":"\u003cp\u003eHail is a primary contributor to insurance losses from severe convective storms (SCSs), with the USA accounting for 50\u0026ndash;80% of insured SCS losses (Gallagher Re 2024), resulting in annual damages between USD 8 and 14\u0026nbsp;billion (Podlaha et al. 2020). These losses are comparable to those caused by landfalling hurricanes (Gunturi and Tippett \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Single hail events exceeding USD 1\u0026nbsp;billion in damage have been recorded in the USA, Australia (Schuster et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), and Europe (Kunz et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), with northern Italy experiencing a record USD 6\u0026nbsp;billion loss in 2023 (Panosetti and Tomassetti \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Small hail (\u0026lt;\u0026thinsp;2 cm) can cause damage to agriculture (Rana et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), while damage to infrastructure, which contributes most to economic losses, increases sharply as hail becomes larger than 5 cm (P\u0026uacute;čik et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Despite the higher socioeconomic impacts of very large hail (VLH), research on global climatologies has focused on hail of smaller size (Prein and Holland \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, Bang and Cecil \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). These studies were also constrained by the coarse resolution of satellite and reanalysis datasets (ERA-Interim; Dee et al. 2011) available at the time. The ERA5 reanalysis (Hersbach et al. 2020) offers higher spatial and temporal resolution, enhancing the representation of hail-favorable environmental conditions, although it does not simulate hail explicitly. To overcome this limitation, we apply a statistical model (AR-CHaMo; R\u0026auml;dler et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, Battaglioli et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023a\u003c/span\u003e) to generate a global climatology of VLH based on atmospheric variables such as instability, vertical wind shear, and cloud base height. This model has proven effective in reconstructing the spatial distribution of VLH and the associated trends, even in regions lacking extensive ground-based observations (Battaglioli et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023a\u003c/span\u003e). Regional trend analyses have been conducted in the USA (Tang et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Gensini et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), Europe (Punge et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, R\u0026auml;dler et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e)\u003csup\u003e,\u003c/sup\u003e Australia (Raupach et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and China (Zhang et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) however, a comprehensive global analysis remains absent. This work aims to address the existing knowledge gap by evaluating globally the long-term trends of VLH and the associated economic losses.\u003c/p\u003e\n\u003ch3\u003eThe global climatology of very large hail\u003c/h3\u003e\n\u003cp\u003eThe AR-CHaMo models were applied to the ERA5 reanalysis, assigning a probability of VLH to every grid point (0.25\u0026deg; \u0026times; 0.25\u0026deg; spatial resolution) for each 3-hourly period of the past 74 years, from 1950 to 2023, In total, more than 80\u0026nbsp;billion atmospheric profiles were used to construct the global climatology of VLH (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe global maximum in VLH frequency is found in northern Argentina, where the mean annual number of VLH events is around 0.50, the equivalent of a VLH event every two years per grid box. This aligns with previous studies (Mezher et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), which report the highest frequency in the northern lee of the Argentina Andes. VLH frequency peaks from November to February although it has decreased in recent decades, particularly in December and January. Other areas of high VLH frequency include the tri-border region of Uruguay, Paraguay, and Brazil. Smaller but non-zero VLH frequency extends also into the Central Andes region of Bolivia and southern Peru. Across South America, the necessary ingredients for hailstorms frequently overlap. The Andes foothills provide a focus for lift (Rasmussen and Houze 2011), while the South American low-level jet advects warm and moist air from the Amazon. This, combined with high deep vertical wind shear, provides an ideal environment for hailstorms (Bechis et al. 2022).\u003c/p\u003e \u003cp\u003eIn the USA, the Great Plains exhibit a high VLH frequency with a maximum located between Kansas, Nebraska, and Colorado, aligning with radar-based climatologies (Murillo et al. 2021, Wendt and Jirak \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The overlap of moisture from the Gulf of Mexico (Molina et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), Elevated Mixed Layers generated over the high terrain of the southwest (Andrews et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and cyclonic activity in spring and early summer (Li et al. 2021) provide a combination of high instability and wind shear which makes the region a hail hotspot. A pronounced east-west gradient is observed across northern Texas, where VLH frequency peaks between March and June. The VLH seasonal cycle in Dallas corresponds with the dates of major hail events: 5 April 2003 (Marshall et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), 24 May 2011 (Brown et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), and 11 June 2023 (Gallagher RE \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In Canada, the frequency is lower, with the most affected areas being southern Saskatchewan and Alberta, in line with previous studies (Brunet and Brimelow \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSouth Africa is another VLH hotspot, with a frequency exceeding 0.30 events per year in the northeast and along the Eastern Cape, where the Agulhas current provides moisture and the eastern Escarpment Mountains a focus for lift (Blamey et al. 2016). In Durban, over the last two decades, the peak month of occurrence has shifted from November to October. VLH is also present in equatorial Africa, the western Sahel, and northern Africa however a lack of ground-based reports and uncertainties in the accuracy of satellite-based estimates (Ferraro et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) complicate comparisons with observations in these regions.