The Liguria region, characterized by complex orography and proximity to the sea, every year experiences intense meteo-hydrological events, such as flash floods, severe thunderstorms, hail, downbursts and strong winds. These events are generally localized and typically develop in correspondence with the convergence between the southerly flow over the Ligurian Sea with the northerly cold outbreak from the Po valley across the gaps in the orography (Buzzi et al., 2014; Fiori et al., 2017). They can cause significant socio-economic impacts, occasionally even casualties, while their predictability is still quite limited. The available observations from the Regional Agency for the Environmental Protection of Liguria region (ARPAL), including a dense network of weather stations, enriched by open hardware ACRONETWORK stations (Loglisci et al., 2024), operational radars, and continuous satellite monitoring, combined with numerical simulation analysis, have highlighted the key role of orography, sea-breeze interactions and SST patterns in enhancing convective activity (e.g., Cassola et al., 2016; Meroni et al., 2018), but the understanding of small-scale processes remains limited. In particular situations to be further investigated, storms originating from mesoscale convergence over the Ligurian Sea or northern Tyrrhenian Sea can move further inland in the Tuscan hinterland and, more rarely, even downwind the main orographic barrier, affecting the Emilia-Romagna region, as happened on 13–14 September 2015 during the devastating flash floods in Val Trebbia and Val di Nure (Segadelli et al., 2020).

In the Veneto and Friuli Venezia Giulia (FVG) regions, a series of particularly intense phenomena, such as supercells and intense convective systems, have occurred in the last decades. Remarkable examples are shown in Fig. 3: the F4 (category 4 of the Fujita scale) Mira-Dolo tornado of July 2015 (Zanini et al., 2017); the violent downburst in Verona in August 2020; the recurrent summer hailstorms, which peaked in July 2023, when the European record for hailstone size was broken (De Martin et al., 2025b). Other severe weather events were directly or indirectly related to cyclonic activity, such as the famed Vaia storm of 2018 (Davolio et al., 2020; Sioni et al., 2023), the impressive “acqua alta” in Venice on 12 November 2019 (the second highest level in the historical record; Ferrarin et al., 2021; Miglietta et al., 2023), the intermediate-duration floods (12–24 h long) over the pre-Alpine area near Vicenza that tested the efficiency of the recently implemented operational detention basins, and the back-building storms responsible for flash floods in eastern FVG in November 2025. A statistical study by the Regional Agency for the Environmental Protection of Veneto region (ARPAV) has identified a marked increase in high precipitation threshold exceedances during the last 30 years for what concerns the short and very short accumulation timescales (5 min–12 h), while no significant trend was found in extreme precipitations in FVG (Manzato et al., 2025). The increased interest in intense events has brought ARPAV to maintain a register of intense rainfall events and tornadoes in the region, as part of its weather forecasting practice. ARPAV continuously monitors these episodes with its ground-based high-resolution observational network, including over 300 hydro-meteorological stations, 2 C-band radars, 8 disdrometers, a temperature and humidity profiler, and a historical archive of images from installed webcams.

Overall, the Alpine region appears to exhibit increased hydrogeological sensitivity, emerging during several heavy downpours during the summer season in the form of recurrent debris flows. In fact, the increased frequency and intensity of extreme weather events (e.g., Dallan et al., 2022) resulted in a rise in hydrogeological risk and in the vulnerability of ecosystems and infrastructure. These phenomena, often characterized by strong spatial and temporal variability and very difficult to predict, are particularly impactful in urban areas (Gambini et al., 2025). For example, the Milan hydraulic node represents an emblematic case, where the Seveso, Olona, and Lambro rivers cause recurrent inundations in the metropolitan area, leading to widespread damage to infrastructure and urban mobility (Ceppi et al., 2022). Furthermore, wide urban areas (such as Milan) may alter the dynamics of severe storms, by intensifying updrafts and shifting storm tracks (De Martin et al., 2025c), thus adding an additional level of complexity.

While northern Italy has been the subject of several studies and some conceptual models are now available to identify conditions conducive to severe weather, the central and southern Apennine range remains largely under-researched, although they constitute a steep, coastal-proximal mountain range occasionally subject to heavy precipitation, strongly modulated in terms of timing and persistence by mesoscale circulations, land-sea contrasts and complex topography. In the Apennine setting, modest but steep relief modulates convective initiation by: (i) channeling low-level flow through gaps and valleys, (ii) forcing ascent on the windward flanks, and (iii) focusing near-shore convergence where the marine boundary layer impinges on the coastal range. The efficacy of these pathways depends on the synoptic flow, boundary-layer stability, shear, and sea-land thermal contrasts, which together set diurnal phasing and storm organization. The available literature clearly documents how the interaction between coastal orography and marine inflows over Italy prolongs event duration and shifts precipitation peaks relative to the upstream forcing (see Miglietta and Davolio, 2022, for a review), leading to intense inland flooding (Rosso and Ceppi, 2023).

