1. Introduction
Coastal areas are periodically subject to significant economic and environmental damages due to sea storm-induced processes [
1]. Extreme marine storms may result in severe indirect (e.g., land subsidence, water supply contamination, salt intrusion) and direct (e.g., coastal erosion, flooding, vegetation destruction, damage to property and infrastructures) impacts on low-elevation coastal areas [
2].
Mediterranean coasts are characterized by a very high population density that induces high social vulnerabilities when coastal infrastructures are directly exposed to wave action [
3]. The occurrence of numerous coastal cities, river deltas, islands, low-elevation coastal areas and cliffs over the Mediterranean coastal zone recalls the need for considering local features in coastal planning and possible future impacts associated with climatic change. Today, storms in the Mediterranean Sea are frequent, especially during the winter [
4], and their impact on coastal areas can be highly hazardous. A potential increase in hazards due to Mediterranean cyclones, or hurricanes induced by global warming is suggested by several studies [
5], which document and predict longer, stronger, and more hazardous tropical-like cyclones associated with stronger winds and rainfall, despite a general decrease in frequency of intense meteoric events. Storm surges may display significant differences in various regions of the Mediterranean, due to the morphology of different areas and the characteristics of individual storms [
6]. Several episodes of significant wave height (above 3 m) are often recorded each year in specific sectors of the Mediterranean, such as Lion Gulf and Genoa Gulf [
7]. The increasing trends in frequency of Tropical-Like Cyclones (TLC), also referred to as Medicanes (Mediterranean Hurricanes), on the southern-central Mediterranean lead to increased probability for the development of very high storm waves. In recent years, Medicanes “Qendresa” that on 7 November 2014 affected Malta and Sicily [
8], “Zorba” that on 29 September 2018 affected Ionian Sea–Peloponnese (Greece) (see at link:
https://www.youtube.com/watch?v=hco9DNoc1vY accessed on 6 September 2021), and “Ianos” that on 18 September 2020 affected Ionian Sea and Western coast of Greece [
9] originated huge waves striking the coast. The magnitude of sea surface elevation extremes has increased variably in different Mediterranean regions (mostly Southern Adriatic, Balearic, and Tyrrhenian Sea) during the 21st century [
6].
The Gulf of Naples area (
Figure 1) is affected by subtropical dry-summer climate (Mediterranean climate Cs, according to Köppen-Trewartha classification [
10]), characterized by dry summers, wet and rainy winters with rainfall amount more than three times compared to the summer period [
11]. The surface marine circulation of the Gulf of Naples is mainly dominated by the local wind field [
12,
13,
14]. During the winter months, the most intense winds come from the NNE and NE directions. In spring and fall, the main directions are NE and SW, whereas, during summer, winds are generally weaker than the rest of the year [
15]. This general wind regime is associated with numerous low-pressure systems passing over the Gulf of Naples area, with frequent thunderstorms and wind events during winter and fall. In contrast, in spring and summer, the presence of the Azores and African anticyclones induces a stable and moderate breeze system [
16,
17]. The sea wave characteristics in the Gulf of Naples denotes a strong seasonality, characterized by strong swells in winter and fall and lower wave heights in spring and summer [
12,
15,
18]. Based on analysis using of HF radar data, the average wave heights in the inner sub-basins of the Gulf of Naples range from 1.19 m to 1.70 m, whereas the Tyrrhenian offshore areas yield values ranging from 1.77 m to 2.38 m, with typical wave periods between 6 and 8 s [
19]. Storms in the Bay of Naples mainly occur during autumn and winter. They tend to involve large areas and may last in the offshore sectors for a significant time, so that large waves can develop. These storms are frequently of extratropical type (i.e., related to the oscillation of the polar front), even though, in some cases, storms can also be driven by Mediterranean low-pressure systems [
20]. During such extreme events, wave heights can be significantly higher than average records and often exceed 3 m. Under these circumstances, storm waves directly affecting the waterfront of Naples and its infrastructures are likely to result in very high levels of coastal hazards [
21].
This study presents a multidisciplinary analysis of a severe sea-storm that occurred during the last days of 2020 producing exceptionally high damage along the waterfront on Naples (Via Partenope) (
Figure 1). The aim of this research is to document the importance of the local and microscale hydrodynamic processes in controlling the wave directions near the coastline and enhancing the mean wave height. A better understanding of the processes leading to these effects may provide a support to better address design engineering criteria for future coastal planning and protection.