\u003c/p\u003e \u003cp\u003eVLH is less frequent in Europe, with the occurrence maximized in the lee of mountain ranges. High VLH frequency is found across northeastern Spain, southwestern France, and northern Italy which stands out as the European hotspot, in agreement with previous studies (Punge et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The seasonal cycle in Milan peaks in summer, although there is a tendency towards an earlier onset of VLH occurrence in spring. Other local maxima are observed in Switzerland, southern Germany, and southeastern Austria, with VLH occurrence extending into southeast Europe. Further east, VLH frequency peaks locally in Turkey, Iran, and the Arabian Peninsula, though climatological studies to confirm these patterns are lacking.\u003c/p\u003e \u003cp\u003eAsia experiences the lowest VLH activity, with northern Pakistan and Bangladesh having the highest frequency at around 0.12 events per year. VLH also peaks locally in northeastern China during spring and summer, while lower but non-zero VLH frequency is found over the Tibetan Plateau and southern China, consistent with ground-based report climatologies (Ni et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn Australia, VLH is most frequent in the southeast, as shown by Brook et al. (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). A second maximum is present in the southwest interior, though reports in this area are rare due to low population density, and the limited radar coverage prevents verification of these findings.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eThe global climate trends of very large hail\u003c/h2\u003e \u003cp\u003eTrends in VLH frequency (1950\u0026ndash;2023) are not globally homogenous but exhibit a strong zonal dependence: positive and significant trends are almost exclusively present in the Northern Hemisphere, while negative trends are limited to the Southern Hemisphere (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eEurope is the only continent to exhibit a widespread and statistically significant increase in VLH frequency, extending into northern Africa and the Middle East. These findings align with pan-European (R\u0026auml;dler et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, P\u0026uacute;čik et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, Taszarek et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2021a\u003c/span\u003e) and regional analyses from single countries (e.g., Switzerland; Wilhelm et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Northern Italy shows the largest global increase in VLH frequency (up to +\u0026thinsp;0.03 events per decade). In Milan, VLH frequency has consistently exceeded the 1950\u0026ndash;2023 average over the last two decades, with 2021\u0026ndash;2023 marking the most active years. This period corresponds with a record number of hail reports in Italy, including a 19-cm hailstone in July 2023 (ESWD 2024). Positive and significant trends are not limited to Italy but extend across most of mainland Europe.\u003c/p\u003e \u003cp\u003eNorth America displays positive VLH trends, particularly in southern Canada, where southern Alberta has the largest increase (+\u0026thinsp;0.01 events per decade). An increase of similar magnitude is also seen in the mountainous regions of Mexico. Trends are mostly not significant in the USA except for very localized increases in the Great Plains. Compared to previous work focusing on hail-prone environments (Tang et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), the increase is more localized, although a tendency for very large hailstones to become more frequent has been shown in recent studies (Gensini et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In Dallas, VLH frequency has increased by 43.4% since the 1950s, and 2023 was the most active year, with a\u0026thinsp;+\u0026thinsp;105% departure from the 1950\u0026ndash;2023 average. This aligns with the record losses reported across Texas in 2023 (Munich RE \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn northern Argentina, VLH frequency has decreased the most. The decrease is steady in Mendoza, where every year since 2006, except for 2015 and 2016, has experienced below-average VLH frequency (mean = -29.8%). This decrease is supported by a reduction in hail reports (Beal et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Localized positive trends are found in Bolivia and Peru.\u003c/p\u003e \u003cp\u003eSouth Africa has the second-largest VLH decrease, especially in the northeast, where the decadal trend reaches \u0026minus;\u0026thinsp;0.02 events per decade. This trend is also supported by a general reduction in severe thunderstorm environments (Taszarek et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e). Negative trends are observed locally in Equatorial Africa, although they remain unverified due to a lack of direct observations. In contrast, northern Africa shows a positive VLH frequency trend comparable in magnitude to that of mainland Europe, as noted by R\u0026auml;dler et al. (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn Asia, VLH frequency trends are generally small or non-existent. However, positive and significant trends are present across Pakistan, the Tibetan Plateau, and northeastern China. In Beijing, above-average VLH activity occurred in the 1980s, followed by a decreasing trend in hail days from 1980 to 2005 (Zhang et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Negative and significant trends in the region are limited to the eastern coast of India.\u003c/p\u003e \u003cp\u003eFinally, in Australia, trends are largely negative in continental areas of the north while coastal regions show constant VLH frequency. In Brisbane, a strong year-to-year variability is present, and trends are not significant, as also shown by Raupach et al. (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eHail trends attribution and relationship with global warming\u003c/h3\u003e\n\u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, we display the grid-based correlation between the annual average VLH frequency and the 2-meter temperature yearly mean (1950\u0026ndash;2023). The relationship between temperature change and VLH trends is not uniform globally.