ARCHIMEDE—An Innovative Web-GIS Platform for the Study of Medicanes

Abstract

Mediterranean hurricanes, also known as medicanes, can cause significant damage to coastal communities. Consequently, they have been the subject of intense study in recent decades. Geographic Information Systems (GISs) and Web-GIS applications are essential for processing and analyzing geographic data from various sources, particularly in the field of spatial planning. Due to their ability to integrate heterogeneous information, these tools can provide experts with a comprehensive overview of relevant geospatial data. Here, we present ARCHIMEDE, a relational geodatabase connected to an open-source Web-GIS platform focused on Mediterranean hurricanes that contains data from recent research on these extreme weather phenomena. These datasets comprise climatic and oceanographic data obtained from remote sensing techniques as well as seismic and geomorphological data obtained from field observations. Consequently, this Web-GIS platform can enhance our understanding of Mediterranean cyclones by relating the physical properties of these natural phenomena with their impacts on coastal regions. ARCHIMEDE is an innovative tool for the dissemination of geographic information to stakeholders, researchers, and decision-makers, offering valuable support for the development of robust coastal management strategies aimed at mitigating the challenges posed by Mediterranean hurricanes.

1. Introduction

Historically, Mediterranean coastal regions have been exposed to several cyclones that share visual characteristics with tropical cyclones (TCs), such as a central eye and the presence of a convective eyewall. Observations of intense rotations around a deep, central warm core as well as reduced environmental wind shear compared to extratropical cyclones suggest that this similarity is not limited to their visual appearance but also includes their physical processes and mechanisms [1,2]. The sustained wind speeds of these Mediterranean hurricanes—also known as “medicanes”—can be as high as a Category 1 hurricane (sustained wind speed is about 33–42 m/s) on the Saffir–Simpson scale [3,4,5]. Consequently, these extreme weather events are associated with destructive winds, torrential rainfall, and storm surges, resulting in flash floods, debris flows, and landslides that can cause substantial damage to local communities [6].

Several studies have significantly improved our knowledge of medicanes and their peculiarities compared to extratropical cyclones, especially in terms of (i) the ability to analyze the physical processes responsible for their genesis and development, (ii) understanding the variations in environmental parameters prior to their development, (iii) their impacts on coastal regions and the relative vulnerability of different environmental settings to these kinds of events, and (iv) the development of new techniques to study and monitor their occurrence [7,8,9]. While there is currently no formal definition for a “medicane”, Gutierrez-Fernandez et al. [10] identified the typical baroclinicity values as well as some large- and mesoscale parameters that characterize the development of warm-core cyclones. Recent climatological studies have also indicated that the intensity and duration of medicanes may significantly increase in a warming climate, even if the overall number of occurrences is expected to remain relatively stable [3,4,11]. Pytharoulis [12] and Stathopoulos et al. [13] noted that sea surface temperatures (SSTs), which are strongly increasing due to global warming [14], appear to be an important parameter that influences medicane intensification. Scardino et al. [15] analyzed 52 Mediterranean cyclones between 1969 and 2023 and observed a significant drop in SSTs (approximately 1.6 °C on average) in the period preceding cyclogenesis. After comparing these data with SST data from satellites, Argo float records, and a coastal temperature gauge located in southeastern Sicily, they deduced that changes in SST prior to medicane formation could be used as an early warning indicator [15].