2. Material and Methods
This study is based on integrated data sets and methods including analysis of meteorological data acquired by weather stations, wave measures by sea buoy, wind and wave analyses, and interpretation of detailed digital elevation model (DEM) and orthophoto derived by UAS (Unmanned Aircraft System) surveys.
The first step in the analysis of the 28 December 2020 sea storm event has been the reconstruction of the meteorological scenario both at a large scale (Europe and central Mediterranean area) and at a local scale. The large-scale meteorological study has been carried out based on forecasting and reanalysis maps derived by meteorological models and EUMETSAT satellite images (see link at:
https://www.eumetsat.int, accessed on 6 September 2021). These data are available from open access International Meteorological Service [
22] and have been used to reconstruct the main features and evolution of the atmospheric systems that caused the storm event discussed in this study, with particular attention to the evolution of wind intensity through time. Local meteorological data, along with tidal and wave data used for this research, have been collected from various institutions:
- -
“Consiglio Nazionale delle Ricerche-Istituto di Scienze Marine” (CNR-ISMAR)–meteorological data by a Campi Flegrei network [
23];
- -
“Istituto Superiore per la Protezione e la Ricerca Ambientale” (ISPRA)–“Rete Mareografica Nazionale” [
24]; meteorological and tidal data in the Port of Naples;
- -
“Istituto Nazionale di Geofisica e Vulcanologia-Osservatorio Vesuviano” (INGV-OV)–“MEDUSA Project”-tidal data by instrumented buoys operating in the Gulf of Pozzuoli [
25]; and
- -
University of Naples “Parthenope”–waves data by instrumented buoy operating in the Gulf of Naples (courtesy of prof. G. Budillon, Parthenope University).
The CNR-ISMAR meteorological monitoring network (
Table 1) is composed of four weather stations, equipped with Davis Instruments–Vantage Pro2 model [
26], located along the coastal area of the Naples district (
Figure 2). The network has been implemented to investigate the influence of meteorological factors on geomorphological coastal processes, such as cliff retreat, landslides, beach erosion, and sea storms. The four weather stations (WS) are described in
Table 2 and
Table 3. The Integrated Sensor suite detects temperature, relative humidity, barometric pressure, rain, and wind (speed and direction).
The sensors of ISPRA meteorological and tidal station WS-05-Port of Naples (40°50’29.06” N–14°16’09.10” E) are described on the institutional ISPRA website and include an ultrasonic anemometer [
27]. The characteristics of the tide gauge sensor of the INGV-OV buoy in the Gulf of Pozzuoli (40°48’32.47’’ N, 14°8’39.33’’ E) are documented on the institutional INGV-OV website (Medusa Project) [
25]. Sea waves data are recorded by an active buoy managed by the University of Naples “Parthenope”, located in the southern part of the Gulf of Naples (
Figure 2) off the coast of Massalubrense (40°37.127’ N, 14°19.415’ E; bathymetry: −50 m). The buoy has a floating body in polyethylene and includes a central metal structure on which the different modules are installed. The system is equipped with a datalogger for data acquisition; it records the measurements on the internal memory and transmits data in real time to a remote computer through a UTMS (Universal Mobile Telecommunications System) modem. The sensor mounted on the turret of the buoy body is based on a GNSS (Global Navigation Satellite System) receiver for the detection of height and direction of the waves. The buoy presents a main exposure angle optimized for wave directions somewhat different from that of study area; therefore, this study is based on an estimation of sea wave motion, based on the transferring of energy from wind to sea surface. The sites of installation of all sensors are reported in
Figure 2.
The marine sector where waves propagate before impacting the seawall of Via Partenope (
Figure 1d) is characterized by a complex seabed morphology. The fluid dynamic processes affecting wave motion, depending on the local seabed morphology, sharply differs from the case of a sandy coast with a medium or low slope. The methodological approach applied in this study takes into account this situation. In detail, spilling breaking type and refraction phenomena have been considered absent or negligible. The three parameters involved in the process are wind speed, wind duration, and fetch length. In our analysis, Dorrestein nomogram [
28] has been used. The most important wave parameters evaluated through the Dorrestein nomogram are H
c (characteristic wave height) and T
c (characteristic wave period). They are considered active up to the depth of −10 m and then undergo a breaking process as the depth decreases below −10 m. As the breaking process occurs very close to the coast, the energy dissipation is not significant.
