Medicanes (Mediterranean hurricanes) are among the most hazardous high-impact weather systems affecting the Mediterranean basin, producing intense precipitation, severe winds, coastal flooding, and widespread socio-economic impacts in densely populated coastal regions. Recent advances in satellite Earth Observation are progressively transforming the monitoring of these tropical-like cyclones from retrospective classification toward near real-time physical characterization of their lifecycle and tropical-transition processes.
This work presents the near real-time multi-sensor analysis of cyclone Jolina (March 2026), one of the most recent observational examples of medicane, according to the new medicane definition recently included in the glossary of the American Meteorological Society (https://glossary.ametsoc.org/wiki/medicane/). According to this framework, a medicane is defined as a mesoscale Mediterranean cyclone exhibiting a warm core extending into the upper troposphere, spiral cloud bands, an eye-like structure, and a nearly symmetric surface wind circulation with maximum winds concentrated close to the storm center.
Cyclone Jolina originated as a baroclinic disturbance and progressively evolved toward a compact, diabatically driven warm-core system approaching the Libyan coast. Infrared and passive microwave satellite observations revealed the transition from an asymmetric cold-core cyclone to a tropical-like structure characterized by upper-tropospheric warm-core signatures, spiral cloud organization, and enhanced rotational symmetry. Particularly remarkable was the occurrence of this tropical transition during an atypical season and under relatively cold sea surface temperature conditions, highlighting the complex interplay between upper-level dynamics and diabatic processes in medicane development.
The analysis combines geostationary MSG-SEVIRI observations, passive microwave radiometry (e.g., ATMS, AMSU/MHS, as well as the newly available Arctic Weather Satellite and EPS-SG Microwave Sounder), scatterometer-derived ocean surface winds, and Synthetic Aperture Radar (SAR) measurements from Sentinel-1A. Remarkably, near real-time cyclone tracking is provided through the DeMeTra deep-learning algorithm, which estimates the medicane rotational center every 5 minutes using SEVIRI Airmass RGB imagery sequences. Passive microwave observations around the oxygen absorption complex near 55 GHz are exploited to identify upper-tropospheric warm-core anomalies, while humidity sounding channels around 183.31 GHz are used to detect signatures of stratospheric dry-air intrusion and associated potential vorticity anomalies contributing to cyclone intensification and warm-core development.
Beyond its meteorological significance, Jolina also demonstrated the substantial socio-economic relevance of Mediterranean tropical-like cyclones. Severe weather conditions triggered emergency measures across southern Italy and Libya, including school closures, transport disruption, flooding, landslides, evacuations, and even casualties. Interestingly, some of the strongest impacts occurred during the earlier stages of the cyclone, before the complete tropical transition occurred, emphasizing the importance of considering the whole lifecycle and both synoptic-scale interactions and medicane-scale processes in hazard assessment.
This case study demonstrates the increasing capability of integrated multi-sensor Earth Observation systems, combined with artificial intelligence approaches, to identify and monitor medicane formation and tropical transition in near-real time. The methodologies developed within the ESA MEDICANES project provide new opportunities for operational monitoring, early warning, and improved risk assessment of Mediterranean cyclones under changing climatic conditions.
How to cite: Panegrossi, G. and the ESA MEDICANES Project Team: Mediterranean Cyclone Jolina: Near Real-Time Detection of Medicane Formation Using Multi-Sensor Earth Observation, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-45, https://doi.org/10.5194/egusphere-plinius19-45, 2026.
Medicanes are rare Mediterranean tropical-like cyclones characterized by small spatial scales, rapid evolution, and in particular a warm core, a cloud free eye, and a closed ring of strong winds, leading to potentially severe coastal impacts, which require robust near-real-time detection and tracking from high-frequency geostationary satellite imagery. Their automatic detection and tracking remain challenging because labelled events are scarce, cyclone morphology is highly variable, and satellite-based signatures may be confused with other organized cloud systems.
This work presents DeMeTrA, a deep-learning framework for medicane detection and rotation center localization exploting MSG SEVIRI Rapid Scan Service (RSS) Airmass RGB image sequences. The system combines self-supervised spatiotemporal representation learning based on VideoMAE pretrained vision transformer model, with supervised downstream modules for cyclone presence estimation and center tracking. The architecture was developed through a two-scale detection and tracking strategy motivated by subsequent analyses of basin-scale inference. High-resolution video Transformer backbones provide powerful spatiotemporal representations, but their use in this context requires the Mediterranean domain to be processed through fixed-size small local crops compatible with the pretrained input geometry. This local formulation may limit the availability of full-domain spatial context during the initial detection stage. Under realistic Mediterranean conditions, where organized cloud systems, frontal structures, and peripheral spiral-like patterns can resemble medicane signatures, independent local decisions may lead to false alarms or to multiple candidate centers that are not physically consistent at basin scale. To address this limitation, DeMeTrA integrates a lightweight first-pass module operating on the full Mediterranean basin at reduced spatial resolution. This module estimates cyclone presence and provides a coarse cyclone-center location, which is then used to guide the high resolution VideoMAE tracking stage over a physically consistent region of interest. By separating basin-scale event screening from local center refinement, the framework preserves the advantages of pretrained VideoMAE representations while restoring the large-scale contextual information required for robust detection. This two-scale design limits false-alarm generation, avoids inconsistent candidate centers from independent spatial crops, and provides a single coherent center-track estimate for each detected event, supporting near-real-time medicane monitoring from geostationary infrared satellite observations.