Medicanes can have a severe impact on coastlines [16], leaving geomorphological evidence such as (i) flooded surfaces and the deposition of sediment in low-lying areas and rocky coasts; (ii) the triggering of gravity-driven processes such as cliff slumps, landslides, and debris flows; and (iii) alluvial flooding. Diakakis et al. [6] studied the impact of Medicane Ianos on Greece by focusing on the geomorphological influence of the medicane on mass movements in both inland and coastal areas. In the context of coastal erosion, they observed small slumps at the base of the cliffs caused by shear waves; specifically, rotational and planar slips, debris flows, and rock falls were identified in the wake of the storm. The study described how these phenomena were influenced by the lithology, the geological structure, and the geotechnical conditions of the cliff material [6]. Washout phenomena were identified across beach environments in various regions, including the washout area and the backshore. In some cases, coastal line regression was identified. Kotsi et al. [17] employed UAS-Aided Photogrammetry to monitor and quantify the geomorphic effects of extreme weather events in tectonically active mass waste-prone areas. Their findings revealed a significant change within the studied timeframe (between July and October 2020) attributed to Medicane Ianos. An analysis of rainfall and earthquake data revealed that no other significant events occurred during that period. Additional geomorphological evidence was documented along the Sicilian coasts, where comparisons between common storms and medicanes have been conducted [18]. Scardino et al. [19] used the LEUCOTEA system to evaluate the impact of waves and storm surges on the coastline, primarily in the form of boulder displacements. A subsequent study by Scardino et al. [20] highlighted the occurrence of storm surge values higher than those typically seen in seasonal storms; these affected southeastern Sicily during the hybrid cyclone Helios [21]. Helios is described as a hybrid storm: it exhibited features characteristic of both extratropical cyclones (upper-level cold core) and tropical cyclones (low-level warm core); however, it did not exhibit the intense convection during the mature stage that is typical of Mediterranean tropical-like cyclones. Wave heights were found to be comparable to those observed in previous medicanes, such as Zorbas and Apollo. Medicanes have also been studied as a seismic source. Several authors have highlighted the strong relationship between the amplitude of a particular seismic signal, the sea state [22,23,24,25], and the presence of cyclonic activity, including hurricanes [26], typhoons [27], tropical cyclones [28], and medicanes [7,8,29]. Borzì et al. [8] analyzed eight medicane events and four extratropical storms to define a characteristic seismic signature for medicanes using a method that exploits the coherence of continuous seismic noise and the strength of the medicane from a seismic perspective; the authors called this the Microseism Reduced Amplitude. In this context, the term “seismic source” does not imply that medicanes induce seismic events in the traditional sense, such as earthquakes. Instead, it means that the medicanes generate a strong wave motion that can be registered by seismometers due to the energy transfer from the sea waves to the seafloor [7].

The integration of these diverse studies on medicanes, which span a wide variety of geophysical disciplines, as well as their contextualization within varied fields underscore the necessity for a unified geodatabase on medicanes. While the literature offers a significant amount of climatological and oceanographic data on Mediterranean cyclones, accurate descriptions and analyses of the impacts that these events have on coastal areas are scarce. Therefore, the ability to link the physical processes of these phenomena to their real-world impacts could serve as a key tool for defining the vulnerability of different regions. In this manuscript, we describe the development of ARCHIMEDE, a Web-GIS platform that contains information from recent studies on Mediterranean cyclones as well as new data obtained from remote sensing instruments and direct field observations (Figure 1). The integration of data from different sources, including information relevant to the social sciences and impact studies, represents a new approach to monitoring and measuring the impacts of extreme weather- and climate-related events. We anticipate that ARCHIMEDE will improve the accessibility of geospatial data, foster collaboration, and facilitate informed decision-making. By conducting an in-depth analysis of these key aspects, this study aims to provide an enhanced understanding of the geographical features of the Mediterranean basin.

2. Materials and Methods

2.1. Geodatabase Development

The ARCHIMEDE geodatabase was initialized by separating extratropical cyclones and Mediterranean tropical-like cyclones. The latter were categorized based on their wind speed and mean sea level pressure (MSLP), following the standards of the German Meteorological Service, which proposed an unofficial classification based on the average peak wind speed of intensity v, following the Saffir–Simpson scale for tropical cyclones (

Table 1. The classification of Mediterranean tropical-like cyclones based on an intensity scale. Only extratropical cyclones with an intensity matching this table were considered in this study.

	Mediterranean Tropical Depression	Mediterranean Tropical Storm	Mediterranean Hurricane
Wind Speed (km/h)	<63	64–111	>112
MSLP (hPa)	1006–1015	994–1005	974–993

) [30]. A dataset of 51 Mediterranean cyclones (Mediterranean hurricanes, Mediterranean tropical storm, Mediterranean tropical depression, Mediterranean tropical disturbance and extratropical cyclones) was selected; these events date back to as early as 1969 (Supplementary Material) [15]. Extratropical cyclones were selected based on their severity and their significant impact on coastal areas [31,32,33].

