Abstract. Remote hail detection and hail size estimation using weather radar observations has the advantage of wide spatial coverage and high spatial and temporal resolution. Switzerland National Weather Service (MeteoSwiss) uses two radar-based hail metrics: the probability of hail at the ground (POH) to assess the presence of hail, and the maximum expected severe hailstone size (MESHS) to estimate the largest hailstone diameter. However, radar-based metrics are not direct measurements of hail and have to be calibrated with and verified against ground-based observations of hail, such as crowdsourced hail reports. Switzerland benefits from a particularly rich and dense dataset of crowdsourced hail reports from the MeteoSwiss app. We combine a new spatiotemporal clustering method (ST-DBSCAN) with radar reflectivity to filter the reports and use the filtered reports to verify POH and MESHS in terms of the Hit Rate, False Alarms Ratio (FAR), Critical Success Index (CSI), and Heidke Skill Score (HSS). Using a 4 km × 4 km maximum upscaling approach, we find FAR values between 0.3 and 0.7 for POH and FAR > 0.6 for MESHS. For POH, the highest CSI (0.37) and HSS (0.52) are obtained for a 60 % threshold, while for MESHS the highest CSI (0.25) and HSS (0.4) are obtained for a 2 cm threshold. We find that the current calibration of POH does not correspond to a probability and suggest a recalibration based on the filtered reports.

Remote hail detection and hail size estimation using weather radar observations has the advantage of wide spatial coverage and high spatial and temporal resolution. Switzerland's National Weather Service (MeteoSwiss) uses two radar-based hail metrics: the probability of hail on the ground (POH) to assess the presence of hail and the maximum expected severe hailstone size (MESHS) to estimate the largest hailstone diameter. However, radar-based metrics are not direct measurements of hail and have to be calibrated with and verified against ground-based observations of hail, such as crowdsourced hail reports. Switzerland benefits from a particularly rich and dense dataset of crowdsourced hail reports from the MeteoSwiss app. We combine a new spatiotemporal clustering method (Density-Based Spatial Clustering of Applications with Noise, ST-DBSCAN) with radar reflectivity to filter the reports and use the filtered reports to verify POH and MESHS in terms of the hit rate, false-alarm ratio (FAR), critical success index (CSI), and Heidke skill score (HSS). Using a 4 km × 4 km maximum upscaling approach, we find FAR values between 0.3 and 0.7 for POH and FAR > 0.6 for MESHS. For POH, the highest CSI (0.37) and HSS (0.52) are obtained using a 60 % threshold, while for MESHS the highest CSI (0.25) and HSS (0.4) are obtained using a 2 cm threshold. We find that the current calibration of POH does not correspond to a probability and suggest a recalibration based on the filtered reports.

Hail is a major threat associated with severe thunderstorms, and estimating the hail size is important for issuing warnings to the public. For the validation of existing operational, radar-derived hail estimates, ground-based observations are necessary. Automatic hail sensors, for example within the Swiss Hail Network, record the kinetic energy of hailstones to estimate the hail sizes. Due to the small size of the observational area of these sensors (0.2 m2), the full hail size distribution (HSD) cannot be retrieved. To address this issue, we apply a state-of-the-art custom trained deep learning object detection model to drone-based aerial photogrammetric data to identify hailstones and estimate the HSD. Photogrammetric data of hail on the ground were collected for one supercell thunderstorm crossing central Switzerland from southwest to northeast in the afternoon of 20 June 2021. The hail swath of this intense right-moving supercell was intercepted a few minutes after the passage at a soccer field near Entlebuch (canton of Lucerne, Switzerland) and aerial images were taken by a commercial DJI drone, equipped with a 45-megapixel full-frame camera system. The resulting images have a ground sampling distance (GSD) of 1.5 mm per pixel, defined by the focal length of 35 mm of the camera and a flight altitude of 12 m above the ground. A 2-dimensional orthomosaic model of the survey area (750.4 m2) is created based on 116 captured images during the first drone mapping flight. Hail is then detected using a region-based convolutional neural network (Mask R-CNN). We first characterize the hail sizes based on the individual hail segmentation masks resulting from the model detections and investigate the performance using manual hail annotations by experts to generate validation and test data sets. The final HSD, composed of 18 207 hailstones, is compared with nearby automatic hail sensor observations, the operational weather-radar-based hail product MESHS (Maximum Expected Severe Hail Size) and crowdsourced hail reports. Based on the retrieved data set, a statistical assessment of sampling errors of hail sensors is carried out. Furthermore, five repetitions of the drone-based photogrammetry mission within 18.65 min facilitate investigations into the hail-melting process on the ground.