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eEurope is the only region where an increase in 2-meter temperature significantly correlates with higher VLH frequency, especially in southern Europe (\u003cem\u003er\u003c/em\u003e\u0026thinsp;\u0026asymp;\u0026thinsp;0.7), the Mediterranean, Turkey, and the Middle East. This positive correlation reflects an increase in instability and thunderstorm intensity (Taszarek et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e, Battaglioli et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023a\u003c/span\u003e). In the USA, the correlation is weaker and spatially heterogeneous, with positive values in the east and negative ones in the west, where drying in the mid-levels reduces storm potential (Taszarek et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e, Andrews et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn South America, a bimodal pattern appears along the lee of the Andes, with positive correlations (\u003cem\u003er\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.5) in Peru and Bolivia but negative correlations in Argentina. Similar patterns occur in southern and central Brazil however correlations remain small (-0.3\u0026thinsp;\u0026lt;\u0026thinsp;\u003cem\u003er\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;+\u0026thinsp;0.3), suggesting other factors might better explain VLH frequency trends. In Africa, correlations are generally small and non-significant, except in the Sahel region, where \u003cem\u003er\u003c/em\u003e \u0026lt; -0.5. This area coincides with a marked decrease in convective precipitation (Taszarek et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021b\u003c/span\u003e)\u003c/p\u003e \u003cp\u003eCorrelations vary by region but remain small in Australia, ranging from \u0026minus;\u0026thinsp;0.3 inland to +\u0026thinsp;0.3 on the coast. The correlation mirrors moisture patterns (Supplementary Fig.\u0026nbsp;1), with increasing coastal moisture and inland drying, which reduces storm potential (Raupach et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn conclusion, although temperature as a stand-alone parameter does not represent an ingredient for storm or hail formation, changes in its values can correlate with changes in storm ingredients, influencing VLH frequency.\u003c/p\u003e\n\u003ch3\u003eNorthern Italy and Northwestern Argentina: Quantitative Trend Attribution\u003c/h3\u003e\n\u003cp\u003eObserving that upward 2-meter temperature trends coincide with both increasing and decreasing trends in VLH frequency prompts us to investigate the drivers of these changes. We focus on two regions with the most pronounced and opposite trends, northern Italy and northwestern Argentina, and analyze the contributions of the various predictor parameters of the VLH model (Supplementary Table\u0026nbsp;1) to the trends depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eIn northern Italy, the sum of all predictors\u0026rsquo; contributions is +\u0026thinsp;73%. While thunderstorms have become slightly less frequent (-6%), their severity has substantially increased. More specifically, the probability that a storm will be capable of producing VLH has risen by 79%. This increase is mainly due to rising atmospheric instability in the cold portion of the storm, which results from enhanced latent heat release and higher water vapor content. This aligns with previous findings (R\u0026auml;dler et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, Pilguj et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) suggesting that increased instability leads to stronger storms in the mid-latitudes.\u003c/p\u003e \u003cp\u003eIn northwestern Argentina, the sum of all predictors\u0026rsquo; contributions is -25%. Thunderstorms have both become less frequent (-6%) and less prone to producing VLH (-16%). The reduction in thunderstorm likelihood is consistent with patterns of increasing drought (Feron et al. 2024) and is primarily attributed to declining mid-level atmospheric relative humidity and instability.\u003c/p\u003e \u003cp\u003eA detailed breakdown of parameter contributions for both northern Italy and northwestern Argentina is provided in Supplementary Table\u0026nbsp;2.\u003c/p\u003e\n\u003ch3\u003eTrends in insured losses across Europe, the United States, and Australia\u003c/h3\u003e\n\u003cp\u003eUnderstanding the relationship between changes in VLH frequency and hail losses is crucial, given the significant economic impacts of hail (P\u0026uacute;čik et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). We analyzed trends in hail loss events in regions with high insurance penetration: Germany, Austria, Benelux countries (Europe), the eastern two-thirds of the USA, and populated coastal regions of Australia. To compare long-term trends in hail loss events with VLH frequencies, we aggregated VLH probabilities over the selected domains from 1993 to 2023. Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e shows the comparison between the number of hail loss events (see Methods) and the accumulated VLH probabilities in the three regions.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eOver the past thirty years, hail loss events have increased in Europe (Germany, Austria, Benelux), the USA, and Australia, driven by a combination of social, economic, and environmental factors (Kron et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). The modeled VLH frequency generally reproduces yearly variations in annual loss events well, especially in Europe between 1998 and 2007. The correlation between the trends in VLH frequency and those in hail loss events suggests that increasing VLH occurrence strongly contributes to the rise in hail loss events in Europe. In contrast, the USA and Australia display different patterns. While VLH frequency increased only slightly in the USA and decreased in Australia, hail loss events continued to rise in both regions. This discrepancy suggests that socio-economic factors, such as changes in exposure and vulnerability, are more important drivers of the increase in hail loss events than meteorological factors. These findings are consistent with previous research on tornadoes in the USA (Strader et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, Bouwer \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e"},{"header":"Summary and Discussion","content":"\u003cp\u003eWe reconstructed the global occurrence of VLH over a period of 74 years (1950–2023). using an additive logistic regression model (AR-CHaMo) trained with lightning observations, hail reports, and atmospheric predictors from the ERA5 reanalysis across Europe, the USA, and Australia. VLH is most frequent across northern Argentina and the border regions of Uruguay, Paraguay, and Brazil. High VLH frequency is also modeled in the USA Great Plains and parts of South Africa. While VLH is less frequent in other regions (Europe, Africa, Oceania, and Asia), the reconstruction aligns with regional climatologies in both the simulated spatial distribution and the seasonal patterns. Additionally, the recent extreme hail activity in the USA and Europe is well represented by the model, with 2023 standing out as the year with the highest VLH frequency.