Across the Apennines, orographic precipitation encompasses both cold-season events, driven by moisture transported from the sea (stratiform or embedded convection in the presence of long-fetch advection) and warm-season events, characterized by convection initiated by the orography, including hail-producing storms. Capozzi et al. (2023a) identified a limited set of recurrent heavy-precipitation spatial patterns for Campania region (southern Italy) that are systematically associated with synoptic types and the regional orography; importantly, such patterns occur in autumn-winter under long-fetch maritime moisture transport impinging on the mountain ranges, highlighting the key role of moisture-fluxes in organizing and sustaining heavy rainfall events (Fig. 4).

Atmospheric rivers (ARs) have emerged as one of the most relevant global drivers for extreme hydro-meteorological events in many areas of the globe. Defined as narrow corridors of enhanced horizontal transport of low-level moisture, ARs typically form over midlatitude regions pulling up water vapor charge also from subtropics and feeding precipitation systems in the areas where they make landfall and are forced to raise over the orography. Whilst the science of ARs is well developed in the US, where they have been recognized as responsible for extreme precipitation especially along the Pacific coast, only in the last decade the role of ARs has become evident over Europe (Lavers and Villarini, 2015). Over the Mediterranean basin, ARs were found relevant for recent extreme rainfall and major floods over southern Spain (Lorente-Plazas et al., 2019), southern France (Francis et al., 2025), the Dinaric Alps (Martinković et al., 2017), the eastern basin (Bozkurt et al., 2019) and over northern and central Italy (Davolio et al., 2020; Davolio et al., 2023; Sioni et al., 2023).

Focusing on the Alpine chain southern slopes, the presence of moisture transport, which may originate from either remote or local sources, is critical for the development of orographic rainfall and convection. Heavy precipitation often develops eastward of an advancing upper-level disturbance, a trough or a cyclone deepening over the basin. The latter is responsible for steering warm and moist low-level (unstable) air from the sea towards the steep slopes of the Alps and the Apennines (Miglietta and Davolio, 2022). At the mesoscale, precipitation systems are likely triggered by the direct orographic uplift or by the deflection or blocking action exerted by the orography (Miglietta and Buzzi, 2004), which can also renew convection triggering at the same location.

The role of ARs in heavy precipitation has been recently disclosed in a climatological framework over northern Italy (Davolio et al., 2026), as well as in case study analyses (Davolio et al., 2020; Sioni et al., 2023), providing evidence of a strict connection between extreme organized water vapor transport and extreme events affecting both the Alps and the Apennines (Grazzini et al., 2020). Conversely, it is well known that the thermodynamic characteristics of the impinging moist flow are fundamental to determine the interaction with the orography (Miglietta and Rotunno, 2014, among others), thus the intensity and distribution of rainfall. In this context, ARs have been so far described as characterized by moist-neutral stratification in the lowest layers (∼3 km) of the atmosphere, as observed from dropsondes released during many field campaigns over the Pacific (Ralph et al., 2020). This weak stability makes AR prone to rise over mountain barriers, producing orographic precipitation where convection is likely embedded, which can greatly amplify the amounts of precipitation.

Italy is among the European countries with the highest tornado frequency and intensity. Most Italian tornadoes fall within the categories of damage F0 and F2 of the Fujita scale, but intense cases (F3, F4 or even F5) have also been reported, as documented by the European Severe Weather Database (ESWD; Dotzek et al., 2009). Some climatological studies (Miglietta and Matsangouras, 2018; Avolio and Miglietta, 2022) have shown that the Italian areas most prone to tornado development include some southern areas (Ionian regions and Sicily), the Po Valley (in particular, its eastern part), and the Tyrrhenian coasts (Fig. 5). These events occur predominantly in autumn (in the southern regions) and summer (in the northern regions).

Some recent studies have improved our understanding of tornadic environments in Italy. The conditions typically identified include the presence of strong wind shear (especially in the low-level) and significant Convective Available Potential Energy (CAPE) (Ingrosso et al., 2020). In southern Italy, SST anomalies also play a key role (Miglietta et al., 2017a; Bagaglini et al., 2021; Avolio and Miglietta, 2021) as well as the convective initiation by the orography (Miglietta et al., 2017b). More recently, a conceptual model based on the proximity of three air masses with different characteristics has been proposed for the Po valley (De Martin et al., 2024), showing some similarities with the US Midwest tornadic environments, although at a much smaller spatial scale. Due to its small-scale, this feature was not identified earlier in reanalysis studies. A detailed analysis of a recent IF3 (category 3 of the International Fujita scale) tornado in the same area reaffirmed the applicability of this conceptual model (De Martin et al., 2025d).