A relational database was used to store and manage the cyclone data; PostgreSQL was selected because it is an open-source database system that offers excellent support for geospatial datasets through the PostGIS extension and is also capable of integrating a variety of spatial datasets in both vector and raster forms in a web-based application. The database includes several different geospatial datasets related to each event:

(1)

Cyclone tracks and position derived from MSLP extracted from ERA-5 reanalysis, with tracks and positions based on every 6 h mean sea level pressure;

(2)

SST differences (following the methods reported in Scardino et al. [15]) obtained from the reanalysis of CMEMS (Copernicus Marine Environment Monitoring Service) and satellite data;

(3)

Wind speeds extracted from ERA-5 reanalysis, considering the eastward wind component (U wind) and the northward wind component (V wind), 10 m above the surface with an hourly temporal span;

(4)

Seismic data;

(5)

Old and new geomorphological data.

This database was connected to a QGIS environment in order to develop a Web-GIS application through the QGIS2Web plugin. This plugin generates a simple web application containing all of the related files and resources in an organized manner, including basic JavaScript, CSS, and HTML, while all of the layers are contained in an expanded group system without any interactive graphics. The JavaScript and HTML files were subsequently modified to organize the layers into different groups (i.e., tropical-like and extratropical cyclones) and sub-groups (e.g., Mediterranean hurricanes and tropical storms). Additional features and graphics were also created to produce an interactive web map application. Different base map services, such as Google satellite maps and ESRI world topographic maps, served as the global canvas extent (Figure 1). The tracks dataset, obtained from the ERA-5 reanalysis, contains detailed information about the path that the cyclones took, including latitude, longitude, date, and time of occurrence. Consequently, this dataset allows for the visualization of the movement of cyclones over time. SST measurements obtained at different time intervals during the storm events are crucial for understanding the relationship between cyclone intensity and sea conditions (especially sudden drops in SSTs), while the MSLP dataset provides insights into variations in atmospheric pressure. Geomorphic datasets provide information about the physical changes and impacts caused by these cyclone events on the landscape.

2.2. Development of Web-GIS Platform

The QGIS2Web plugin (Appendix A) was used to develop the Web-GIS platform that was used to analyze the geomorphological effects of medicanes and extratropical cyclones; this application leverages the capabilities of the QGIS software (version-3.30.1) to create interactive web-based maps [34]. This platform will allow researchers and stakeholders to analyze and visualize geographic data associated with extreme weather events in an accessible and user-friendly manner. In particular, the QGIS2Web plugin extends the functionality of QGIS projects by converting them into interactive web maps. We performed the following steps to develop the Web-GIS system described in this manuscript (Figure 2).

(1)

Data preparation: We collected and compiled relevant spatial data to create the geodatabase. This included various weather and climate datasets, including SST, MSLP, and cyclone tracks integrated with geomorphological data, including flooding limits, coastal erosion, and boulder displacements observed through geotagged images. We also integrated the results of seismic analysis, including the “seismic” positions and tracks of each medicane.

(2)

QGIS project configuration: We configured the QGIS project by using the following styling for each of the different data layers: (1) polylines for the medicane tracks (both satellite and “seismic” tracks); (2) isobars for MSLP and flooding limits; (3) raster maps of SST variations as well as DTMs and DSMs; (4) geotagged points with geomorphological evidence associated with medicanes. This process involved the selection of appropriate symbology as well as the definition of relevant attribute data to facilitate a comprehensive analysis and effective visualization. We also employed refined cartographic techniques to enhance the representation of geomorphological features to more accurately depict the impacts of storms.

(3)

QGIS2Web Plugin: This plugin allowed us to convert the QGIS project into a web-based map. We used the plugin settings to customize the map’s appearance, including the base map options, layer visibility, and potential interactions.

(4)

Exporting the Web Map: We exported the QGIS project as a web map using the QGIS2Web plugin. This process resulted in the production of web-friendly files, including HTML, JavaScript, and CSS files, which could be hosted on a web server or shared through other web platforms. The exported web map retained the symbology, labeling, and attribute data defined in the original QGIS project configuration.

(5)

Web Map Deployment: We hosted the exported web map files on a web server and deployed them on a cloud-based platform, allowing users to access the web map through a standard web browser. We ensured that the web map was easily accessible for the intended audience and provided the appropriate documentation and instructions for its use.