Characteristic and significant wave motion are model descriptions useful to represent the real wave motion and can be adopted as proxies of the average effects produced by the wave process on coastal areas. As “characteristic wave height Hc”, we indicate the height of most developed waves appearing during the growth of the swell, characterized by relatively regular period and height. As “significant wave height Hs”, we denote the height (constant) of a regular stationary sequence of waves (sinusoidal transverse profile), with height equal to the arithmetic average of the highest third of the waves that compose a specific marine event and cross a given marine area without interference between waves and seafloor.
Gröen and Dorrestein [
28] developed a set of curves showing the characteristic height Hc and period Tc versus fetch and duration. These graphs are distinct from significant height (H
1/3) and mean period (T
z). There is uncertainty concerning the relationship between the visually and instrumentally derived quantities, but some bias (H
c and T
c both being slightly higher than H
1/3 and T
z, respectively) may be likely involved when using this type of graph. However, systematic errors are generally small compared with the random errors in individual observations and the differences are negligible for all practical purposes [
29]. For the estimation of the characteristic wave height H
c in relation to the wind force, known the duration of the anemometric event and the fetch, we used the Gröen-Dorrestein nomogram [
28] and cross-checked the values through Breugem-Holthuijsen nomogram [
30].
UAS (Unmanned Aircraft System) photogrammetry combined with the Structure from Motion (SfM) technique [
31,
32] has been used to analyze the seawall morphology along Via Partenope sections (
Figure 1d). These methodologies represent a well-established method for mapping and creation of digital elevation models (DEMs) and ortho-images. In order to investigate the coastal section involved by the sea storm, a DJI Phantom 4 RTK aircraft with an on-board RTK (Real Time Kinematic) has been used to perform all the aerial imagery acquisitions.
The DJI Phantom 4 RTK UAS mounted with a camera equipped with FC6310R lens (f = 8.8 mm), resolution of 5472 × 3648 pixels. The GNSS RTK equipped UAS simplified the construction of highly accurate digital models through SfM photogrammetry, thus avoiding the need for ground control points (GCPs) [
33] that can be difficult in a busy urban context. In this way, the high precision GNSS, allows the detection of the 3D position of the camera at the time of each capture, within a few centimeters. The position of the camera can be estimated automatically by the firmware and the coordinates are stored internally. Exif metadata have been referenced to the camera. The flight altitude was set to 50 m above the ground with a vertical image acquisition axis and Ground Sampling Distance (GSD) of 1.21 cm/pix. The GSD represents the distance between the centers of two consecutive pixels in the image on the ground [
34]. For a given flying height H (m), the GSD is given by:
where P is the pixel size of the sensor (micro), and f is the focal length.
Ninety-nine images were acquired following a single–grid plan, thus collecting pictures along a given direction only. In this case, the dataset consisted of nadiral imageries, and the entire coastal section was mapped. The SfM photogrammetric processing was performed by Agisoft Metashape
® Professional software package [
35].
The SfM-based photogrammetric processing resulted in a dense cloud point, 3D model, 2D orthomosaic, and DEMs data from the set of overlapping images for the area of interest. The obtained final value of root mean square error (RMSE), expressed in pixel units, was below 1 pixel RMSE for all the performed projections. The high resolution of the orthophotos and the accuracy of the entire post photogrammetric processing allowed for obtaining precise measurements within an accuracy range of ±2 cm.
3. Study Area and Historical Events
On 28 December 2020, the Tyrrhenian coastline of the Campania region (
Figure 1) was struck by a severe sea storm, due to a deep Mediterranean low and related cold front, moving from NW to SE. Particularly, the Gulf of Naples was heavily involved during this sea storm. The recorded wind gusts caused widespread damages to coastal infrastructures, especially because of the action of associated huge sea waves.