Significant field campaigns in Italy, focused on observations of high-resolution surface wind fields during thunderstorms, have been conducted using a wind monitoring network, which integrates ultrasonic anemometers, three LiDAR profilers, and one scanning Lidar, installed in the commercial ports of the Ligurian Sea (Solari et al., 2012; Repetto et al., 2018). This system has captured numerous thunderstorm-related wind events over the past few decades (e.g., Burlando et al., 2017, 2018, 2020; Canepa et al., 2020; Solari et al., 2015; Zhang et al., 2018). Canepa et al. (2024) included a database of twenty-nine thunderstorm wind records acquired by ultrasonic anemometers within the monitoring network. Several of these events were simultaneously captured by multiple instruments located within the same area, enabling spatiotemporal cross-correlation analyses and providing a depiction of different thunderstorm outflows. This level of spatial and temporal resolution makes the dataset unique in the national context.

In terms of direct measurements of thunderstorm outflows and their potentially destructive impacts on the ground, full-scale wind observations remain scarce due to the limited spatial and temporal extent of the phenomenon. When available, high-sampling anemometric records capture only a single, event-specific realization, shaped by both the intrinsic characteristics of the downburst (e.g., size, intensity, duration, and interaction with background flow) and the relative position of the measuring instrument with respect to the evolving phenomenon. Consequently, field data seldom provide a complete description of the spatial flow structure or of its temporal evolution, particularly when the parent storm moves relative to the sensor. In-situ measurements alone do not permit a quantitative reconstruction of the interplay between the various flow components that collectively define the downburst system. As a result, much of the research on downburst dynamics relies on laboratory-scale physical modelling and numerical simulations. Recent large-scale experimental campaigns have produced valuable datasets, enabling statistical analyses of the spatiotemporal evolution of downburst flows and their interaction with the near-surface PBL and the translating storm systems. Subsequent experiments incorporated thermodynamic effects by imposing temperature differentials between the downdraft and ambient air, highlighting the key role of thermal gradients in shaping the downburst structure and dynamics (Canepa et al., 2026). Nevertheless, the quantitative influence of thermal processes in natural conditions remains largely unresolved.

Aerosols are the seeds upon which cloud droplets and ice crystals form. Their ability to act as ice-nucleating particles (INPs) or cloud condensation nuclei (CCN) regulate microphysical processes, e.g. increasing the number of liquid droplets (Twomey, 1977) or initiating cloud glaciation (Murray et al., 2021). Aerosol particles play an important role in the climate by controlling the radiative budget of the atmosphere (aerosol-radiation interaction; ARI) and by directly modifying the microphysical properties of clouds (aerosol-cloud interaction; ACI).

In fact, variations in aerosol properties can dramatically affect convective storms as they may: modify droplet activation and warm-rain formation; shift the height and temperature of primary glaciation; influence secondary ice processes such as rime-splintering; determine the number and type of ice particles, graupel, and hail embryos; affect latent heating profiles, updraft strength, and storm intensity; control the rate and efficiency of precipitation development; modulate storm electrification and lightning activity.

Despite limited observational evidence, coarse-mode aerosol particles, particularly mineral dust, have recently been linked to hail formation through their ability to act as efficient ice-nucleating particles. Several studies reported positive associations between hail occurrence and enhanced Saharan dust loadings (Brennan and Wilhelm, 2025), which is extremely important for Italy, given its frequent exposure to Saharan dust outbreaks (Vogel et al., 2025).

Currently, within the research community there is considerable interest in hail over Italy for several reasons. While Italy has emerged as a hail hotspot in Europe over the last decade (e.g., Punge et al., 2017) and Northern Italy has shown the largest global increase in very-large hail frequency (up to +0.03 events per decade; Battaglioli et al., 2026), the most recent hail climatologies show remarkable differences across the country. Some studies indicate a frequency peak over the foothills of Northern Italy (e.g., Giordani et al., 2024; Cui, 2025; Manzato et al., 2022b), while others show a maximum over the sea (e.g., Laviola et al., 2022; Kahraman et al., 2024). Furthermore, there has been growing awareness of the substantial economic damage caused by hail to crops, buildings, and vehicles (Púčik et al., 2019; Panosetti and Tomassetti, 2024). A record of USD 6 billion loss has been experienced in Northern Italy in 2023 (Battaglioli et al., 2026).