Applications of DeMeTra to documented medicane cases will be shown to analysie its capabilities and skills during the storm development and mature phases. The methodology is currently being extended to the MTG Flixible Combined Imager FCI) for future applications.
Keywords: medicanes; MSG SEVIRI; deep learning; vision transformer; cyclone detection; cyclone tracking
How to cite: D'Armiento, D., Sebastianelli, S., D'Adderio, L. P., Sanò, P., Casella, D., and Panegrossi, G.: DeMeTrA: A Two-Stage Coarse-to-Fine Deep Learning Framework for Medicane Detection and Tracking from MSG SEVIRI Image Sequences, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-126, https://doi.org/10.5194/egusphere-plinius19-126, 2026.
The NASA TROPICS (Time-Resolved Observations of Precipitation structure and storm Intensity with a Constellation of Smallsats) Earth Venture (EVI-3) mission, was successfully launched into orbit on May 8 and May 25, 2023 (two CubeSats in each of the two launches into 550-km orbits with approximately 33-degree inclination, with swaths extending to almost 40 degrees North and South latitude, thus providing observations of the southern half of the Mediterranean region). Over the course of the mission, TROPICS provided more spaceborne microwave soundings than any operational program, and the combined forecast impact was larger and more spatially coherent than that of any individual passive microwave platform, illustrating the benefit of constellation-based temporal sampling for constraining rapidly evolving tropical convection. Prior to the deorbit of the last TROPICS spacecraft in December 2025, observations of 3-D temperature and humidity, as well as cloud ice and precipitation horizontal structure, at high temporal resolution were used to conduct high-value science investigations of tropical cyclones and other severe weather phenomena. TROPICS has provided rapid-refresh microwave measurements (median refresh rate of better than 60 minutes early in the mission with four functional CubeSats) in twelve channels spanning 92 to 205 GHz over the tropics that can be used to observe the thermodynamics of the troposphere and precipitation structure for storm systems at the mesoscale and synoptic scale over the entire storm lifecycle. Thousands of high-resolution images of tropical cyclones have been captured by the TROPICS mission, revealing detailed structure of the eyewall and surrounding rain bands. The new 205-GHz channel in particular (together with a traditional channel near 92 GHz) has provided new information on the inner storm structure, and, coupled with the relatively frequent revisit and low downlink latency, has informed tropical cyclone analysis at operational centers. The suite of TROPICS products is publicly available with much improved median revisit rates and were provided with data latencies that are sufficient to enable their use in operational tropical cyclone forecasting applications. In this presentation, we highlight the use of these high-revisit thermodynamic data from TROPICS to better characterize storm structure and environmental conditions over a variety of cases over the 30-month mission lifetime.
How to cite: Blackwell, W. and the TROPICS Science Team: New Capabilities for Observing Precipitation Intensity and Structure Provided by the NASA TROPICS Microwave Radiometer Constellation Mission, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-48, https://doi.org/10.5194/egusphere-plinius19-48, 2026.
Short-term prediction of severe hailstorms remains a major challenge, especially when forecasts must preserve the location, intensity, and internal structure of rapidly evolving high-reflectivity cores. Although Numerical Weather Prediction (NWP) models have improved synoptic-scale forecasting, physics-based models are limited at short lead times by model spin-up. This makes it harder to accurately capture rapidly evolving hail-producing storms. Useful hail nowcasting also requires guidance that can represent storm scale evolution with finer spatial and temporal fidelity. In this work, we present a radar-based AI nowcasting framework for short time range forecasting of intense convection over the Colorado-Wyoming region. The model is trained using composite radar reflectivity fields and is designed to generate high-resolution forecasts out to 1 to 3 hours.
To assess performance, the framework was evaluated on 15 severe hail events over the Colorado-Wyoming region using a combination of verification metrics. These include Probability of Detection (POD), False Alarm Ratio (FAR), Critical Success Index (CSI), Fractions Skill Score (FSS), and Structural Similarity Index Measure (SSIM), with a focus on the 40 dBZ reflectivity threshold as a proxy for hail-relevant storm intensity. The results show that the model provides high skill at shorter lead times and retains meaningful forecasts through 180 minutes. Comparisons against traditional extrapolation based nowcasting approaches and physics-based forecast models, including the High-Resolution Rapid Refresh (HRRR), indicate that the AI based framework is particularly effective at preserving storm structure, intensity, and spatial placement. Hail-producing storms are usually characterized by localized high-reflectivity maxima embedded within rapidly evolving convective morphology, so maintaining both intensity and spatial realism is essential for forecast usefulness. These findings suggest that radar-based AI driven nowcasting is a good tool for severe hail prediction.