Measurements of hailstone diameters and kinetic energy, collected by the Swiss network of automatic hail sensors, are available in three regions of Switzerland for the period between September 2018 and August 2023. In this study, we propose the use of double-moment normalization for modeling the hail size number distribution (HSND), which is defined as the number of hailstone impacts measured, for each diameter size, by one instrument during one hail event. This method uses two of the empirical moments of the HSND to compute a normalized distribution. While the HSND is dependent on the duration and intensity of the event and on the detection area of the sensor, we show that the normalized distribution has limited variability across the three geographical regions of deployment of the sensors. Thanks to its invariance in space and time, a generalized gamma function is used to model the normalized distribution, and its parameters have been determined through a fit over approximately 70 % of the events. The fitted model and the previously chosen pair of empirical moments can be used to reconstruct the HSND at any location in Switzerland. The accuracy of the reconstruction has been estimated over the remaining 30 % of the dataset. An additional evaluation has been performed on an independent HSND, made of estimates of hail diameters measured by drone photogrammetry during a single event. This HSND has a much larger number of hailstone impacts (18 000) than those of the hail sensor events (from 30 to 400). The double-moment normalization is able to reproduce well the HSND recorded by the hail sensors and the drone, albeit with an underestimation of the number of impacts at small diameters. These results highlight the invariance of the normalized distribution and the adaptability of the method to different data sources.

From June 18 to July 31, 2021, a series of exceptional hailstorms occurred over Switzerland, causing major damages to buildings, cars, and crop fields. The available estimates from the insurance companies suggest that these events will be among the most expensive of the last decades. At the same time the events provide a unique research opportunity as the hailstorms were well captured by various observing systems: a newly set-up network of automatic hail sensors that report the size and kinetic energy of individual hail stones with very high temporal and size resolution, the crowdsourcing function of the MeteoSwiss app, and two radar-based operational hail products. The recently established radar-based Swiss hail climatology shows that the events of 2021 were extreme with high return periods both in terms of the reported hail stone sizes and in their spatial extent. Using the data captured by those complementary hail-dedicated observing systems, we review the hail activity in Switzerland during the period of interest and investigate two particularly intense hail days: June 28 (HD1) and July 8 (HD2). On HD1, the storms originated in western Switzerland, moved along the northern flank of the Swiss Alps in a Southwest to Northeast motion, and one storm evolved in a mesoscale convective system. On HD2, the storms originated in Northern Italy and moved over Southern Switzerland (Ticino) in a South to North motion. We look at the synoptic-scale situation, mesoscale environment, and storm tracks of HD1 and HD2 in details and demonstrate their exceptional character with respect to the climatology. We touch upon the new research avenues opened by the automatic hail sensors measurements both individually, as they allow to capture the time evolution of the hail stones size distribution, and in combination with the crowdsourcing and radar data (cross-validation of the radar-based hail algorithms).

Abstract Global Navigation Satellite Systems (GNSS) are not only a state‐of‐the‐art sensor for positioning and navigation applications but also a valuable tool for remote sensing. Through the usage of L‐band carrier frequencies, GNSS acts as an all‐weather‐operation system, offering substantial benefits compared to optical systems. Nevertheless, severe weather can still have an impact on the strength of signals received at a ground station, as we show in this study. We investigate GNSS Signal‐to‐Noise Ratio (SNR) observations during two severe convective storm events over the city of Zurich, Switzerland. We make use of a GNSS‐SNR‐based algorithm originally developed for the detection of hail particles from volcanic eruptions. Results indicate that, although GNSS observations are considered to be fairly insensitive to the presence of hydrometeors, convective storm events are visible in SNR observations. SNR levels of affected satellites show a significant drop during event periods, which are determined by weather radar observations.

Measuring the properties of hailstorms is a difficult task due to the rarity and mainly small spatial extent of the events. Especially, hail observations from ground-based time-recording instruments are scarce. We present the first study of extended field observations made by a network of 80 automatic hail sensors from Switzerland. The main benefits of the sensors are the live recording of the hailstone kinetic energy and the precise timing of the impacts. Its potential limitations include a diameter-dependent dead time, which results in less than 5 % of missed impacts, and the possible recording of impacts that are not due to hail, which can be filtered using a radar reflectivity filter. We assess the robustness of the sensors' measurements by doing a statistical comparison of the sensor observations with hailpad observations, and we show that, despite their different measurement approaches, both devices measure the same hail size distributions. We then use the timing information to measure the local duration of hail events, the cumulative time distribution of impacts, and the time of the largest hailstone during a hail event. We find that 75 % of local hailfalls last just a few minutes (from less than 4.4 min to less than 7.7 min, depending on a parameter to delineate the events) and that 75 % of the impacts occur in less than 3.3 min to less than 4.7 min. This time distribution suggests that most hailstones, including the largest, fall during a first phase of high hailstone density, while a few remaining and smaller hailstones fall in a second low-density phase.