Generally, hail is not explicitly forecast by regional and national weather services and is generally treated as one of several possible hazards associated with severe storms. While this approach is partly justified by the uncertainty in hail prediction, other weather services (e.g., NOAA in the United States) issue daily hail outlooks, suggesting that such forecasts are feasible and could be tested during the TIM campaign. Notably, OSMER-ARPA FVG mentioned in their forecast for 24 July 2023 the possible occurrence of severe hailstorms, a few hours before a record-breaking hail event occurred.

Despite the progress made in the model resolution and physical parameterizations, the forecast of convective events is challenging for numerical weather prediction (NWP) models because of the multitude of physical processes and scales involved (Hu and Franzke, 2020) and the intrinsic limited predictability of convection.

Current weather and climate models are progressing towards finer spatiotemporal scales, which is largely due to the increased computational capabilities that have become available. Improved model performance can be also achieved by steering the model using observations, i.e. applying data assimilation (DA). Several operational systems around the world use convection permitting DA (e.g. Ballard et al., 2016), that is regional-scale DA with a horizontal grid spacing of approximately 1–4 km, where convection is explicitly resolved rather than parametrized. These systems provide improved nowcasts and short-term (0–24 h) forecasts particularly suitable for convective storms (Milan et al., 2020).

Several studies have been conducted over Italy focusing on the assimilation of different data sources, often at convection resolving resolution, as radar (Gastaldo et al., 2021; Maiello et al., 2017; Lagasio et al., 2019b; Mazzarella et al., 2020), GNSS (Lagasio et al., 2019a; Torcasio et al., 2023; Faccani et al., 2005), lightning (Federico et al., 2024, 2026), weather stations (Maggioni et al., 2023), rainrate (Davolio et al., 2017; Torcasio et al., 2024), among others. These studies used different DA techniques, namely 3DVar, 4DVar, nudging, and Ensemble Kalman filter applied to different models used in Italy (e.g., WRF, MOLOCH, COSMO).

Improving the prediction of localized convective storms requires a better understanding of the local conditions leading to convective initiation. Although some recurrent patterns for heavy rain have been recently identified in Italy (e.g., Capozzi et al., 2023b; Grazzini et al., 2020, 2021; Iacomino et al., 2025), and the role of equilibrium and non-equilibrium conditions has been analyzed (Molini et al., 2011), the characteristics of the synoptic and mesoscale conditions associated with other severe storms should be clarified, identifying the observational/model-based precursors for intense convection in the different regions. This investigation will allow improving the capacity of forecasting severe thunderstorms and, combined with observational analysis, should help identify the typical trajectories followed by the storms in their evolution.

Within the framework of orographic convection in the Alpine areas, the experience gained during the TEAMx–DECIPHER campaign has emphasized the need to better understand the PBL mechanisms leading to convective initiation and, in particular, any mechanism that may produce vertical motion sufficient to lift equivalent potential temperature high-θe air mass to the level of free convection. In this framework, Manzato et al. (2022a) provided the first climatology of convective initiation over northern Italy based on lightning strikes. The campaign stimulated several scientific questions, both methodological and operational: which dynamic and thermodynamic preconditions govern the initiation and lifecycle of orographic cumulus clouds? which atmospheric and terrain factors govern their location and persistence? how do local slope/valley circulations, atmospheric stability, and direction of the synoptic flow interact to determine the stationarity or propagation of convective clouds?

Different initiation mechanisms may favor the development of convective systems, such as convergence induced by local breeze systems, land-coast gradient boundaries, sunrise-sunset induced thermal gradients, moisture gradients due to advection (e.g., from the Adriatic Sea). In addition, cold-pools, bores and gravity waves may trigger new convective events (the so-called “secondary convection”). However, it is difficult to discriminate which initiation mechanism is effective. This issue is related to a key question: why on days with similar instability and “weather-type” conditions, do severe storms sometimes occur and sometimes not? This point should be analyzed for heavy rain events both in the convective season (summer), in the period characterized by “strong flux of rain” (autumn), and in the transition between the two regimes (Manzato, 2007a; Davolio et al., 2016).

In the past, Manzato (2007b) and Manzato et al. (2019) developed a neural network–based scheme to perform a multivariate analysis for forecasting the occurrence and intensity of rainfall events using multiple sounding-derived indices in FVG. Building on this work, the TIM campaign aims also to better understand whether a relationship between instability indices and severe precipitation that can lead to flash flood episodes may exist. Given the limitations identified in this kind of studies, new aspects should probably be incorporated in research, in particular the role of aerosol concentration and of warm rain in enhancing storm dynamics and rain efficiency. At a later stage of development, another issue is the transition of convective storms from mountains to plains, which seems neither well understood from a storm-dynamics perspective nor generally well predicted by numerical models (Fischer et al., 2025).