(6)

On the Web-GIS platform, polylines of flooding limits were reported for the following events: Qendresa, Zorbas, and Helios. Pictures associated with cyclone impacts in several regions across the Mediterranean were stored as geotagged images. The Web-GIS application also allowed for the display of points with georeferenced images.

The Web-GIS portal presented in this manuscript can be found at the following URL: https://archimedemedicane.it (accessed on 1 July 2024).

2.3. Remote Sensing and Reanalysis Data

We analyzed satellite and reanalysis data from CMEMS for the Mediterranean basin (Supplementary Table S1) to obtain important parameters associated with the cyclonic events contained in our database. These parameters included SST, obtained from the Copernicus Marine Data Store [35,36,37], MSLP, wind speed, significant wave height, and sea surface height, obtained from the ERA-5 portal [35,38,39,40]. MSLP data were obtained from ERA-5: the fifth generation of the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis. This dataset contains comprehensive global climate and weather data from the last eight decades (1940 to the present day). These datasets cover a temporal range that includes the medicanes and extratropical cyclone events analyzed in the literature. The reanalysis amalgamates model data and observations globally, yielding a coherent and exhaustive dataset that includes observations across the globe. ERA-5 offers hourly estimates for various atmospheric, oceanic, and land surface parameters [41]. Wind data were extracted from ERA-5 [41] for (i) the zonal wind component (U wind at an elevation of 10 m) and (ii) the meridional wind component (V wind at an elevation of 10 m). These components were combined to obtain wind speeds during storm events (Equation (1)):

$$WS=\sqrt{U^2+V^2}$$

(1)

Wind speed data were reported as daily raster maps during the lifespan of each cyclone.

Copernicus satellite-derived Level 4 SST data were used for the analysis of historical cyclone events from 1995 to the present day. These SST products are derived from nocturnal infrared images captured by sensors on a wide range of satellite platforms, and have a spatial extent that encompasses the Southern European Seas. We used significant wave height during the lifetime of cyclone events to describe their hydrodynamic features; these data were obtained from ERA-5 [41]. Significant wave height refers to the average wave height of the top one-third of the highest waves generated by local wind. The significant wave height of each cyclone event was reported as raster maps, highlighting their variations during the lifetime of the cyclone. Sea surface height was obtained from Copernicus reanalysis [35,36] and was considered to represent the water level above the geoid. These data were also provided in the geodatabase as raster maps and can be visualized by the end-user. The layers described above were included in the geodatabase with the following parameters (Figure 3):

(1)

MSLP (hPa) is stored in a vector format in the form of isolines;

(2)

Wind speeds (m/s) are presented using a raster format with geotagged pictures;

(3)

SSTs (°C) are presented using a raster format in the form of a GeoTIFF;

(4)

Significant wave heights (m) are presented using a raster format with geotagged pictures;

(5)

Sea surface height (m) is presented using a raster format with geotagged pictures.

The resolution of each data layer is reported in Supplementary Table S1 and the data are available in the github repository (https://github.com/alokkush2024/Archimede_datasets, accessed on 1 July 2024).

2.4. Old Geomorphological Data

Geomorphological evidence associated with historical medicane occurrences was collected from diverse sources, including published articles, reports, and in situ surveys [18,42,43,44]. The following features were included in the geodatabase:

(1)

Coastal flooding presented in the form of vector polylines representing the inundation limit reached by the water level during the cyclone event;

(2)

Out-of-size deposits presented in the form of vector polylines representing boulder perimeters at different times;

(3)

Geotagged pictures of the direct observations and landforms (landslides, rock falls, debris flow, riverbank erosion, coastal erosion) reported on the web and in the scientific literature.

Coastal flooding surfaces are the most common type of geomorphological evidence associated with the impacts of medicanes; these surfaces represent the landward deposition of marine sediment [18,45,46]. Marine sediments associated with the impacts of Medicane Ianos (September 2020) were detected in Peloponnesus (Greece) [42,45]. Wave impacts and rainfall during medicane occurrences can also trigger gravitationally driven processes, such as the cliff slumping observed in Cephalonia during Medicane Ianos [47]. Medicane-triggered landslides have also been detected in inland areas, usually caused by the intense rainfall associated with medicane events [48,49,50].