Global Navigation Satellite Systems (GNSS) have become a valuable tool for remote sensing, as signals can be used for monitoring soil and snow properties as well as water vapor in the atmosphere. By using L-band carrier frequencies, GNSS acts as an all-weather-operation system. Nevertheless, severe weather can still have an impact on the strength of signals received at a ground station, as we show in this study. We investigate Signal-to-Noise Ratio (SNR) from the Global Positioning System (GPS) during two thunderstorm events, which produced excessive amounts of rain and hail. We make use of a GPS-SNR-based algorithm, developed for the detection of hail particles from volcanic eruptions. Results indicate that the investigated thunderstorm events are visible in SNR observations. Affected satellites show a significant SNR drop during event periods, which are determined by weather radar observations. Thus, results suggest the possibility of detecting severe weather systems using GNSS-SNR observations.

We present an investigation of the first observations from the Swiss Hail Network Project, a network of 80 fully automatic hail sensors (Kopp et al., 2022) installed between 2018 and 2020 in the three most hail-prone regions of Switzerland: the Jura, the Napf and Ticino (NCCS 2021). Those sensors provide new hail ground-based information, not only about each hailstone size (estimated by an indirect measures of its kinetic energy), but also about the precise timing of hailstone individual impact (Löffler-Mang et al., 2011), allowing to obtain a time-resolved hail size distribution (HSD). More specifically, we investigate the point (local) duration of hailfalls, the event hit rate (impacts per second) and time-resolved HSD. We also present and discuss the potential sources of uncertainty specific to the hail sensor, such as the dead time (minimum time between two consecutive observations). We then compare our observations to measurements from an hailpads network in northeastern Italy (Manzato et al., 2022). While our sample is still limited (around 10’000 hailstone impacts registered during 4 warm seasons) with respect to the hailpads records (29 warm seasons), we found that the HSD obtained with both measurements’ devices are very close to each other. Finally, we discuss the further combination of the sensor data with radar hail products (Kopp et al., 2022), the exceptionally high density of crowdsourced hail reports collected in Switzerland (Kopp et al., 2022) and the recent drone measurements of hail (Martin Lainer et al., abstract ECSS2023-5), which could pave the way to new and exciting research avenues on hail understanding and forecasting. Kopp, J., Schröer, K., Schwierz, C., Hering, A., Germann, U. and Martius, O. (2022), The summer 2021 Switzerland hailstorms: weather situation, major impacts and unique observational data. Weather. https://doi.org/10.1002/wea.4306 Löffler-Mang, Martin, Dominik Schön, and Markus Landry. 2011. « Characteristics of a New Automatic Hail Recorder ». Atmospheric Research 100 (4): 439‑46. https://doi.org/10.1016/j.atmosres.2010.10.026. Manzato, Agostino, Andrea Cicogna, Massimo Centore, Paolo Battistutta, and Mauro Trevisan. 2022. « Hailstone Characteristics in NE Italy from 29 Years of Hailpad Data ». Journal of Applied Meteorology and Climatology, août. https://doi.org/10.1175/JAMC-D-21-0251.1. NCCS (2021) National Centre for Climate Services : Hail climatology Switzerland. https://www.nccs.admin.ch/nccs/en/home/the-nccs/priority-themes/hail-climate-switzerland.html, accessed 26 December 2022

There is a growing need for past weather and climate data to support science and decision-making. This paper describes the compilation and construction of a global multivariable (air temperature, pressure, precipitation sum, number of precipitation days) monthly instrumental climate database that encompasses a substantial body of the known early instrumental time series. The dataset contains series compiled from existing databases that start before 1890 (though continuing to the present) as well as a large amount of newly rescued data. All series underwent a quality control procedure and subdaily series were processed to monthly mean values. An inventory was compiled, and the collection was deduplicated based on coordinates and mutual correlations. The data are provided in a common format accompanied by the inventory. The collection totals 12452 meteorological records in 118 countries. The data can be used for climate reconstructions and analyses. It is the most comprehensive global monthly climate dataset for the preindustrial period so far.