The presence of the Mediterranean Sea introduces an additional degree of complexity, making the prediction of convective initiation and evolution particularly challenging, especially in the case of stationary or back-building convective systems, as identified in some heavy rain events in the Liguria region (Cassola et al., 2016). A better understanding of the role of SST anomalies has implications not only for a better detection of observed events, but also in terms of reduction of false alarms.

Extreme convective storms often form in environments characterized by strong horizontal and vertical aerosol gradients – for example, dust intrusions, biomass-burning plumes, polluted boundary layers, or marine aerosol inflow. The ability to detect these gradients over large areas is essential for interpreting storm initiation and spatial evolution; in fact, depending on their distribution within the atmosphere, aerosols may either invigorate or suppress convective development (Stier et al., 2024). Unfortunately, direct observations in convective environments remain scarce. Limited airborne measurements reveal substantial variability in aerosol concentrations and properties throughout the cloud column, suggesting that aerosol–cloud interactions are highly sensitive to local microphysical, dynamical and thermodynamic conditions (e.g. Zanatta et al., 2023).

Moreover, the understanding of microphysical processes leading to hail formation and extreme precipitation is still incomplete and leads to a limited consensus on the impacts of aerosol particles on precipitation, making microphysics parametrization in numerical models a critical issue (Stier et al., 2024). Also, the impact of ACI on cloud formation and precipitation is nonlinear (Zhao et al., 2024), leading to contrasting effects, i.e. invigorating or weakening microphysics effects at cloud base that accelerate or delay precipitation. Lastly, aerosol-mediated effects on hail development remain a potentially important but poorly constrained process.

Despite recent progress, wide observational gaps limit our knowledge of aerosol-cloud interaction and their influence on Mediterranean severe storms: do aerosols act as CCN and/or as INP? Do they promote the formation of rain and/or graupel, and if so, through which mechanisms and relationships? Recently, the role of aerosols was identified in modulating the precipitation in Mediterranean areas exposed to Saharian dust advection (e.g., Ferrari et al., 2024); however, the generality of these results is still uncertain and requires further investigation.

Within the Mediterranean, climate-change induced desertification and heatwaves modify the local concentration of aerosols (e.g. Cristofanelli et al., 2018; Vogel et al., 2025), hence the availability of CCN and INPs. Even though numerical modelling studies correlate these aerosol perturbations with changes in convective systems, the inability of reconciling observations and modelling data complicates the attribution of observed changes to aerosol-cloud interactions (Stier et al., 2024).

Along similar lines, Napoli et al. (2022) showed through cloud-permitting simulations over the Great Alpine Region (GAR) that the lifetime of clouds is generally increased at high levels of aerosols, while convective clouds can decrease at high levels of pollution as a result of the reduction in strong updrafts associated with an increased air column stability. All in all, it is crucial to study observations related to the 3D distribution of aerosol in connection with convective initiation (e.g., Brennan and Wilhelm, 2025, have identified for hail in Switzerland an “optimal concentration” of aerosol to maximize hail probability).

A central challenge is to disentangle aerosol-driven microphysical effects from dynamic forcing. Accordingly, open questions are: what are the background aerosol states (composition, CCN/INP abundance, vertical structure) that do not promote storm intensification, and how do they differ from perturbations associated with Mediterranean severe convective weather? Under which dynamical and thermodynamic conditions (CAPE, shear, humidity, synoptic forcing) do Mediterranean aerosol perturbations transition from weakening to invigorating convective storms? How can in-situ observations be used to attribute observed storm variability to aerosol–cloud interactions rather than to meteorological dynamics alone?

The potential changes in storm frequency and intensity under future climate scenarios is an open question that is still subject of strong debate in the scientific community. While there are strong evidences that potential instability increased in recent years (Cavalleri et al., 2026), and will be enhanced in a warmer climate, the signal in rainfall observations from regional network is not so clear (e.g. Manzato et al., 2025, found no significant trend of precipitation in northeastern Italy during the convective season), emphasizing the need for the analysis of multi-decadal trends in different data to detect possible shifts in precipitation regimes (Mazzoglio et al., 2025). Conversely, attribution studies based on pseudo-global warming simulations have clearly identified the role of a warmer atmosphere and sea in enhancing the intensity of severe convective events in the Mediterranean (González-Alemán et al., 2023; Calvo-Sancho et al., 2026). The fact that the observed super-Clausius–Clapeyron scaling of extreme precipitation may be explained by a shift from stratiform to convective rain type (Da Silva and Haerter, 2025; Haslinger et al., 2025) opens new questions on the role of climate change in modifying the precipitation formation pathways (Hawkins et al., 2023). In this framework, and following recent evidence for the Apennine areas that the intensity and frequency of heavy precipitation events are increasing particularly during autumn (Capozzi et al., 2023b), it is worth investigating how this result applies to different types of precipitation (e.g., autumn-winter orographic rainfall episodes driven by long-fetch maritime moisture transport in the Apennines).