From a geomorphological perspective, medicane events can result in the deposition of (i) out-of-size landforms, such as washover fans and the accumulation of mega-boulders; and (ii) horizons of high-energy sediments interbedded between the coastal plain and shelf deposits. In particular, the deposition of coastal boulders (up to 40 m³ and weighing up to 40 tons) is strongly correlated with the occurrence of medicanes and storm events [44,51,52]. These coastal boulders can be part of a variety of sedimentary assemblages [53,54,55], including isolated boulders, fields of scattered boulders, boulder clusters (spaced, stacked, or imbricated), and boulder ridges. Although these deposits can also be generated by other extreme events, such as tsunamis [56,57], over the last few decades, two strong medicanes have been observed to transport boulders along the rocky coast of the Peninsula of Maddalena (Siracusa, southeastern Sicily). Video evidence provided by the Marine Protected Area of Plemmirio highlighted the movement of boulders associated with Medicane Zorbas [44] and Medicane Apollo (October 2021) [19], representing the first case of video evidence for the movement of boulders associated with medicanes.

These studies involved the collection and integration of geomorphological evidence related to medicane events from a variety of sources, including diverse landforms, sedimentary deposits, and erosion features. The latter were analyzed to understand the intensity, duration, and impact of these events on coastal and inland environments. The collected data were geotagged for precise spatial context. This comprehensive approach will contribute to the assessment of medicane characteristics, their effects, and their potential influence on ecosystems.

2.5. New Geomorphological Data

Following the occurrence of cyclone Helios in 2023, we performed post-event surveys using terrestrial laser scanning (TLS) and global positioning system real-time kinematic (GPS RTK) techniques along the coasts of southeastern Sicily. This allowed us to assess, with centimetric accuracy, the flooding limits associated with marine inundation in eight coastal areas in southeastern Sicily: Acitrezza, the Thapsos Peninsula, the Marine Protected Area of Plemmirio, Arenella, Punta del Cane beach, Avola, and Marzamemi Town (Figure 4). These represent the areas that have been most severely impacted by medicanes over the last decade.

TLS was performed in combination with GPS-RTK surveys using a Faro Focus phase-shift instrument to obtain the Digital Terrain Model (DTM) and Digital Surface Model (DSM) of the coastal areas. Additional reference points of pre-cyclone features were measured using orthophotos and a separate DTM provided by regional authorities and surveyed using airborne photogrammetric and LiDAR (Light Detection and Ranging, 2 × 2 m cell size) techniques (provided by ex Ministero dell’Ambiente). To investigate coastal flooding limits and sediment erosion and accumulation, we conducted a comprehensive geospatial analysis of the sandy coast of Arenella, which features a nearby back lagoon that was the subject of previous studies on medicane impacts [18]. Specifically, we calculated the Digital Elevation Model (DEM) of Difference obtained from LiDAR in 2009 (1 × 1 m cell size) as well as the dataset obtained via TLS (0.5 × 0.5 m cell size) following cyclone Helios in 2023 [21]. We then compared the flooding patterns observed during several extratropical storms, revealing that inundation extents were greater during medicane events as well as cyclone Helios.

In addition, we conducted a comprehensive survey of the Maddalena Peninsula (Sicily, Italy), a rocky coastal area, to evaluate the boulder displacements associated with various storm events that have occurred since 2009. This survey used a systematic approach that combined field observations, measurements, and data analysis techniques to understand the dynamic nature of the coastal environment and the impact of extreme weather events on the movement of large boulders. We first identified and selected representative boulders along the Maddalena Peninsula, taking into account their size, shape, and proximity to the shoreline. These boulders were then carefully monitored over time to document any changes in their position. The displacement of these boulders was quantified using precise measurement methods such as GPS and TLS surveys, allowing us to accurately record the locations of each boulder and track their movements over multiple storm events.

The data acquired from these surveys were included in the geodatabase (https://github.com/alokkush2024/Archimede_datasets, accessed on 1 July 2024) in the following formats:

(1)

Points: geotagged pictures of geomorphological evidence observed in the field (Figure 5).

(2)

Polylines: coastal flooding limits as observed in the field; isolines of sediment thickness were assessed using a DEM of Difference.