\u003c/p\u003e \u003cp\u003eAlthough no universally valid relationship between VLH trends and global warming has been identified, some significant regional correlations exist. In Europe, VLH frequency has increased the most, correlating with rising temperatures. Trends are especially pronounced in northern Italy, driven by an increase in low-level moisture and instability in the cold part of storms. Other regions with positive significant trends include the Middle East, southern Canada, parts of Mexico, and localized areas in the USA. Increasing temperatures do not always correlate with an increase in VLH frequency: in the western USA and continental Australia, drying has reduced the storm potential. Negative trends in VLH frequency are observed mainly in the Southern Hemisphere, with the strongest declines in South Africa and northern Argentina, where they are primarily driven by a decrease in mid-level humidity and instability.\u003c/p\u003e \u003cp\u003eFinally, we examined the relationship between changes in VLH frequency and hail losses. Hail loss events have increased in parts of Europe, the USA, and Australia. In Europe, rising VLH occurrence contributes to the increase in hail losses, alongside growing exposure and vulnerability. In the USA and Australia, changes in meteorological conditions do not correlate with the increase in hail loss events; instead, exposure and vulnerability drive the increase. To confirm this hypothesis, future studies should dive deeper into the topic of loss normalization or investigate trends in exposure and vulnerability.\u003c/p\u003e \u003cp\u003eSome limitations must be considered when interpreting these results. The VLH model, trained on three mid-latitude regions, was applied globally therefore, results should be interpreted with caution in areas where regional climatologies do not allow for verification or known biases in ERA5 are present (e.g., the tropics; Taszarek et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2021c\u003c/span\u003e). Future research should incorporate newly available databases from regions like South America, Canada, and China to better capture hail environments on a global scale. Another area of future research concerns the use of reanalysis data at a higher resolution to explicitly simulate storm development, which was not possible in ERA5.\u003c/p\u003e \u003cp\u003eDespite the limitations, this is the first analysis of the climatological occurrence of VLH, the associated long-term trends in the context of a warming climate, and the relationship with losses on a global scale.\u003c/p\u003e"},{"header":"Data \u0026 Methods","content":"\u003ch2\u003eAR-CHaMo models\u003c/h2\u003e\u003cp\u003eThe Additive Regressive Convective Hazard Model (AR-CHaMo; Rädler et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, Battaglioli et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023a\u003c/span\u003e) is a logistic regression model that yields the probability of VLH as a function of environmental predictors from the atmospheric reanalysis (listed in Supplementary Table\u0026nbsp;1) that were calculated using the ThundeR rawinsonde package (Taszarek et al. 2023). AR-CHaMo calculates this probability by combining the likelihood of a thunderstorm (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{lightning}\\)\u003c/span\u003e\u003c/span\u003e) and the conditional probability of VLH given a thunderstorm (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{VLH|lightning}\\)\u003c/span\u003e\u003c/span\u003e). \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{lightning}\\)\u003c/span\u003e\u003c/span\u003e was trained simultaneously using lightning observations (over a billion) from the Arrival Time Difference network (ATDnet; Enno et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) across Europe (34.5°–63.5°N, 9.0°W–46.0°E), the National Lightning Detection Data (NLDN; Koehler \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) from the USA (29.0°–41.5°N, 109.0°–79.0°W) and the Global Position and Tracking Systems (GPATS; Dowdy and Kuleshov \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) from Australia (10.5°– 40.0°S, 119.0°–153.5°E). \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{VLH|lightning}\\)\u003c/span\u003e\u003c/span\u003e was trained simultaneously using hail reports (over 4500) from the European Severe Weather Database (ESWD; 45.0°–54.0°N, 5.0°–22.0°E), the Storm Prediction Centre (SPC) Storm Archive (30.5°–41.5°N, 105.0°–82.0°W) and the Australian Bureau of Meteorology (BOM; 22.5°– 37.5°S, 149.5°–153.5°E). \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{lightning}\\)\u003c/span\u003e\u003c/span\u003e was improved by using a grid-based ratio between lightning observations from the Earth Networks Global Lightning Detection Network (ENGLN; DiGangi et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and modeled lightning occurrence, improving the lightning distribution, especially across (sub-)tropical regions where AR-CHaMo was not trained. The mean annual expected number of VLH events was calculated as the sum of all three-hourly probabilities in a year multiplied by three. To do so, we assumed three-hourly probabilities to be independent of each other, although this may not strictly be the case.\u003c/p\u003e\u003ch3\u003eInsurance loss data\u003c/h3\u003e\u003cp\u003eData on hail losses were obtained from the NatCatSERVICE database of Munich RE (Kron et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Within the database, each loss entry includes its type, location, date, loss, and description. Using the description, we filtered only for events where the loss was at least partly caused by hail, ensuring that events caused exclusively by other convective hazards were not taken into consideration. Single events often covered multiple days, meaning that those entries included multiple severe convective storms. In these cases, loss location coordinates indicate the loss center. To account for inflation, urban growth, and regional wealth differences, the estimated overall losses were normalized to the 2019 levels of destructible wealth. For 2020 to 2023, we used the inflation-adjusted values of the losses. Normalization was done by considering the Gross Domestic Product discretized on a 1° × 1° grid.\u003c/p\u003e\u003cp\u003eThe accuracy of the loss estimates depends on the insurance penetration in a given country for a specific hazard (Kron et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). In Europe, we selected Germany, the Benelux countries, and Austria. Australia was also considered, given the fact that hail causes substantial losses (Schuster et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), similar to the USA (Munich RE 2023). For the USA, we applied a threshold of USD 1\u0026nbsp;billion to capture the most impactful events and to limit the chances that “minor” events were missed in the early times.