Some studies report a marked increase in large hail events in the Po Valley (Battaglioli et al., 2023), others over the Alps (Cui et al., 2025), while in the plain of FVG, located in northeastern Italy, no significant hail trends were observed from 29 years of data collected by a network of about 300 hailpad stations, except for hail size (Manzato et al., 2025). These inconsistencies may ultimately be related to the scarcity of systematic ground-based hail observations, a limitation that could only be addressed through an extensive and dense network of automated hailpads (e.g., Kopp et al., 2023) and disdrometers (Jameson et al., 2015).

While conceptual models for precipitation have been developed for the Alpine areas, for other regions specific conceptual models need to be developed. The open issues include: (i) better investigating the convective initiation along the Adriatic/Tyrrhenian flanks, discriminating between terrain-forced convergence versus synoptically-forced cases; (ii) identifying under which combinations of SST anomalies, stability and shear, orographic channeling most effectively trigger convection and favors hail growth in central Italy; (iii) assessing the transferability of hail proxies calibrated in other Mediterranean regions, such as the Alps and northeastern Italy (Manzato, 2012; Giordani et al., 2024), to the interior Apennine basins with different relief scales and coastal geometry.

ARs associated with severe organized convective storms, over or at some distance from the orography, have been rarely documented. For example, Moore et al. (2012) analyzed a heavy rainfall event that occurred over the Central US, where an AR fed two quasi-stationary mesoscale convective systems (MCSs), linearly organized and exhibiting back-building mechanisms. Recent events affecting the Mediterranean basin revealed the presence of intense convection associated with ARs, as in the case of the dramatic flood in Valencia in October 2024 or the record-breaking rainfall over Liguria and Piedmont in October 2020 (Davolio et al., 2023). Therefore, given the availability of various AR datasets, both global and Mediterranean-specific, it is timely and worthy to investigate whether, and to what extent, ARs are associated with severe convective storms that develop near the Alpine chain during different seasons.

Significant limitations persist in monitoring and predicting rapidly evolving storms, such as tornadoes or downbursts, hence raising key challenges. These include: (i) sparse and heterogeneous observational coverage, including radar gaps and non-uniform scanning strategies, that limit the monitoring of mesocyclones and small-scale storms; (ii) underreporting and inconsistent intensity classification; (iii) the intricate geomorphology of Italy and the Mediterranean, which can enhance or hinder small-scale circulations, including sea–land and orographic interactions; (iv) the absence of tornado-specific forecasting or warning systems in Italy, as discussed in Miglietta and Rotunno (2016).

While some progress has been made in identifying the tornadic environment (Bagaglini et al., 2021; De Martin et al., 2024), more attention should be paid to radar patterns associated with severe convection. A climatology and a systematic analysis of the environments conducive to downbursts is also missing, although they represent a significant cause of damage in some Italian regions (Burlando et al., 2018). Future advances in downburst understanding will require integrated observational networks combining high-frequency anemometry, Doppler Lidar and radar systems, and thermodynamic profilers capable of resolving temperature, humidity, and pressure gradients at fine spatial and temporal scales. The coupling of these measurements with numerical weather prediction (NWP) models and Computational Fluid Dynamics (CFD) simulations could enable real-time reconstruction and nowcasting of downburst structures.

Understanding of the physical processes leading to hail formation remains limited, particularly for giant hail events (Allen et al., 2020). In Italy, this issue is further complicated by complex topography, as recently demonstrated by De Martin et al. (2025b) for northeastern Italy, and Ricchi et al. (2023a, b) for a giant hail event in Central Italy, where the Apennine topography, together with SST anomalies, guided the low-level flow and focused coastal convergence, effectively preconditioning convective initiation. This uncertainty has important implications. For instance, why does large hail appear to be a rare phenomenon along the Ligurian and Tyrrhenian coasts, despite the very high frequency of severe thunderstorms? Conversely, why is large hail so common along the Alpine foothills? Uncertainty in the mechanisms leading to large hail production also affects hail forecasting. For example, how important are environmental kinematics with respect to thermodynamics?

Beyond direct hail observation and forecasting, substantial effort is currently devoted to estimating hail size using remote sensing, like radar data (Allen et al., 2020; Areggeret et al., 2025). An effective hail size detection algorithm, combined with a tracking algorithm, could represent a powerful tool for early warnings of severe hailstorms. Regional weather services in Emilia-Romagna and Piedmont are testing several algorithms for this purpose (Fornasiero et al., 2024), and the data collected during the TIM campaign could significantly support their calibration.