(3)

Rasters: DEMs obtained from TLS data in the form of GeoTIFFs.

2.6. Seismic Data

Seismic data from ~100 seismic stations installed along the Italian, Maltese, Greek, and French coasts were downloaded to analyze medicanes from a seismic perspective. This included seismic data from six medicanes (Rolf, Qendresa, Trixie, Zorbas, Ianos, and Apollo) as well as the hybrid cyclone Helios [58]. Due to the long interval between the earliest (Rolf, 2011) and the latest storm event (Helios, 2023), the number of available seismic stations varied slightly due to new installations, malfunctions, or disposal. Seismic stations were selected based on the following characteristics: they must be (i) installed in coastal areas and (ii) equipped with 3-component broadband seismic sensors. In addition, 15 seismic stations installed in the Etnean area were used for array analysis.

Following the approach of Borzì et al. [29], the seismic data were first corrected for instrument response, after which spectral and amplitude analyses were performed. The seismic signal was processed according to the method described by Welch [59], utilizing a time window of 81 s. This allowed for the computation of spectra at hourly intervals. The hourly spectral data were then collated and represented as spectrograms, with the time on the x-axis, the frequency on the y-axis, and the logarithm of the Power Spectral Density (PSD)—which provides information about the “microseism energy”—represented by the color bar. Since the microseism is a continuous seismic signal, it is not possible to apply the conventional methods used to locate earthquakes to determine their source. Consequently, the source of the microseism was located using amplitude decay and array analysis [29,60]. The microseism source was stored in the geodatabase in a vector format in the form of points that indicated the time of acquisition (Figure 6). The location and time of acquisition reflect the moment at which the wave interference was recorded.

2.7. Data Analysis

Several datasets were analyzed in order to improve the Web-GIS application. We first examined the MSLP obtained from ERA-5 reanalysis data to determine the track of each cyclone event following the method described by Chauvin et al. [61]. The data were reported in the form of isobars since the cyclone tracks follow the displacement of the pressure minimum. Daily SSTs derived from Copernicus satellite observations were used to conduct a comprehensive time series analysis spanning summer through winter in years featuring cyclone occurrences. The time series data were extracted through resampling of the cyclone source area based on the coordinates of cyclogenesis, as presented in Supplementary Table S2. This resampling was performed using 110 km cells to determine the daily average SST. Thermal drops were assessed by calculating the difference in the SST at the end of the cyclone’s lifetime and the SST recorded 10 days prior to cyclone formation [15]. SST values from 10 days prior to the event were used in order to determine the maximum thermal drop, which can be used to characterize cyclone evolution. We also analyzed significant wave height and wind speed data from ERA-5 for all strong cyclone events, showing that intense wind speed and waves were concentrated along the path of the cyclone. This analysis highlights the importance of accurate cyclone tracking and forecasting for timely evacuation and disaster management. Post-event geomorphological evidence from different sources (such as published articles, reports, and social media) was collected to further assess the impact of these events. This analysis contributes to the assessment of medicane characteristics, their effects, and their influence on the coastal ecosystems.

3. Results

Here, we describe the development of a Web-GIS platform for the systematical mapping of weather, oceanographic, geomorphological, and microseism data associated with Mediterranean cyclones, with a specific emphasis on medicane events. The Web-GIS platform contains information on 51 cyclone events that occurred between 1969 and 2023 and has been specifically developed to visualize the features of two main categories of extreme weather systems observed in the Mediterranean basin: tropical-like cyclones and extratropical cyclones.

The Web-GIS platform contains three main datasets for each extreme weather event:

(1)

Mesoscale features, such as mean sea level pressure (MSLP), sea surface temperature (SST), seismological data, and hydrodynamic parameters;

(2)

Geomorphological evidence of the impact of past medicanes;

(3)

Field evidence of flooding and coastal erosion as observed through direct surveys and remote sensing.

The MSLP data are reported as isobars at the early stages of each event. As described in Section 1, the minimum sea level pressure can be used to help classify a given weather event and has been used in the literature to produce composite maps that are representative of specific circulation patterns, such as the Cyprus Low (CL) [62], or to identify the cyclone tracks [63].