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eF.B., P.G., and T.P. contributions were supported by ESSL. M.T. was funded by a grant from the Polish National Science Centre (grant no. 2020/39/D/ST10/00768). A.R. contribution was supported by Munich RE. The ERA5 reanalysis computations were performed in the Poznań Supercomputing and Networking Center (grant no. 648).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAuthor Affiliations\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEuropean Severe Storms Laboratory e.V., Wessling, Germany\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFrancesco Battaglioli, Pieter Groenemeijer\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEuropean Severe Storms Laboratory - Science \u0026amp; Training, Wiener Neustadt, Austria\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePieter Groenemeijer, Tomas P\u0026uacute;čik\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDepartment of Meteorology and Climatology, Adam Mickiewicz University, Poznan, Poland\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMateusz Taszarek\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMunich Re, Munich, Germany\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAnja R\u0026auml;dler\u003c/p\u003e\n\u003cp\u003eContributions\u003c/p\u003e\n\u003cp\u003eF.B. conceived the idea and designed this study. F.B. performed the analysis, wrote the manuscript, and produced all figures apart from Fig. 4 (A.R). M.T. calculated environmental predictors from the ERA5 reanalysis and contributed to the interpretation of the results. P.G. and T.P. contributed to the scientific analysis of the results together with A.R., who contributed especially to the section \u0026ldquo;Trends in insured losses across Europe, the United States, and Australia\u0026rdquo;.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorresponding author\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrespondence to Francesco Battaglioli ([email protected]).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declaration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eERA5 hourly data on single levels can be obtained from the Copernicus Climate Data Store at ​​\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\u003c/span\u003e\u003cspan address=\"https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=overview\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Convective parameters from ERA5 data are calculated using the open-source ThundeR package that is available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://bczernecki.github.io/thundeR/\u003c/span\u003e\u003cspan address=\"https://bczernecki.github.io/thundeR/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. USA hail reports are available from the Storm Prediction Centre Storm Archive (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.spc.noaa.gov/exper/archive/events/\u003c/span\u003e\u003cspan address=\"https://www.spc.noaa.gov/exper/archive/events/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), Australian hail reports are available from the Bureau of Meteorology (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.bom.gov.au/australia/stormarchive/\u003c/span\u003e\u003cspan address=\"http://www.bom.gov.au/australia/stormarchive/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and European ones can be obtained from the European Severe Weather Database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://eswd.eu/\u003c/span\u003e\u003cspan address=\"https://eswd.eu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). ENGLN global lightning observations are publicly available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://thunderhours.earthnetworks.com\u003c/span\u003e\u003cspan address=\"http://thunderhours.earthnetworks.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e while regional ones were obtained for this study and are property of the Met Office (ATDnet), Vaisala (NLDN), and the Bureau of Meteorology (GPATS). The NatCatSERVICE dataset is property of Munich RE and was made available exclusively for this study. The AR-CHaMo datasets can be made available upon request to the corresponding author, Francesco Battaglioli ([email protected]).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eCode availability\u003c/h2\u003e \u003cp\u003ePredictor parameters were calculated using the ThundeR rawinsonde package, which is openly accessible at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.rawinsonde.com/thunder_app/\u003c/span\u003e\u003cspan address=\"http://www.rawinsonde.com/thunder_app/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Code for the model calculation can be made available upon request to the corresponding author, Francesco Battaglioli ([email protected]).\u003c/p\u003e \u003c/div\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAndrews, M. S., V. A. Gensini, A. M. Haberlie, W. S. Ashley, A. C. Michaelis, \u0026amp; Taszarek M. Climatology of the Elevated Mixed Layer over the Contiguous United States and Northern Mexico Using ERA5: 1979\u0026ndash;2021. J. Climate, 37, 1833\u0026ndash;1851 (2024) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JCLI-D-23-0517.1\u003c/span\u003e\u003cspan address=\"10.1175/JCLI-D-23-0517.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBang, S., \u0026amp; Cecil D. Constructing a Multifrequency Passive Microwave Hail Retrieval and Climatology in the GPM Domain. J. Appl. Meteor. Climatol. 58, 1889\u0026ndash;1904 (2019) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/jamc-d-19-0042.1\u003c/span\u003e\u003cspan address=\"10.1175/jamc-d-19-0042.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBattaglioli, F., P. Groenemeijer, T. P\u0026uacute;čik, M. Taszarek, U. Ulbrich, \u0026amp; Rust H. Modeled Multidecadal Trends of Lightning and (Very) Large Hail in Europe and North America (1950\u0026ndash;2021). J. Appl. Meteorol. Clim. 62(11), 1627\u0026ndash;1653 (2023a) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JAMC-D-22-0195.1\u003c/span\u003e\u003cspan address=\"10.1175/JAMC-D-22-0195.