Further work is needed to clarify the spatial distribution of hailstorms over Italy. Extending hail climatologies to include economic hail damage – potentially involving insurance companies – should improve hail risk maps and lay the groundwork for impact-oriented hail warnings.

As stated, there are some studies over Italy addressing the problem of DA at the convective scale. However, very high-resolution simulations (i.e. with a grid spacing of less than 1 km) will be necessary to study local low-level circulations (e.g., up-valley and down-valley flows, mountain-valley breeze, …) that are relevant for the initiation of convection. Furthermore, the sensitivity of the model to parameterization schemes (e.g., convection, turbulence, microphysics) should clarify the role of convection-relevant and not yet fully studied physical processes, such as the impact of aerosols on cloud microphysics. Therefore, there are many challenges requiring further work and the TIM campaign can help to cope with some of them: (i) quantifying the impact of different data sources to guide future observing network design for the prediction of convection; (b) quantifying the impact of a background correlation matrix, which is aware of the “error of the day”, compared to climatological approaches in order to better designing DA systems for the convective scale; (c) using different observations at high spatio-temporal resolution to consider explicitly the correlation among observations; (d) refining the use of currently available observations, and assimilating new observation types.

Satellite observations are limited by their relatively coarse space and time resolution; however, the integration of geostationary and polar data remain a privileged tool to explore and monitor storm evolution, from pre-storm conditions and convection initiation to the mature and decaying stages of the systems. In addition, flash observations from the Lightning Imager onboard MTG are very useful to identify, track and forecast convective storms.

New satellite techniques are available to identify and monitor hailstorms. The Multi-sensor Approach for Satellite Hail Advection (MASHA) is a new multi-instrument blended technique conceived for real-time detection of hail-bearing clouds. MASHA (Fig. 6) identifies hail clouds from satellite measurements and monitors the evolution of hail-bearing systems every 5 min by combining the strength of the MicroWave Cloud Classification-Hail (MWCC-H) method, which detects hail through the whole Global Precipitation Measurement (GPM) sensor constellation (Laviola et al., 2020a, b), with the high temporal rate of the Meteosat Rapid Scan Service (MSG-RSS). MASHA has been applied to various case studies demonstrating strong capability in identifying severe events and high robustness to detect small-scale hailstorms in mountain catchments. These results open new perspectives for investigating hailstorms and hydro-meteorological events in mountain areas, whereas traditional methodologies often miss events or underestimate their severity. Although further experiments are needed, MASHA's performance paves the way towards pseudo-operational applications in support to hailstorm nowcasting and assimilation in regional numerical weather predictions and appears mature for application during the TIM campaign.

Finally, the data provided by the new satellite constellations (e.g., Meteosat Third Generation, MTG), including vertical profiling, will represent a tremendous opportunity to better characterize the environment where convective storms develop and to track their evolution.

Investigating convective initiation requires high-temporal resolution observations of the vertical velocity and moisture distribution in the lowest part of the troposphere (first 2–3 km). Doppler radar wind data with overlapping scan domains (at least every 5 min) would be necessary to estimate the vertical velocity spatial distribution using the classical continuity equation method (Dual-Doppler Analysis, DDA) or more modern techniques (e.g. López-Carrillo and Raymond, 2011; Tsai et al., 2025). For northeastern Italy, these techniques can be applied to the available Loncon and Fossalon C-band Doppler radars. In the Emilia-Romagna region, an X-band radar will be soon installed and two C-band radar sites updated.

The Italian Civil Protection Department expressed its willingness to join the TIM campaign, being particularly interested in analyzing the ground impact of severe convective storms and in integrating the damage surveys with radar data. A mobile X-band radar (2 m high antenna) is available for the campaign but requires power supply stations in the target areas and some funds to assure the transport of the instrument. Also relevant for the campaign is a radar situated on mount Zoufplan at 2000 m height, at the border with Austria.

The recent installation of two X-band dual-polarization radars in the Lombardy Region (Desio and Flero) aims to strengthen the regional monitoring network and enhance nowcasting capabilities using machine learning models (Franch et al., 2025). In addition to supporting operational forecasts (Xu et al., 2025), these radar-based estimates will serve as inputs for hydrological models, whether physically based and spatially distributed (hybrid systems) or entirely AI-driven (full AI systems). Consequently, integrating AI-based radar nowcasting with hydrological models will represent a significant advancement in improving the effectiveness of operational warning systems and the dynamic management of structural mitigation measures, providing reliable and accurate short-term forecasts, particularly for summer convective events (Ceppi et al., 2025).