How to cite: Chandrasekar, C. V., Biswas, S., and Radhakrishnan, C.: A Radar-Based AI Framework for Nowcasting Severe Hailstorms, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-34, https://doi.org/10.5194/egusphere-plinius19-34, 2026.
The Mediterranean region is characterised for its strong spatial and temporal variability in precipitation extremes. Such a variability, driven by complex atmospheric dynamics and intensified by climate change, results in a high exposure to natural hazards, particularly short-duration and high-intensity convective precipitation that usually generate flash floods in small to medium catchments. These phenomena pose significant challenges for precipitation estimation, monitoring, and risk management, especially in highly urbanised or topographically complex catchments.
In October 2024, the eastern Mediterranean coast of Spain was impacted by an extreme meteorological cut-off low associated with intense convective instability and exceptional rainfall accumulations up to 778 mm in 24 h and 79.4 mm in 1 h. The El Poyo Ravine catchment (385 km2) was the most affected area with severe damage to infrastructure, 238 fatalities, 4 500 buildings and 120 000 vehicles. It highlighted the urgent need for reliable methodologies to characterise convective precipitation and support early warning and preventive actions.
This research aims to identify and assess suitable spatial interpolation methods of precipitation for high localised convective storms, with a focus on improving precipitation characterisation for future forecast and operational applications. Based on a comprehensive literature review, eight interpolation methods were selected, including deterministic, geostatistical, and mathematical-formulation-based approaches. Quantitative validation was performed using metrics such as RMSE and the Nash-Sutcliffe Efficiency coefficient between observations and estimates.
The October 2024 flood event in the El Poyo Ravine catchment was selected as case study. Rainfall fields at each time step were generated by using 15-min observations at 12 rain-gauging stations of the real-time system (SAIH) of the Júcar River Basin Authority. Rainfall fields were generated with the eight interpolation methods of precipitation considered in the study.15‑min rainfall observations at 9 rain-gauging stations of the crowdsourced ECOWITT network were used for validation purposes, after undergoing a strict quality control process to discard either poor data or inconsistent stations. The analysed period spanned from 28 October 2024 at 07:00 to 30 October 2024 at 06:45 with a 15‑minute temporal resolution, focusing on the most intense convective phases of the event. Validation considered only time steps with available observations at both networks.
Results show that accuracy and computational cost strongly depend on the interpolation method. While the Inverse Distance Weighting (IDW) method provides results close to observations with low computational cost, local ordinary kriging requires a much higher computational cost to achieve comparable performance. In addition, the mathematical-equation-based method is more suitable for small catchments, as accuracy increases with decreasing catchment size.
Although this exploratory research focuses on a single catchment and one extreme event, the proposed methodology provides a transferable framework that can be applied to additional events and independent datasets. Therefore, this work can contribute to improving precipitation estimation strategies for Mediterranean catchments and supporting more effective hazard monitoring and risk management.
Acknowledgment: This research was supported by INECO (Ingeniería y Economía del Transporte S.M.E. M.P., S.A.) through funding provided under the project ‘Application of stochastic methods for flood assessment in urban areas’. Ángela Zaragoza-Paredes would like to thank the Fundación José
Entrecanales Ibarra for its financial support through a PhD research grant.
How to cite: Zaragoza-Paredes, Á. M., Mediero, L., Fernández-Fidalgo, F. J., and Lama-Pedrosa, B.: Comparison of precipitation interpolation methods applied to the October 2024 extreme rainfall event in Valencia (Spain), 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-18, https://doi.org/10.5194/egusphere-plinius19-18, 2026.
Rainfall detection and monitoring rely heavily on ground observation stations, meteorological radars, and microwave/infrared sensors onboard satellite platforms. However, the irregular distribution of observation stations, radar limitations due to ground clutter and topographic blockage, and the low spatial and temporal resolutions of polar-orbiting satellites—which carry the microwave sensors providing direct precipitation data—complicate rainfall tracking in urban areas during flash floods. To overcome these limitations, there is growing interest in approaches that utilize signal attenuation data from Commercial Microwave Links (CML) within telecommunication networks. Rain-induced attenuation on these CML signals provides critical information for estimating precipitation type and intensity.