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeal, A., R. Hallak, L. D. Martins, J. A. Martins, G. Biz, A. P. Rudke, \u0026amp; Tarley C. R. T. Climatology of hail in the triple border paran\u0026aacute;, Santa Catarina (Brazil) and Argentina. Atmos. Res. 234, 104747 (2020) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.atmosres.2019.104747\u003c/span\u003e\u003cspan address=\"10.1016/j.atmosres.2019.104747\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBechis, H., \u0026amp; Coauthors. A case study of a severe hailstorm in Mendoza, Argentina, during the RELAMPAGO-CACTI field campaign. Atmos Res. 271, 106127 (2022) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.atmosres.2022.106127\u003c/span\u003e\u003cspan address=\"10.1016/j.atmosres.2022.106127\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlamey, R. C., C. Middleton, C. Lennard, \u0026amp; Reason C. J. C. A climatology of potential severe convective environments across South Africa. Clim. Dyn. 49, 2161\u0026ndash;2178 (2017), doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00382-016-3434-7\u003c/span\u003e\u003cspan address=\"10.1007/s00382-016-3434-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBouwer, L. M. Observed and Projected Impacts from Extreme Weather Events: Implications for Loss and Damage. \u003cem\u003eLoss and Damage from Climate Change.\u003c/em\u003e Climate Risk Management, Policy and Governance, 63\u0026ndash;82 (2019) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-319-72026-5_3\u003c/span\u003e\u003cspan address=\"10.1007/978-3-319-72026-5_3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrunet, D. \u0026amp; Brimelow J. A Hail Climatology for Canada Using a Lightning Proxy. J. Appl. Meteor. Climatol. 63, 1227\u0026ndash;1240 (2024) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JAMC-D-23-0214.1\u003c/span\u003e\u003cspan address=\"10.1175/JAMC-D-23-0214.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrook, J. P., J. S. Soderholm, A. Protat, H. McGowan, \u0026amp; Warren R. A. A radar- based hail climatology of Australia. Mon. Wea. Rev. 152(2), 607\u0026ndash;628 (2024) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/mwr-d-23- 0130.1\u003c/span\u003e\u003cspan address=\"10.1175/mwr-d-23- 0130.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown, T. M., W. H. Pogorzelski, \u0026amp; Giammanco I. M. Evaluating hail damage using property insurance claims data. Weather, Climate, and Society. 7(3), 197\u0026ndash;210 (2015) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/wcas-d-15-0011.1\u003c/span\u003e\u003cspan address=\"10.1175/wcas-d-15-0011.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDiGangi, E. A., M. Stock, \u0026amp; Lapierre J. Thunder Hours: How Old Methods Offer New Insights into Thunderstorm Climatology. Bull. Amer. Meteor. Soc. 103, E548\u0026ndash;E569 (2022) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/BAMS-D-20-0198.1\u003c/span\u003e\u003cspan address=\"10.1175/BAMS-D-20-0198.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDowdy, A. J., \u0026amp; Kuleshov Y. Climatology of lightning activity in Australia: spatial and seasonal variability. Australian Meteorological and Oceanographic J. 64(2), 103\u0026ndash;108 (2014) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1071/ES14010\u003c/span\u003e\u003cspan address=\"10.1071/ES14010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEnno, S., J. Sugier, R. Alber, \u0026amp; Seltzer M. Lightning flash density in Europe based on 10 years of ATDnet data. Atmos. Res. 235, 104769 (2020) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.atmosres.2019.104769\u003c/span\u003e\u003cspan address=\"10.1016/j.atmosres.2019.104769\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eESWD. European Severe Weather Database, accessed 6 April 2024, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://eswd.eu\u003c/span\u003e\u003cspan address=\"https://eswd.eu\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeron, S. \u0026amp; Coauthors. South America is becoming warmer, drier, and more flammable. Commun. Earth Environ. 5, 501 (2024) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s43247-024-01654-7\u003c/span\u003e\u003cspan address=\"10.1038/s43247-024-01654-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFerraro, R., J. Beauchamp, D. Cecil, \u0026amp; Heymsfield G. A prototype hail detection algorithm and hail climatology developed with the Advanced Microwave Sounding Unit (AMSU). Atmos. Res. 163, 24\u0026ndash;35 (2015) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.atmosres.2014.08.010\u003c/span\u003e\u003cspan address=\"10.1016/j.atmosres.2014.08.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGallagher RE. 2023: Natural Catastrophe and Climate Report: 2023, Available online: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ajg.com/gallagherre/-/media/files/gallagher/gallagherre/news-and-insights/2024/january/natural-catastrophe-and-climate-report-2023.pdf\u003c/span\u003e\u003cspan address=\"https://www.ajg.com/gallagherre/-/media/files/gallagher/gallagherre/news-and-insights/2024/january/natural-catastrophe-and-climate-report-2023.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(accessed on 3 Dec 2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGensini, V. A., W. Ashley, A. C. Michaelis, A. M. Haberlie, J. Goodin, \u0026amp; Wallace B. C. Hail size dichotomy in a warming climate. npj Climate and Atmos. Sci. 7, 185 (2024) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41612-024-00728-9\u003c/span\u003e\u003cspan address=\"10.1038/s41612-024-00728-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGunturi, P., \u0026amp; Tippett, M. K. Managing severe thunderstorm risk: Impact of ENSO on U.S. tornado and hail frequencies (Tech. Rep.). Minneapolis, MN, USA: Willis Re (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHersbach, H., \u0026amp; Coauthors. The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc. 146, 1999\u0026ndash;2049 (2020) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/QJ.3802\u003c/span\u003e\u003cspan address=\"10.1002/QJ.3802\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, F., D. R. Chavas, K. A. Need, N. Rosenbloom, \u0026amp; Dawson D. T. The role of Elevated Terrain and the Gulf of Mexico in the Production of Severe Local Storm Environments over North America. J Clim. 34(19), 7799\u0026ndash;7819 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JCLI-D-20-0607.1\u003c/span\u003e\u003cspan address=\"10.1175/JCLI-D-20-0607.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoehler, T. Cloud-to-Ground Lightning Flash Density and Thunderstorm Day Distributions over the Contiguous United States Derived from NLDN Measurements: 1993\u0026ndash; 2018. Mon. Wea. Rev. 148(1), 313\u0026ndash;332 (2020) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/MWR-D-19-0211.1\u003c/span\u003e\u003cspan address=\"10.1175/MWR-D-19-0211.