AEROLAB (AERosol mObile LABoratory) is a CNR-ISAC mobile platform designed for deployment in remote and non-instrumented areas for the study and characterization of atmospheric particulate matter through in situ sampling. AEROLAB is housed in an ISO10 container, certified for operation also on-board research vessels, temperature-controlled, and equipped with dedicated sampling lines for multiple instruments. The remote sensing operational observational capability of the mobile platform consists of a co-located Doppler wind lidar, a 35 GHz cloud radar, and a lidar, enabling an integrated characterization of convection-relevant processes from the surface layer to the cloud column.

Together, the suite links aerosol structure, inflow kinematics, and near-cloud vertical motions to the cloud and precipitation response, enabling process studies and model evaluation of triggering, microphysics–dynamics feedbacks, and aerosol-mediated convective intensification.

A possible correspondence between severe storms and ARs can be identified based on the exploitation of satellite analysis and ground-based observations. Being characterized by large-scale moisture transport, ARs can be identified using analysis fields and detection algorithms. However, to describe precisely the low-level thermodynamic profiles of the impinging moist flow and infer the precipitation characteristics, the availability of on-demand radio-soundings, dropsondes and vertical moisture/wind measurements, including weather drones, especially upstream of convection, is hoped for. Additional instruments, like optical lidar and powerful wind profilers, which have been extensively employed during the TEAMx campaign, should monitor the PBL at high-resolution and the low-level wind, including its vertical component.

A substantial upgrade of in-situ observations is also desirable, mainly in areas which have been excluded from the mainstream scientific interest, such as the Apennines. Dense, elevation-stratified raingauges with ≤10 min logging along coastal-to-ridge transects and windward crests may complement the available disdrometer network (Adirosi et al., 2023) to resolve drop size distributions, phase and fall speeds in mixed-phase and convective events. An extensive hail monitoring network is also desirable to document occurrence, size spectra and kinetic energy, ideally co-located with disdrometers for cross-validation.

The available ACI in-situ observations are rarely coordinated with large scale dynamic campaigns (Kunz et al., 2022). The Italian contribution to TIM aims to provide systematic observations of both aerosol properties and cloud microphysics at the WMO/GAW observatory of Monte Cimone (CNR-ISAC), a facility representative of the Mediterranean troposphere. However, without vertical profiles of wind, stability, humidity, and aerosol distribution (Fig. 7), it remains impossible to separate microphysical aerosol effects from dynamical forcing. Implementing remote-sensing and profiling capabilities in parallel to in-situ aerosol-microphysical observations during the TIM campaign is therefore desirable to enable robust attribution of aerosol impacts on Mediterranean severe storms.

Airborne platforms are the only way of providing vertically and horizontally resolved observations of aerosol properties, thermodynamic structure, and cloud development in the vicinity of convective systems. To address this need, we propose the integrated deployment of two available Italian research aircrafts: the SkyArrow ERA and the Piper Seneca III, in coordination with aerosol remote sensing (Lufft CHM – Nimbus 15k). Used in combination, the Cimone observatory and the two mobile platforms may characterize the environmental conditions from the surface to the mid-upper troposphere with complementary measurements.

Resolving aerosol concentrations in 3D is crucial considering that convective storms source their inflow differently, depending on their mode (supercell, MCS, etc.) and on the fact they are elevated or surface based. For supercells, different flanks of the storm even have different inflow origins, which has implications for hail growth (e.g., Lin and Kumjian, 2022). Near the orography, inflows might be even more complex and changing over time. Thus, monitoring of aerosols, moisture, and winds would be a unique and impactful observation set, yielding a 3D, multi-scale picture of the storm environment over a wide geographic domain, capturing horizontal gradients in aerosol concentration, shifts in aerosol types, thermodynamic stability, moisture and temperature gradients, all of which are critical for convective initiation and strongly modulate the spatial evolution and intensity of convective storms.

A multi-institutional, multi-platform field campaign entails practical and coordination challenges, which require rigorous operational frameworks, such as instrument interoperability, data-sharing protocols, real-time decision-making during IOPs, and the logistical complexity of deploying airborne platforms alongside ground networks. In fact, different institutions bring instruments with different hardware, data format, and calibration standards, which will require establishing compliance profiles and creating centralized preprocessing scripts to align all outputs, while centralising multiple observations requires the transition to federated, cloud-based data repositories that enforce interoperable formatting like netCDF for sharing. In addition, mission-planning strategies to rapidly evolving conditions will be required to ensure multi-aircraft coordination and to allow simultaneous data collection in the same atmospheric column from fixed-point ground networks and fast-moving airborne platforms.