This study aims to perform the first CML-based high-resolution rainfall estimation in Türkiye, focusing on İstanbul—a megacity with a population of 16 million, complex micro-climatic features, and high flood vulnerability. The methodological framework integrates signal data from thousands of CML lines operated by Vodafone Türkiye with a network of 45 rain gauges belonging to the Turkish State Meteorological Service (TSMS). The proposed model architecture includes wet-dry classification of signal data, dynamic correction of additional signal losses caused by antenna wetting (wet antenna effect), and optimization of the attenuation-rain rate relationship tailored to local precipitation characteristics. This paper presents the data processing infrastructure, station-based validation methods, and the potential contributions of this pioneering CML-based rainfall retrieval system to the urban hydro-meteorological monitoring capacity of İstanbul.
Keywords: Commercial Microwave Link (CML), Rainfall, İstanbul, Türkiye.
How to cite: Akın, M., Durmaz, A., Öztopal, A., Tekeli, A. E., Öztürk, A., Yilmaz, F., Tatli, Z., Çoban, D., Gül, G., Anit Türkan, S., and Öztopal, Z.: First Implementation of CML-Based Rainfall Estimation in Türkiye: Focus on İstanbul , 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-122, https://doi.org/10.5194/egusphere-plinius19-122, 2026.
Mediterranean environments are characterized by highly intermittent hydro-meteorological
processes, including intense rainfall events, flash floods and rapidly evolving atmospheric
conditions. These phenomena generate considerable forecasting challenges despite recent
advances in Machine Learning (ML), Deep Learning (DL) and foundation models applied to
environmental prediction.
While most studies focus on comparing forecasting algorithms, the intrinsic relationship
between the statistical structure of environmental variables and their forecastability remains
poorly understood. We hypothesize that predictive performance is strongly constrained by
the statistical organization of the observed processes and not solely by model complexity.
To investigate this question, we present SAPHIR DATA SERVICE, a cloud-native en-
vironmental intelligence platform developed at SPE lab, in the University of Corsica in
collaboration with LiSN Lab in the University of Paris Saclay . The platform supports
the continuous acquisition, storage, visualization and exploitation of heterogeneous envi-
ronmental observations originating from meteorological stations, hydrological sensors, IoT
monitoring devices and institutional environmental services.
SAPHIR DATA SERVICE relies on a dual-layer data architecture. A real-time database
supports operational monitoring, environmental surveillance and nowcasting activities, while
a historical database supports long-term analyses, retrospective studies, machine learning
training and scientific investigations. This architecture enables the simultaneous manage-
ment of operational and research-oriented workflows within a unified framework.
The backend infrastructure is continuously supervised through Grafana dashboards pro-
viding real-time monitoring of acquisition pipelines, database services, sensor status and
environmental observations. A complementary web frontend provides access to environmen-
tal indicators, historical analyses, forecasting products and decision-support services.
Beyond data management, the long-term objective of SAPHIR DATA SERVICE is the
implementation of a continuous environmental intelligence pipeline linking observation, in-
gestion, learning, prediction and decision support.
Within this framework, we investigate whether forecastability can be considered an in-
trinsic property of environmental variables and whether it can be explained through their
statistical signatures. To address this question, we introduce a characterization framework
combining temporal autocorrelation, spatial intercorrelation, spatio-temporal structure func-
tions, fractional moments, skewness, kurtosis and intermittency.
These descriptors are evaluated against forecasting performances obtained from a cata-
logue of statistical, machine learning and foundation models, including persistence baselines,
ensemble methods and modern deep-learning architectures.
The Porto-Vecchio study area provides a real-world Mediterranean testbed for evaluat-
ing how environmental statistical signatures relate to achievable forecasting skill within an
operational nowcasting framework. The proposed approach aims to establish a quantitative
relationship between environmental data structure and predictive performance, providing
new perspectives for rare-event forecasting, hydro-meteorological risk management and en-
vironmental decision-support systems.
How to cite: Silvani, X., Duchaud, J.-L., Muzy, J.-F., Paoli, C., and Al Agha, K.: SAPHIR DATA SERVICE: Environmental Forecastability andHydro-Meteorological Nowcasting, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-65, https://doi.org/10.5194/egusphere-plinius19-65, 2026.
This study introduces a network-based framework for characterizing and clustering the spatial organization of drought and wet extreme regimes at multiple accumulation timescales, applied to the 0.5° gridded SPEI Global Drought Monitor dataset over Italy. Extreme events are identified through run theory, extracting a set of statistical descriptors for each pixel and temporal scale: total number and frequency of events, temporal event occurrence fraction, mean and maximum duration, mean and maximum severity, mean and maximum intensity, global and local inter-event temporal irregularity, and the dominant season of event peak occurrence. These descriptors constitute a 12-dimensional feature space, which is transformed into a similarity network using Mahalanobis distance to account for inter-feature correlations. Community detection is then performed via the Louvain algorithm to identify pixel clusters sharing statistically similar extreme regimes. Partition quality is assessed through modularity Q and feature-space silhouette width and validated against a 1,000-sample permutation-based null distribution.