The Gulf of Naples represents a marginal basin of the south-eastern sector of the Tyrrhenian Sea (
Figure 1c). It is characterized by a complex bathymetry, including two deep canyons, and submerged volcanic structures [
18]. The southern limit of the Bay is formed by the Sarno River plain and the Sorrento peninsula, represented by the carbonate reliefs of the Lattari Mountains (
Figure 1c). The entire coastal area, from Pozzuoli to Naples, Pompei and Sorrento is heavily urbanized and inhabited by ca. 1.5 million of people that cause an intense anthropic pressure to coastal and marine environments [
36,
37]. Despite this high anthropic impact, the coastal part of the Gulf of Naples is characterized by the occurrence of habitats with high natural value [
38] and numerous coastal archaeological sites [
39,
40]. In this sector, the former beaches and natural environments have been lost due to urbanization and the shorelines are stabilized with revetments and seawalls. The waterfront construction was necessary to protect local houses and the several economic and touristic activities along this coastal zone. The morphology of Naples coastal sector is mainly characterized by volcanic landforms with the Campi Flegrei caldera hills to the north-west and the Somma-Vesuvius volcanic complex to the east, minor coastal plains associated with fluvial processes (e.g., the Sebeto plain), and erosive traces of wave action (cliffs, erosion terraces), resulting in alternating small bays and cliffed promontories [
40]. Depositional sequences of volcaniclastic deposits, reworked by alluvial and/or colluvial processes, are present in the coastal plain of Chiaia [
41]. This area is characterized by an overall subsidence in the last 4000 years with a prevailing prograding/aggrading trend due to the sedimentary input coming from the hillslopes and significant anthropic pressure [
40,
41]. In the near high rocky coast sector, the Neapolitan Yellow Tuff (NYT) unit crops out along the paleo cliffs of the Pizzofalcone promontory and Castel dell’Ovo islet (
Figure 3a).
The study site is located along the Chiaia coastal segment of the Naples seafront between Castel dell’Ovo and Mergellina, particularly at Via Partenope, the area most intensely hit by the storm (
Figure 1d).
The dominant waves mainly come from the SW sector with fetches less than 400 nautical miles for the Gulf of Naples. The medium wave height generally ranges between 0.9 and 2.2 m, although heights of up to 4.7 m have been recorded, especially during the winter. These values may double with a critical wave height of 7.4–8.4 m if we consider the recording data in a period of 50 years for the widest fetch [
42]. Extreme waves characteristics evaluated for Via Caracciolo and Via Partenope sectors and for T = 100 years (T = return period) present a significant wave height H
s =2.55–2.91 m for the wave exposure angle 165°–225°N [
42]. Concerning storm surge processes triggered by extreme swells, local authorities (e.g., Autorità di Bacino Nord Occidentale della Campania) [
43] evaluated H
s100 = 2.91 m and the “Run up” R
100= 3.95 m a.s.l. over coastal infrastructures for T = 100 years.
The local marine setting, where waves propagate before affecting the seawall of Via Partenope, is characterized by an articulated seafloor morphology and by the occurrence of Castel dell’Ovo islet. In front of “Castel dell’Ovo”, the water depth is greater than −10 m up at ca. 150 m off the tuff cliff, and this depth almost reaches the base of the Castel dell’Ovo western bastion (
Figure 3a).
After 1950, sea storms of intensity similar to the one that occurred on 28 December 2020 have affected the Via Caracciolo–Via Partenope waterfront in three cases, namely on 4 November 1966, 23 December 1979, and 11 January 1987 (
Figure 3).
The 1966 sea storm event in the Gulf of Naples was generated by the same cyclonic system triggering the great historical flood of Florence and the highest tide ever recorded in Venice (+194 cm) [
44]. Long lasting south-easterly winds affected Tyrrhenian and Adriatic Seas along the direction of their maximum extension, leading to the development of large waves (
Figure 4b). The 1979 event was represented by a nocturnal south-east windstorm (“Scirocco” wind), that concentrated its action in the Mergellina area with devastating effects on the road platform of via Caracciolo and the sinking of numerous boats and a hydrofoil in the Mergellina harbor (
Figure 3c). The third event occurred on 11 January 1987 (
Figure 3d), with a southerly wind direction in the phase of maximum intensity. This storm presents a clear analogy with the one observed on 28 December 2020, even though the meteorological scenario over Europe was quite different (
Figure 4).
5. Sea Conditions Analysis of the Sea Storm Event of 28 December 2020
The coastal area of Naples hit by the sea storm is showed in
Figure 3a. The entire waterfront of Via Caracciolo/Via Partenope was affected by the storm, but the effects appear to increase in magnitude from Mergellina towards Castel dell’Ovo (from West to East), likely due to the relatively lower protection offered by the Posillipo promontory to the eastern part of the waterfront with respect to specific direction of wave motion (SSW-NNE) (
Figure 9). Via Partenope was the waterfront area most affected by damage, as highlighted in yellow in the inlet image of
Figure 10. The main wave exposure angle is the angular sector encompassing the directions from which the largest waves can affect the study area. This angle (in reference to the north) is determined by taking the half-lines tangent to the promontories that mostly restrict the direction to the open sea, in our case, the western margin of Capri Island (named Punta Arcera) and Capo Posillipo (
Figure 9).