A cross-correlation between MSLP and seismic data was performed to provide a seismic perspective on cyclone dynamics. The data were collected from seismic stations installed near regions that are closely associated with storm surges generated by medicanes. Spatial analyses were also performed to determine the location and track of the cyclone during its lifetime. The locations constrained by the amplitude-decay-based grid search approach and array analysis were validated with the actual locations of medicanes obtained from satellite images, which were collected from https://search.earthdata.nasa.gov/ (accessed on 1 May 2024). Specifically, we utilized the MODIS/Aqua Surface Reflectance Daily L2G Global 500 m SIN Grid Near Real Time (NRT) product (short name MYD09GHK). This product provides one to seven bands that serve as input for generating several land and atmospheric products such as daily surface reflectance, vegetation indices (VIs), thermal anomaly, and snow/ice/clouds. Using the band-1 channel, we analyzed the spiral cloud patterns of the cyclone’s position, and 1-day averages of mean sea level pressure (MSLP) isobars were overlapped for the cross-validation of cyclone positions on the corresponding date (Figure 7).

The Web-GIS platform can be used to display remotely sensed SST data to analyze the thermal changes prior to medicane occurrences (Figure A1). Thermal drops were observed prior to cyclogenesis [15], and decreases in SSTs are typically observed along the track of medicanes. Miglietta et al. [64] and Pytharoulis [12] showed that SSTs play a significant role in controlling the intensity and longevity of cyclones, while Noyelle et al. [65] showed that SST has a minor influence on cyclones’ tracks but a strong influence on their intensities. Increased SSTs tend to promote tropical transitions due to stronger low- and upper-level warm cores as well as lower pressure minima [1]. Hydrodynamic parameters were extracted from the reanalysis datasets of Copernicus satellites and ERA-5; these included geodetic sea level, significant wave height, and wind speed. Raster maps were geotagged at specific points along the cyclone tracks to help visualize the changes in these hydrodynamic parameters. These parameters are particularly important for measuring the intensity and describing the characteristics of cyclone events [66,67]. In particular, there are crucial differences in the distribution of significant wave heights during medicanes and extratropical cyclones. During medicane events, the significant wave height exhibits a quasi-symmetric structure similar to that observed in previous studies of tropical cyclones [68], highlighting that unimodal sea states occur mainly to the right of the cyclone eye with respect to the direction of propagation of the storm, while crossing sea states are confined to the left sectors of medicane tracks [69]. These observations have only been superficially investigated in the scientific literature and require deeper analysis; this is particularly important considering that the hydrodynamic features of sea states during medicane events are important for assessing the geomorphological effects of medicanes along coastal areas [18].

4. Discussion

Web-GIS platforms have been used as a tool for the mapping of geomorphological features in different fields [70,71]. Here, direct observations of major landforms detected after cyclone events are recorded using geotagged pictures. These pictures can be visualized by querying the point associated with the relevant landform locations. The main geomorphological evidence of cyclone impacts is related to coastal flooding and debris flow during the impact of Medicane Zorbas and Medicane Ianos, respectively. Zekkos et al. [50] published a Web-GIS platform that reported the effects of Medicane Ianos; they focused on displaying pictures of landslides and debris flow in the form of geotagged points. There were also many direct observations of the geomorphological impact of Medicane Ianos (Figure 8).

Geomorphological surveys were conducted before and after different Mediterranean hurricanes and extratropical cyclones in southeastern Sicily [18,44]. Geomorphological mapping was performed on the Arenella sandy coast to identify the main landform that characterized this area. TLS surveys performed before and after cyclone events were used to generate DEMs of the coastal landscape. Our findings revealed substantial variations in the elevation of the sandy coast of Arenella following cyclone Helios (8–11 February 2023). DEM of Difference analysis revealed that the area most greatly affected by storms during the Helios event was the coastal area. Specifically, Helios caused significant erosion on the sandy Arenella coastline, primarily along the foreshore area and the nearby back lagoon, resulting in 0.07 km³ of sediment loss. Similar behavior was previously observed in studies on the impact of Medicanes Qendresa and Zorbas [18]. The flooding limits mapped during Medicanes Qendresa and Zorbas are comparable with the flooding limit observed during cyclone Helios. The coastal erosion that occurred during this event significantly altered the coastal landscape and had a strong impact on sediment balance and redistribution. We also observed significant sediment accumulation in specific areas, suggesting that the redistribution of sediments within the coastal zone was a consequence of cyclone Helios.