The Louvain algorithm consistently identifies 4 to 6 communities across all temporal scales for both drought and wet events, with all partitions achieving high statistical significance. For drought events, modularity Q exhibits a clear increasing trend at longer accumulation periods, rising monotonically from SPEI-7 to SPEI-12, indicating that community structure becomes progressively more pronounced as the accumulation timescale increases. F-ratio analysis reveals a clear three-phase regime across timescales. At short scales (SPEI-1, SPEI-3, SPEI-5), the dominant season of event peak occurrence governs cluster discrimination, reflecting the strong seasonality of short-term drought events. At intermediate scales (SPEI-6 to SPEI-9), discrimination shifts toward event-structure metrics such as severity and frequency. At long scales (SPEI-10 to SPEI-12), maximum spell severity and duration reach their highest explanatory power, indicating that persistent multi-month drought structures become the dominant axis of spatial differentiation.
For wet events, modularity Q does not exhibit a monotonic trend across scales and remains lower than its drought counterpart at most timescales, suggesting that wet regimes are spatially less structured than drought regimes. The dominant season of event peak occurrence is nearly uninformative for clustering, indicating an absence of seasonally organized differentiation. From SPEI-1 to SPEI-9, cluster discrimination is primarily driven by maximum duration and severity, peaking at SPEI-4 and SPEI-5. A sharp regime shift occurs at longer timescales (SPEI-10 and SPEI-11), where these duration- and severity-based features lose most of their discriminative power, and are replaced by mean intensity and inter-event temporal irregularity, revealing a scale-dependent reorganization of the drivers of spatial heterogeneity in wet spell regimes.
This study, integrating run theory, multi-scale feature extraction, and network-based community detection, provides a statistically robust and transferable framework for characterizing the spatial organization of hydroclimatic extremes, whose scale-dependent and phase-specific clustering structure would remain hidden to conventional single-scale approaches.
How to cite: Telesca, L.: Spatial clustering of drought and wet spell regimes over Italy: a feature-based network approach across multiple SPEI timescales, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-97, https://doi.org/10.5194/egusphere-plinius19-97, 2026.
Mountain regions are particularly vulnerable to climate change, as warming reduces snow and ice reserves, thus amplifying positive temperature feedbacks. These processes also have consequences for the hydrological cycle and, therefore, leading to wide-ranging impacts on society by altering ecosystem services and products. This highlights the importance of understanding how climate change affects mountain areas. However, the limited availability of long-term climate records at high elevations, due to adverse weather conditions, makes high-resolution regional climate models essential for studying complex terrain.
The CIMAs (Climate Research Initiative for Iberian Mountain Areas) project focuses on analyzing climate variability and the impact of climate change on the Central System of the Iberian Peninsula. The studied area is the largest mountain range of the peninsula, reaching 2.592 m at its highest point (Almanzor Peak) and includes surrounding areas with lower altitudes.
CIMAS observational data, gathered from several institutions in Portugal and Spain, was used to assess the accuracy of regional models over the domain of interest. Model simulations were produced with the WRF and HCLIM models spanning the period 1990-2025, using two nested domains at 12 and 4 km resolution and a nestdown strategy to produce a 1 km resolution simulation domain over the Central System. The domains at 4 and 1 km resolution were configured as convection-permitting. Both models were driven by the same boundary conditions provided by the ERA5 reanalysis, which was also used as a reference to evaluate the added value of increased resolution by each regional model. Specifically, temperature and precipitation variability at daily temporal resolution were used to evaluate model output against observations.
Results show how increasing resolution improves the simulation of temperatura and precipitation at high elevations and allows for better understanding of the climatology in complex terrain. The comparison of the WRF and HCLIM temperature simulations with observations highlights differences, mostly in the reproduction of extremes. Regarding the precipitation, 1 km resolution tends to overestimate the total accumulations due to an excess of simulated wet days.
How to cite: Madera Sánchez, S., González Rouco, J. F., García Bustamante, E., Navarro Montesinos, J., Vegas Cañas, C., Rodríguez Guisado, E., Sánchez Perrino, J. C., Prieto Rico, I., Rodríguez Camino, E., Cardoso Tavares, R. M., Greciano Zamorano, E., Cardoso dos Santos, L., and García Pereira, F.: Assessment of temperature and precipitation variability over a complex terrain. Multi-resolution model evaluación., 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-113, https://doi.org/10.5194/egusphere-plinius19-113, 2026.
Climate change impact studies (e.g. hydrology, agriculture, health) systematically require high-resolution meteorological variables. However, these disciplines rarely need the full suite of physical variables resolved by computationally expensive dynamical Regional Climate Models (RCMs). Their primary need lies in obtaining reliable projections and, crucially, a wide diversity of simulations to properly characterize climate uncertainty. To address this challenge, statistical emulation via deep learning emerges as a computationally efficient alternative.