Considering the coastline point P
0, in correspondence of “Via Partenope”, we defined the main wave exposure angle relative to point P
1, where it is possible to consider undisturbed the wave motion, due to the absence of wave energy dissipation by interference with the seafloor. These last elements modify the direction of propagation of wave fronts due to diffraction phenomena. This point is located at the intersection between the axis of the main wave exposure angle and a water depth H of −34 m (
Figure 10).
The value H = −34 m is equal to the maximum characteristic semi-wave length (L/2) measured during the observed meteorological event:
In the case of the study area, the geographical fetch is wide. Most of the directions included in the sector relative to point P1 intercept the coasts of North Africa (with fetch extending over 600 km) and the northern coasts of Sicily (with fetch extending approximately 300 km). This means that a large possibility exists of wind energy transfer to the sea surface and consequent increase of wave dimensions (length and height).
In the Mediterranean area, the barometric lows generating the major wind events, and greatest wave heights, have not a full compliance with the definition of fetch. Particularly, wind usually does not have constant direction and intensity over the entire length of the fetch. For this reason, the geographical fetch has been locally limited to the value of 250 nautical miles (about 460 km) (
Table 4). The barometric conditions on 28 December over the central Mediterranean showed a high rate of exploitation of fetches in the directions of the main wave exposure angle (
Figure 11), as the anemometric event discussed in this study has constantly maintained a direction of about 210°. In this condition, the role of wave motion from angular sectors external to the main wave exposure angle is not significant, which makes the evaluation of the effective fetch unnecessary.
The wind event triggering the sea storm began in the early morning of 28 December. For convenience, we consider 06:00 UTC as the onset time of the event, when the anemometric diagrams (
Figure 8) indicate a significant increase of the wind speed. Wind intensification occurred during almost the entire morning and afternoon of 28 December 2020, until 19:00 UTC. In detail, two distinct wind corridors were active (
Figure 12).
Corridor 1 is relatively wide and is characterized by high speed, SW winds, whereas corridor 2 is relatively narrow and is dominated by SSW winds. Both corridors pointed towards the southern coastal area of Lazio at 15:00 UTC while approaching the coasts of Campania. These corridors acted extensively on the fetches of the main wave exposure angle of P1. The convergence of two different wave motions likely caused a concentration of the sea storm energy, with enhancement of wave motion and the possible development of rogue waves due to random interaction of the two distinct wave trains.
The maximum development of wave action occurred around 19:00 UTC, following a drop in wind intensity when the waves became regular in their profile, and the wave breaking process started, mostly due to the interaction of wave motion with the seabed, without substantial contribute of windshear (
Table 5). This setting favored the concentration of energy in the wave motion and a wave geometry with almost undulatory profile before the breaking process. At the beginning of the windstorm event (t = 0), we consider the wave motion in the Gulf of Naples equal to the one recorded by the Parthenope University ondametric buoy for the time interval before 06:00UTC:
- -
Hs (m) = 1.00,
- -
Ts (s) = 4.1 → L = 25 m,
- -
Dir (°) = 264.
The wave motion is the result of the previous meteo-marine phase, with the presence of swell waves of modest steepness (low ratio H/L), as confirmed by the analysis of the Gröen-Dorrestein [
28]. In fact, a new forming wave characterized by a height H = 1.00 m, is characterized by a period T < 4 s (and shorter wavelength) for duration of the wind event D < 24 h. This is up to the intensity of 30 m s
−1, with delayed times for reaching the height of H = 1.00 m, depending on the intensity of the wind event and the delayed minimum fetch. The swell height usually decays about 25% in 12 h, and 40% in 24 h [
29]. Given the duration of 14 h for the analyzed wind event, we consider that the initial wave motion, after 14 h, is reduced by 30%, as shown in Equation (4).
The starting of the anemometric event induces the formation and the progressive development of new waves. According to the different period of the characteristic waves (previous and recent waves), the sea surface appears rather chaotic, with irregular distribution of waves with various height and period, but, as the waves grow, the development of characteristic waves is observed. The meteorological stations selected for the analysis of the anemometric event and used to evaluate the meteo-marine conditions in front of Via Partenope, are Capo Posillipo–ISMAR (WS-03) and Port of Naples–ISPRA (WS-05) (
Figure 2). The choice is based on the exposure of these two weather stations, on their relative proximity to the study area that is located along the line connecting WS-03 and WS-05.