These observations were reported as isolines of cumulative sedimentation/erosion on the Web-GIS platform (Figure 9). Temporal elevation data revealed a significant rise in water levels, resulting in extensive coastal flooding. Importantly, the observed inundation surpassed the levels typically observed during the conventional seasonal (i.e., extratropical) storms that affect Sicily, underscoring the exceptional nature of Mediterranean hurricanes and hybrid cyclones (such as Helios) as well as their heightened impact on coastal areas.

The high energy of these events is reflected in the out-of-size deposits transported by waves in the nearshore areas. It should be noted that the reliability of this form of evidence is debatable since they are associated with both tsunami and storm events [54,72,73,74]. Various surveys have been performed since 2009 using UAV and TLS systems to map coastal boulders found in the Greek quarry of Santa Lucia [44,75,76]. The position of the coastal boulders is reported in the Web-GIS platform as colored isolines, with each color representing different survey periods. Each boulder is labeled following the conventions described in Scicchitano et al. [75]. By comparing the initial and subsequent positions of boulders in Santa Lucia, we found that boulder K was displaced from its initial position by 16 m during Medicane Qendresa, while boulder P was displaced by about 8 m during cyclone Helios (2023). This analysis can help determine the magnitude and direction of the displacement induced by different kinds of cyclones (Figure 10). As highlighted by studies on the sandy coast of Arenella, Mediterranean hurricanes and hybrid cyclones appear to be characterized by higher energies than seasonal storms, resulting in significant boulder displacements. Boulder displacements are important in helping to parametrize numerical models for minimum flow assessments of cyclone impacts [77,78].

Our work is constantly undergoing improvements. In particular, this study is focused on visualizing the weather data and geomorphological features associated with Mediterranean cyclones in a relational geodatabase that feeds into a web-based application. Future work could incorporate SAR “Synthetic Aperture Radar” and other satellite datasets to better understand the dynamics and impacts of these events. Additionally, we will provide WMS services that could be used to visualize the layers in other web platforms. A download and upload section will be incorporated, allowing different users to share data. This approach will facilitate research collaborations and contribute to the advancement of geomorphological studies associated with extreme Mediterranean events.

5. Conclusions

This study describes the development of the first Web-GIS platform dedicated to the study of Mediterranean cyclones, with a particular emphasis on medicanes. The interactive nature of the web map allows users to explore different layers, examine geotagged images with relevant information, and measure distances. This tool can help researchers obtain valuable insights into the impacts of extreme weather events on geomorphological features and support decision-making processes related to hazard assessment, environmental planning, and mitigation strategies. Specifically, our Web-GIS platform can be used to help accomplish the following:

(1)

Map the position and track of historical medicanes by using innovative microseism-based methodologies;

(2)

Display the weather features associated with Mediterranean cyclones;

(3)

Map the geomorphological landforms associated with cyclone events;

(4)

Map the coastal erosion and boulder/debris accumulation associated with medicanes.

Through advanced remote sensing, GIS analysis, and the development of a relational geodatabase and Web-GIS platform, we have produced a tool capable of providing critical information to decision-makers and promoting collaboration between stakeholders. These findings contribute to the current scientific knowledge basis, encouraging future management strategies and conservation efforts aimed at safeguarding coastal ecosystems and human communities in the face of cyclone-induced coastal hazards.

In addition, the new data on coastal flooding limits and sediment erosion and accumulation obtained during several medicane events (i.e., Qendresa, Zorbas, Apollo) provide invaluable insights into the unique characteristics of these meteorological phenomena and their impacts on coastal environments. By understanding the underlying mechanisms that drive increased inundation levels as well as the sediment dynamics associated with medicanes, we can enhance our predictive capabilities and develop effective strategies to mitigate potential risks and long-term impacts on vulnerable coastal regions. In particular, there is a lack of geomorphological studies on Mediterranean cyclones compared to meteorological and climatological studies; it is important for additional geomorphological studies to be conducted in order to provide fundamental observational data that can further enhance our understanding and ability to predict these extreme meteorological phenomena.