Hail is a major threat connected to severe thunderstorms and an estimation of the hail size is important to issue warnings for the public. Radar real-time products exist that estimate the size of the expected hail. For the verification of such products, ground based observations are necessary. Automatic hail sensors, as available within the Swiss hail network, can provide information about hail diameters observed on the ground. Unfortunately, due to the small size of these sensors (e.g. 0.2 m2) the estimation of the hail size distribution (HSD) can have large uncertainties. To overcome this issue, aerial drone-based 2D orthophotos can be analyzed by using state-of-the-art custom trained AI-object detection models to identify hail stones in the images and to estimate the HSD. A large right moving supercell with a lifespan of more than 6 hours crossed the midlands of Switzerland from south west in the afternoon of 20th June 2021. The hail swath of this classical supercell was intercepted near Entlebuch and aerial images of the hail on the ground were taken by a DJI Matrice 300RTK drone immediately after the storm has passed. The drone was equipped with a 50 megapixels full frame camera. The average ground sampling distance is 1.5 mm per pixel, which is set by the mounted camera objective with a focal length of 35 mm and a flight altitude of 12 m above ground level. A 2D orthomosaic model of the survey area (soccer field) is created based on 116 captured images during the first drone mapping flight. The orthomosaic covers an area of about 750 m2 and is then used to detect hail by using a region-based Convolutional Neural Network (Mask R-CNN) model. First, we characterize the hail sizes based on the individual hail segmentation masks resulting from the model detections and investigate the performance with respect to manual hail annotations from experts that are used as validation and test data sets. We present the final obtained HSD from more than 18000 hail stones (Dmax = 39 mm, Dmed = 9 mm) and compare it with nearby automatic hail sensor observations and weather radar based hail products like MESHS (Maximum Expected Severe Hail Size). Furthermore, we provide first insights into hail melting processes that can be inferred from the information retrieved from a total of 5 subsequent flights performed with the drone within about 20 minutes after the passage of the supercell.

Extreme precipitation events that occur in close succession can have important societal and economic repercussions. Here we use 42 years of reanalysis data (ERA-5) to investigate the link between Euro-Atlantic large-scale pattern of weather and climate variability and the temporal clustering of extreme rainfall events over Europe. We implicitly model the seasonal rate of extreme occurrences as part of a Poisson General Additive Model (GAM) using cyclic regression cubic splines. The smoothed seasonal rate of extreme rainfall occurrences is used to (i) infer the frequency of significant temporal clustering and (ii) implicitly serves as the baseline rate when modeling the effects of atmospheric drivers on extreme rainfall clustering. We use GAMs to model the association between the temporal clustering of extreme rainfall events and seven predominant year-round weather regimes in the Euro-Atlantic sector as well as a measure of synoptic-scale transient recurrent Rossby wave packets. Sub-seasonal clustering of precipitation events is significant at all grid-points over Europe; the proportion of extreme rainfall events that cluster in time ranges between 2% to 27%. The most relevant weather regime is the Atlantic Trough (corresponding to NAO+ with a southward shift of the jet) explaining most of the significant increase in clustering probability over Europe. The Greenland Blocking regime explains most of the clustering over the Iberian Peninsula. The Scandinavian Blocking regime is associated with a significant increase in clustering probability over the western Mediterranean, with a northwards shift in the signal to central Europe in summer.

Temporal clustering of extreme precipitation events on subseasonal time scales is a type of compound event, which can cause large precipitation accumulations and lead to floods. We present a novel count-based procedure to identify subseasonal clustering of extreme precipitation events. Furthermore, we introduce two metrics to characterise the frequency of subseasonal clustering episodes and their relevance for large precipitation accumulations. The advantage of this approach is that it does not require the investigated variable (here precipitation) to satisfy any specific statistical properties. Applying this methodology to the ERA5 reanalysis data set, we identify regions where subseasonal clustering of annual high precipitation percentiles occurs frequently and contributes substantially to large precipitation accumulations. Those regions are the east and northeast of the Asian continent (north of Yellow Sea, in the Chinese provinces of Hebei, Jilin and Liaoning; North and South Korea; Siberia and east of Mongolia), central Canada and south of California, Afghanistan, Pakistan, the southeast of the Iberian Peninsula, and the north of Argentina and south of Bolivia. Our method is robust with respect to the parameters used to define the extreme events (the percentile threshold and the run length) and the length of the subseasonal time window (here 2 – 4 weeks). The procedure could also be used to identify temporal clustering of other variables (e.g. heat waves) and can be applied on different time scales (e.g. for drought years). For a complementary study on the subseasonal clustering of European extreme precipitation events and its relationship to large-scale atmospheric drivers, please refer to Barton et al.