This study proposes the use of UNet-based neural architectures as direct emulators of EURO-CORDEX RCMs for high-resolution temperature (EUR-11). The central research question is highly operational: is it possible to train an emulator using a single historical GCM-RCM pair and rely on its capacity to regionalize future projections driven by Global Climate Models (GCMs) different from the one used during its training?
To answer this, the model undergoes a rigorous generalization test against the CORDEX ensemble. Once the network is trained specifically using the historical MOHC-HadGEM2-ES and DMI-HIRHAM5 pair, the emulator is fed with the boundary conditions of all other GCMs available for the DMI-HIRHAM5 regional model in the European repository. The validation assesses whether the emulator's outputs, when forced by these new GCMs, deviate significantly from the original dynamical CORDEX projections for those exact pairs.
Preliminary results show that the dispersion (uncertainty spread) generated by the emulator's inference ensemble is equivalent to that of the original CORDEX dynamical ensemble. Nevertheless, zero-shot cross-evaluations reveal that the emulator systematically deviates from its dynamical "mirror" pair. These findings help define the viability of using neural networks to generate on-demand climate ensembles, highlighting both their potential for ultra-fast climate variability reproduction and the limitations associated with transfer biases between global models.
Acknowledgements: This work was supported by the ARUBA (PID2023-149080OB-I00/MCIN/EI/10.13039/501100011033), and the INSIEME (FSRM/10.13039/100007801) projects.
How to cite: Pravia-Sarabia, E. and Montávez, J. P.: Rapid emulation of Regional Climate Models via deep learning for impact studies: evaluating inter-GCM generalization, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-77, https://doi.org/10.5194/egusphere-plinius19-77, 2026.
Short-duration intense rainfall events drive some of the most destructive flood hazards in the Mediterranean, posing considerable difficulties for operational early warning systems. Reliable nowcasting (short-term forecasting) of convective rainfall is essential for hydrological response modelling and risk management. Nevertheless, numerical weather prediction models frequently fail to capture storm initiation and localization, especially over complex terrain.
The present study investigates the integration of polarimetric weather radar data into the Weather Research and Forecasting (WRF) model using a four-dimensional variation (4DVAR) data assimilation technique, to improve rainfall forecasts for flood-relevant time scales. Simulations are performed for selected high-impact precipitation events that occurred over Greece between 2024 and 2026, including cases associated with flash flooding. Through 4DVAR cycling, radar reflectivity and radial wind observations are assimilated, with simulations conducted at 1-km resolution and a 3-hour forecast horizon, aligned with nowcasting time scales. Additionally, humidity, vertical velocity and horizontal wind divergence profiles derived from lightning data at storm locations, are also assimilated with a three-dimensional variation (3DVAR) method. To assess whether data assimilation is sensitive to the choice of initial and boundary conditions, experiments with different initialization data (ICON and GFS) are performed. Results, using primarily the measured reflectivity and radial wind velocity from the weather radar and the proxy lightning data at larger range, indicate that assimilation using these data significantly improves convective initiation, storm structure, and peak rainfall placement during the first forecast hours. These findings suggest that radar-based 4DVAR assimilation has the potential to strengthen operational flood early-warning systems by delivering more reliable rainfall forcing hydrological and decision-support models. Ongoing studies examine its integration within multi-sensor workflows, coupling with meteorological forecasting chains, with the goal of operational implementation in Greece.
How to cite: Katsanos, D., Kalogiros, J., Portalakis, P., Roukounakis, N., and Retalis, A.: Precipitation Nowcasting over complex terrain in Greece, using weather radar and lighting data assimilation, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-17, https://doi.org/10.5194/egusphere-plinius19-17, 2026.
Medicanes are Mediterranean cyclones with the potential to cause devastating floods, storm surges and windstorms, often leading to significant disruption and casualties. During their mature phase, they exhibit tropical-like cyclone features, such as a warm core (WC), a cloud free eye surrounded by spiraling rain bands around the center, a closed vortex associated with strong near-surface winds. Generally, these cyclones originate from extra-tropical cyclones showing a cut-off from the main flow allowing the intrusion of relatively warm stratospheric air resulting in a top-bottom WC development. More rarely, they undergo the so-called tropical transition during their mature phase, exhibiting at some point a deep axi-symmetric WC of diabatic origin. The term Medicane generally refers to both types of cyclones regardless of the processes originating them (https://glossary.ametsoc.org/wiki/medicane /).