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKron, W., M. Steuer, P. L\u0026ouml;w, \u0026amp; Wirtz A. How to deal properly with a natural catastrophe database \u0026ndash; analysis of flood losses Nat. Hazards Earth Syst. Sci. 12, 535\u0026ndash;550 (2012) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/nhess-12-535-2012\u003c/span\u003e\u003cspan address=\"10.5194/nhess-12-535-2012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKunz, M., U. Blahak, J. Handwerker, M. Schmidberger, H. J. Punge, S. Mohr, E. Fluck, \u0026amp; Bedka, K. M. The severe hailstorm in southwest Germany on 28 July 2013: Characteristics, impacts and meteorological conditions. Quart. J. Roy. Meteor. Soc. 144, 231\u0026ndash;250 (2018) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/qj.3197\u003c/span\u003e\u003cspan address=\"10.1002/qj.3197\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarshall, T. P., R. F. Herzog, D. E. Mitchell, S. J. Morrison, and S. Rae, 2004: The Dallas- Ft. Worth, TX hailstorm: 5 April 2003. \u003cem\u003e22nd Conf. on Severe Local Storms\u003c/em\u003e, P9.1. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ams.confex.com/ams/11aram22sls/techprogram/paper_81089.htm\u003c/span\u003e\u003cspan address=\"https://ams.confex.com/ams/11aram22sls/techprogram/paper_81089.htm\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMolina, M. J., R. P. Timmer, \u0026amp; Allen, J. T. Importance of the Gulf of Mexico as climate driver for U.S. severe thunderstorm activity. Geophys. Res. Lett. 43(23), 12,295\u0026thinsp;\u0026ndash;\u0026thinsp;12,304 (2016) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/2016GL071603\u003c/span\u003e\u003cspan address=\"10.1002/2016GL071603\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurillo, E., C. Homeyer, \u0026amp; Allen, J. A 23-Year Severe Hail Climatology Using GridRad MESH Observations. Mon. Wea. Rev. 149, 945\u0026ndash;958 (2023) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/mwr-d-20-0178.1\u003c/span\u003e\u003cspan address=\"10.1175/mwr-d-20-0178.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMunich RE 2024: Natural disasters of 2023: Available online: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.munichre.com/en/company/media-relations/media-information-and-corporate-news/media-information/2024/natural-disaster-figures-2023.html\u003c/span\u003e\u003cspan address=\"https://www.munichre.com/en/company/media-relations/media-information-and-corporate-news/media-information/2024/natural-disaster-figures-2023.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (accessed on 24. Feb 2025)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMezher, R. N., M. Doyle, \u0026amp; Barros V. Climatology of hail in Argentina. Atmos. Res. 114\u0026ndash;115, 70\u0026ndash;82 (2012) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.atmosres.2012.05.020\u003c/span\u003e\u003cspan address=\"10.1016/j.atmosres.2012.05.020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNi, X., A. Muehlbauer, J. T. Allen, Q. Zhang, \u0026amp; Fan, J. A Climatology and Extreme Value Analysis of Large Hail in China. Mon. Wea. Rev. 148, 1431\u0026ndash;1447 (2020) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/MWR-D-19-0276.1\u003c/span\u003e\u003cspan address=\"10.1175/MWR-D-19-0276.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePanosetti, D., \u0026amp; Tomassetti, U. The July 2023 Northern Italy hailstorms from a climatological and (re)insurance market perspective. EGU General Assembly 2024, Vienna, Austria.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePilguj, N., M. Taszarek, M. Kryza, \u0026amp; Brooks H. Reconstruction of Violent Tornado Environments in Europe: High-Resolution Dynamical Downscaling of ERA5. Geophys. Res. Lett. 49(11), (2022) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2022GL098242\u003c/span\u003e\u003cspan address=\"10.1029/2022GL098242\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePodlaha, A., S. Bowen, \u0026amp; M. L\u0026ouml;rinc. Weather, climate and catastrophe insight: 2019 annual report. Aon Annual Rep., 83 pp., (2020) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://thoughtleadership.aon.com/Documents/20200122-if-natcat2020.pdf\u003c/span\u003e\u003cspan address=\"http://thoughtleadership.aon.com/Documents/20200122-if-natcat2020.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrein, A. \u0026amp; Holland G. Global estimates of damaging hail hazard. Wea. and Climate Extremes. 22, 10\u0026ndash;23 (2018) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.wace.2018.10.004\u003c/span\u003e\u003cspan address=\"10.1016/j.wace.2018.10.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eP\u0026uacute;čik, T., C. Castellano, P. Groenemeijer, T. K\u0026uuml;hne, A. R\u0026auml;dler, B. Antonescu, \u0026amp; Faust E. Large Hail Incidence and Its Economic and Societal Impacts across Europe. Mon. Wea. Rev. 147, 3901\u0026ndash;3916 (2019) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/mwr-d-19-0204.1\u003c/span\u003e\u003cspan address=\"10.1175/mwr-d-19-0204.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePunge, H., K. Bedka, M. Kunz, \u0026amp; Reinbold A. Hail frequency estimation across Europe based on a combination of overshooting top detections and the ERA-INTERIM reanalysis. Atmos. Res. 198, 34\u0026ndash;43 (2017) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.atmosres.2017.07.025\u003c/span\u003e\u003cspan address=\"10.1016/j.atmosres.2017.07.025\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRana, V. S., Sharma, S., Rana, N., Sharma, U., Patiyal, V., Banita, \u0026amp; Prasad H. Management of Hailstorms under a Changing Climate in Agriculture: A Review. Environ. Chem. Lett. 20, 3971\u0026ndash;3991 (2022) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10311-022-01502-0\u003c/span\u003e\u003cspan address=\"10.1007/s10311-022-01502-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR\u0026auml;dler, A., P. Groenemeijer, E. Faust, \u0026amp; Sausen R. Detecting Severe Weather Trends Using an Additive Regressive Convective Hazard Model (AR-CHaMo). J. Appl. Meteor. Climatol. 57(3), 569\u0026ndash;587 (2018) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JAMC-D-17-0132.1\u003c/span\u003e\u003cspan address=\"10.1175/JAMC-D-17-0132.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRaupach, T. H., J. S. Soderholm, R. A. Warren, \u0026amp; Sherwood S. C. Changes in hail hazard across Australia: 1979\u0026ndash;2021. npj Climate and Atmos. Sci. 6(1) (2023) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41612-023-00454-8\u003c/span\u003e\u003cspan address=\"10.1038/s41612-023-00454-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRasmusen, K. L. \u0026amp; Houze R. A. Orogenic Convection in Subtropical South America as seen by the TRMM Satellite. Mon Wea. Rev. 139(8), 2399\u0026ndash;2420 (2011) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/MWR-D-10-05006.1\u003c/span\u003e\u003cspan address=\"10.1175/MWR-D-10-05006.