The present work provides an overview of recent studies on the use of satellite-based microwave observations to monitor and characterize the WC, deep convection (DC), the presence of a closed eye surrounded by a ring-shaped band of intense winds during the cyclone evolution, and to identify the tropical transition also inferring the dynamics at the different stages of cyclone’s lifetime. Relevant advances on this topic are being accomplished within the ESA MEDICANES project (https://medicanes.isac.cnr.it/). The detection and characterization of WC and DC are based on satellite passive microwave (PMW) measurements from different radiometers onboard Low Earth Orbit (LEO) satellites. In particular, temperature sounding channels in the 50-60 GHz oxygen absorption band are used to identify the presence of the WC, while high frequency channels (90-190 GHz) are used to identify the presence of the closed eye and the DC areas. In addition, the 183.31 GHz water vapour channels can provide useful insight on the dynamic highlighting the potential vorticity (PV) anomaly resulting from warm dry stratospheric air intrusion. Surface wind structure and intensity are characterized using ocean surface wind products derived from scatterometer missions. Automated analysis through the Medicane Rotational Center Automated Detection (MeRCAD) algorithm identifies the radius of maximum wind, assesses wind-field symmetry, and detects nearly closed ring-shaped bands of intense surface winds typical of mature medicanes. This dynamical characterization complements the thermodynamic information derived from PMW observations, enabling objective identification of tropical-like surface circulation and its temporal evolution. An additional step forward in medicanes’ characterization and monitoring is being carried out exploiting different machine learning (ML) approaches for automated detection of the Medicanes’ features, including the identification of the WC. A semi-supervised deep learning anomaly-detection framework, based on convolutional autoencoders, is used to identify WC signatures as anomalies relative to the dominant non-WC atmospheric states. Training is performed primarily on unlabeled non-WC cases, with only a limited number of labeled WC events, enabling robust learning under strong data imbalance. The system is trained and evaluated on approximately 30,000 satellite overpasses covering nearly 900 Mediterranean cyclones (years 2000–2020), and demonstrates reliable WC discrimination using recall-oriented performance metrics. This work aims to show as a fully MW-based characterization of dynamics, thermodynamics and microphysical processes involved within a medicane is possible.
How to cite: D'Adderio, L. P., Panegrossi, G., Sebastianelli, S., D'Armiento, D., Casella, D., Sanò, P., Camplani, A., Saran, C., Gosset, A., and Herndon, D.: New perspectives and advancements in microwave-based analyses and characterization of Medicanes, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-44, https://doi.org/10.5194/egusphere-plinius19-44, 2026.
Accurate precipitation estimates are paramount for reliable hydrological monitoring of drought. Precipitation constitutes the principal input of most drought indices, and even minor systematic errors can substantially influence the detection, timing, and intensity of drought episodes. This concern is especially pronounced for indices such as the Standardized Precipitation Index (SPI), which depend solely on precipitation records and are extensively employed in operational drought monitoring across various time scales. In areas like Italy, defined by rugged terrain, interactions between coastal and mountain environments, and an irregular network of rain-gauge stations, uncertainties in rainfall measurement can propagate directly into drought evaluations, potentially undermining the dependability of decision-support tools. To overcome these shortcomings, satellite-derived precipitation products have become vital for surface-based observations, offering spatially continuous coverage and near-real-time availability. Nevertheless, their accuracy varies considerably according to retrieval technique, spatial resolution, and the dominant weather conditions, making a thorough evaluation essential prior to their use in drought monitoring applications. This study aims to examine how various satellite precipitation products influence SPI-based drought characterization across Italy. Four widely adopted satellite precipitation datasets, CHIRPS, GPM, PDIRNOW, and SM2RAIN, were chosen to represent a wide spectrum of retrieval strategies, including infrared–station hybrid methods, passive microwave integration, multi-sensor geostationary blending, neural-network-driven infrared approaches, and soil-moisture inversion techniques. Their varied temporal and spatial resolutions render them appropriate for both research purposes and operational monitoring contexts. SPI values derived from each satellite product were systematically compared. The analysis reveals pronounced discrepancies in SPI magnitude, frequency, and duration depending on the precipitation dataset used, evidencing how sensitive drought assessments are to errors in rainfall estimation. The findings show that no individual satellite product consistently surpasses the others, and suggest that combining multiple satellite datasets or adopting hybrid methodologies can enhance the robustness of SPI-based drought monitoring in complex Mediterranean settings. Furthermore, the results highlight the importance of establishing a benchmark dataset. The use of ground-based measurements, even over a geographically restricted area, can help to identify the most appropriate product, ultimately allowing for a more dependable analysis of the spatial patterns of drought events, with direct relevance to water resource planning and management.
This work was funded by the Next Generation EU—Italian NRRP, Mission 4, Component 2, Investment 1.5, call for the creation and strengthening of ‘Innovation Ecosystems’, building ‘Territorial R&D Leaders’ (Directorial Decree n. 2021/3277)—project Tech4You—Technologies for climate change adaptation and quality of life improvement, n. ECS0000009. This work reflects only the authors’ views and opinions; neither the Ministry for University and Research nor the European Commission can be considered responsible for them.
How to cite: Pellicone, G., Coscarelli, R., Caloiero, T., De Marco, A., and Chiaravalloti, F.: Comparing remote sensing products for drought characteristics analysis in Italy , 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-49, https://doi.org/10.5194/egusphere-plinius19-49, 2026.
Mediterranean cyclones can be a threat for human lives and can also have important economical impacts, since are often associated with heavy rainfall and intense winds.
Among mediterranean cyclone, a particular class, known as Medicanes (Mediterranean Hurricanes) has attracted great attention is the last years. Medicanes are interesting from the scientific point of view because of their tropical-like characteristics: a symmetric structure, spiraling clouds, a calm cloud-free eye and a warm core.
Numerical Weather Prediction (NWP) models can be employed to predict Medicane trajectories and impacts. The accuracy of a NWP forecast strictly depends on the representation of the initial state of the atmosphere, which can be improved by data assimilation.
In this work, we focus on the assimilation of the Advanced SCATterometer (ASCAT) radar data into the Weather Research and Forecasting (WRF) model and we consider the impact of ASCAT assimilation for two medicanes: Ianos, which occurred between 15 and 21 September 2020 in the central Mediterranean and made landfall on the west coast of Greece, and Jolina, which occurred between 14 and 19 March 2026 over the central Mediterranean, impacting parts of northern Africa, southern Italy, and Libya.
For both medicanes, simulations are performed using an En3DVar approach with the initial and boundary conditions derived from the European Centre for Medium range Weather Forecast – Ensemble Prediction System (ECMWF-EPS). Using this method the background error covariance matrix is computed from the ensemble and is aware of the meteorological conditions of the day. Two kind of simulations are considered: without ASCAT data assimilation (named CTRL) and with ASCAT data assimilation (named ASCAT).
The forecast trajectories are compared with the best a-posteriori estimate of the trajectory. Results show that ASCAT assimilation into the WRF model positively impacts the prediction of the Medicane trajectory for both cases, and ASCAT trajectories are improved for most members and for all forecasting times compared to CTRL.
How to cite: Federico, S. and Torcasio, R. C.: The impact of ASCAT surface winds data assimilation on medicane prediction: results for two cases., 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-111, https://doi.org/10.5194/egusphere-plinius19-111, 2026.
The Climate research initiative for Iberian Mountain Areas (CIMAs) is a collaborative framework involving several Spanish institutions: the Spanish Meteorological Office (AEMET), Complutense University of Madrid (UCM), Institute of Geosciences (IGEO, CSIC-UCM) and CIEMAT. The main goal of the initiative is to advance the characterization and understanding of climate variability and change in the Central System of the Iberian Peninsula. Mountain regions are particularly sensitive to climate change, however observational data in these environments remain scarce, heterogeneous and difficult to maintain. CIMAs addresses this challenge by integrating multi-source meteorological datasets from institutions with different measurement protocols, temporal resolutions and data formats, such as AEMET, the Guadarrama Monitoring Network (GuMNet), the Portuguese Meteorological Office (IPMA), hydrological agencies operating Automatic Hydrological Information Systems in Spain (SAIH Duero, SAIH Tajo) and the Portuguese National Water Resources Information System (SNIRH).
In this work, we present the comprehensive CIMAs framework, which currently articulates three complementary lines of action. First, the development of the observational database, which includes spatial-temporal harmonization, metadata consolidation, systematic quality control, and a version-controlled architecture. Temperature and precipitation databases are currently operational, supplemented by the integration of wind and snow height products.
This robust observational dataset allows for the development and evaluation of regional climate simulations. The initiative employs WRF and HCLIM models to dynamically downscale ERA5 reanalysis up to very high resolutions (1 km) over the 1990-2025 period. These simulations successfully reduce temperature and precipitation biases over mountain areas. Ongoing developments aim to couple these regional models with the global MPI-ESM to simulate future climate scenarios driven by CMIP6 emission pathways up to the year 2300.
Moreover, CIMAs aims to translate this climate research into climate services tailored for end-users. Building upon the integrated observations and simulations, the goal is to develop tools for sustainable territorial planning, vulnerability assessment and sectoral applications such as hydrology, forestry and energy.
The CIMAs framework provides a structured, interoperable basis for integrating climate observations and simulations across high-mountain areas of the Iberian Peninsula. Supported by a dedicated web platform for data visualization and access, it offers a solid foundation for assessing simulation performance, improving regional climate knowledge, and developing actionable climate services.
How to cite: Vegas Cañas, C., González Rouco, J. F., Rodríguez Guisado, E., Rodríguez Camino, E., Cardoso, R. M., Santos, L. C., Navarro Montesino, J., García Bustamante, E., Madera-Sánchez, S., Greciano-Zamorano, E., Pereira, C., Luna, Y., Morata, A., Robles-Martínez, G., and Hinojal, J. A.: CIMAs: A multi-source climate dataset for high-mountain environments in the Iberian Central System, 19th Plinius Conference on Mediterranean Risks, Murcia, Spain, 6–9 Oct 2026, Plinius19-112, https://doi.org/10.5194/egusphere-plinius19-112, 2026.