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchuster, S., R. Blong, R. Leigh, \u0026amp; McAneney K. Characteristics of the 14 April 1999 Sydney hailstorm based on ground observations, weather radar, insurance data and emergency calls. Nat. Hazards Earth Syst. Sci. 5(5), 613\u0026ndash;620 (2005) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/nhess- 5-613-2005\u003c/span\u003e\u003cspan address=\"10.5194/nhess- 5-613-2005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrader, S. M., W. S. Ashley, T. J. Pingel, \u0026amp; Krmenec A. J. Projected 21st century changes in tornado exposure, risk, and disaster potential. Climatic Change. 141, 301\u0026ndash;313 (2017) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10584-017-1905-4\u003c/span\u003e\u003cspan address=\"10.1007/s10584-017-1905-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang, B., V. Gensini, \u0026amp; Homeyer C. Trends in United States large hail environments and observations. npj Climate and Atmos. Sci. 2(1), 45 (2019) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41612-019-0103-7\u003c/span\u003e\u003cspan address=\"10.1038/s41612-019-0103-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaszarek, M., J. Allen, H. Brooks, N. Pilguj, and Czernecki B. Differing Trends in United States and European Severe Thunderstorm Environments in a Warming Climate. Bull. Amer. Meteor. Soc. 102, 296\u0026ndash;322 (2021a) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/bams-d-20-0004.1\u003c/span\u003e\u003cspan address=\"10.1175/bams-d-20-0004.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaszarek, M., J. T. Allen, M. Marchio, \u0026amp; Brooks H. E. Global climatology and trends in convective environments from ERA5 and Rawinsonde Data. npj Climate and Atmos. Sci. 4(1). (2021b) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41612-021-00190-x\u003c/span\u003e\u003cspan address=\"10.1038/s41612-021-00190-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaszarek, M., N. Pilguj, J. Allen, V. Gensini, H. Brooks, \u0026amp; Szuster P. Comparison of convective parameters derived from ERA5 and MERRA2 with rawinsonde data over Europe and North America. J. Climate. 34(8), 1\u0026ndash;55 (2021c) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/jcli-d-20-0484.1\u003c/span\u003e\u003cspan address=\"10.1175/jcli-d-20-0484.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaszarek, M., B. Czernecki, \u0026amp; Szuster, P. thundeR - a rawinsonde package for processing convective parameters and visualizing atmospheric profiles, 11th European Conference on Severe Storms, Bucharest, Romania, 8\u0026ndash;12 May 2023, ECSS2023-28, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/ecss2023-28\u003c/span\u003e\u003cspan address=\"10.5194/ecss2023-28\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWendt, N., \u0026amp; Jirak I. An Hourly Climatology of Operational MRMS MESH- Diagnosed Severe and Significant Hail with Comparisons to Storm Data Hail Reports. Wea. Forecasting 36, 645\u0026ndash;659 (2021) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/waf-d-20-0158.1\u003c/span\u003e\u003cspan address=\"10.1175/waf-d-20-0158.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilhelm, L., C. Schwierz, K. Schr\u0026ouml;er, M. Taszarek, \u0026amp; Martius O. Reconstructing hail days in Switzerland with statistical models (1959\u0026ndash;2022). Nat. Hazards Earth Syst. Sci. 24, 3869\u0026ndash;3894 (2024) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/nhess-24-3869-2024\u003c/span\u003e\u003cspan address=\"10.5194/nhess-24-3869-2024\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXie, B., Zhang, Q., \u0026amp; Wang, Y. Trends in hail in China during 1960\u0026ndash;2005. Geophys. Res. Lett., 35, L13801 (2008) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2008GL034067\u003c/span\u003e\u003cspan address=\"10.1029/2008GL034067\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, C., Q. Zhang, \u0026amp; Wang, Y. Climatology of hail in China: 1961\u0026ndash;2005. J. Appl. Meteor. Climatol. 47(3), 795\u0026ndash;804 (2008) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1175/2007jamc1603.1\u003c/span\u003e\u003cspan address=\"10.1175/2007jamc1603.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6196143/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6196143/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWe statistically constructed a long-term (1950\u0026ndash;2023) global climatology of very large hail (VLH; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\ge\\:\\)\u003c/span\u003e\u003c/span\u003e 5 cm) and investigated the associated trends in the context of a warming climate. Northern Argentina has the highest VLH occurrence, followed by Uruguay, Paraguay, and southern Brazil. The USA Great Plains and South Africa also exhibit high VLH frequency while Europe, Australia, and especially Asia experience lower activity. VLH occurrence increased the most across Europe, being strongly correlated with rising temperatures and primarily driven by an increase in low-level moisture and instability. Significant decreases in VLH frequency are limited to the Southern Hemisphere and occurred most strongly across South America as a result of decreasing mid-level humidity and instability. Hail-related losses have risen in the USA, Australia and Europe. The more frequent occurrence of VLH contributes to this increase in Europe, while in the USA and Australia, socio-economic factors are mainly responsible for the rise.\u003c/p\u003e","manuscriptTitle":"Very Large Hail in a Warming Climate: Climatology, Trends and Losses","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-30 07:20:07","doi":"10.21203/rs.3.rs-6196143/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-geoscience","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"ngeo","sideBox":"Learn more about [Nature Geoscience](http://www.nature.com/ngeo/)","snPcode":"","submissionUrl":"","title":"Nature Geoscience","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"c92ec04c-4542-4068-894c-e0331e8d163f","owner":[],"postedDate":"April 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47858404,"name":"Earth and environmental sciences/Natural hazards"},{"id":47858405,"name":"Earth and environmental sciences/Climate sciences/Climate change"}],"tags":[],"updatedAt":"2025-12-30T08:05:12+00:00","versionOfRecord":{"articleIdentity":"rs-6196143","link":"https://doi.org/10.1038/s41561-025-01868-0","journal":{"identity":"nature-geoscience","isVorOnly":false,"title":"Nature Geoscience"},"publishedOn":"2025-12-29 05:00:00","publishedOnDateReadable":"December 29th, 2025"},"versionCreatedAt":"2025-04-30 07:20:07","video":"","vorDoi":"10.1038/s41561-025-01868-0","vorDoiUrl":"https://doi.org/10.1038/s41561-025-01868-0","workflowStages":[]},"version":"v1","identity":"rs-6196143","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6196143","identity":"rs-6196143","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}