The real anemometric event, given by the average values of the two weather stations (
Table 6), is transformed into a step diagram (
Figure 13), composed of 14 sub-wind events, each of 60’ in duration (from 06:00 to 19:00 UTC), with a constant wind speed equal to the average value in that specific time interval. Anemometric data (
Figure 13) showed a freshening wind at constant direction, with wind speed that gradually increased from 06:00 UTC to 19:00 UTC without major changes. This almost linear progression allows for a linear interpolation of detected wind speeds over the entire duration of the event (14 h). Hence, it is possible to estimate the characteristic wave height by subtracting 25% of the amount of wind speed increase from the highest wind velocity reached at the end of the storm event and work on the obtained value [
29]. The interpolated wind speeds span from 9.1 m s
−1 (t = 1) to 16.2 m s
−1 (t = 14) (see
Figure 13); then:
For the following analysis, we used a constant wind speed of 14.4 m s−1 over a duration of 14 h.
The reconstruction of the deep-water wave motion (offshore conditions) through the Dorrestein diagram (
Figure 14), using the value above calculated (14.4 m s
−1), allows estimating the characteristic wave height:
- -
Hc = 3.80 m,
- -
Tc = 7.3 s,
- -
Hmax = 1.29 Hs = 4.90 m, and
- -
Tmax = 7.8 s.
These wave height values are confirmed by Breugem and Holthuisjen (2007) wave growth nomogram (
Figure 15), that allows evaluating the significant wave height in:
The absolute maximum height is obtained by adding this value to the wave height recorded at the end of the previous anemometric event, assuming the principle of superposition of effects:
The data measured by the Massalubrense buoy of Parthenope University (
Table 5) are significantly lower than the ones extracted from the Gröen-Dorrestein nomogram. This depends on the smaller fetch as a consequence of the sheltering effect of Capri Island from S-SW winds. Conversely, when S-SW wind is active, unlimited fetches are available at via Partenope.
Tidal Analysis
The wind waves are superimposed on the sea surface level, which oscillates because of astronomical and barometric forcing. Therefore, it is necessary to evaluate the altimetric oscillations of the sea surface to establish the level of the crests and correctly relate them with the coastal infrastructures. The sea level data collected by two tide gauges located in the Port of Naples and Gulf of Pozzuoli were used to evaluate the altimetric oscillations of the sea level during and before the sea storm (
Figure 16).
The maximum uplift of the sea surface level is equal to 0.308 m, including the astronomical and barometric contribution. The maximum astronomical tidal amplitude occurs in the sigitial phase, with the following values for the study area [
43]:
The atmospheric pressure contribution S
Δp(+), the storm wind surge S and the uplift due to breaking wave motion contributions are considered for the definition of the meteorological tide. The sea surface undergoes a change in level of 0.0101329 m for each hPa of variation in the barometric pressure, in reference to the normal atmospheric pressure equal to 1013.2 hPa. In our case, we have:
For the calculation of wind-induced surge, the following relation [
48] to be solved by successive attempts was adopted:
where:
- -
Kp = 3 × 10−6,
- -
G = 9.8065 m s−2,
- -
d = depth of the seabed at the point where the surge is calculated, assumed to be 0 m,
- -
D = limit depth assumed for the continental shelf, assumed to be −200 m,
- -
U = constant storm wind speed (set equal to 14.4 m s−1 for the event under consideration), and
- -
Lp = extension of the continental shelf (distance between d and D) equal to 15,000 m.
The relation has converged by successive attempts to the solution of S ≈ 0.03 m. In the absence of wind shear effects, offshore waves break in sections characterized by a well-defined depth, thus inducing a rise of sea surface level. Along the coastal zone confined between the breakers line and the shoreline, the sea level undergoes a lowering in the wave breaking area (wave set-down) and a subsequent raising towards the shore (wave set-up). The calculation of the two terms (set-down and set-up) can be performed through the following formula:
with:
where:
- -
Hb = Hc,tot = reference wave height at breaking = 4.5 m,
- -
T = period related to reference wave at breaking = 7.8 s,
- -
hb = depth of sea bottom at breaking = 1.3 Hb = 5.85 m,
- -
ΔS = 0.15 hb = 0.88 m,
- -
Sdw = 0.536 Hb(3/2) /g1/2 T = 0.21 m, and
- -
Sup = 0.88–0.21 = 0.67 m.
To evaluate the maximum sea level along the analyzed shoreline we sum up all the different contributions, considering the concomitance of all the factors:
Considering the waves of maximum height (H
1= 5.60 m
)), S
up is slightly higher or equal to 0.80, so:
Given the uncertainty in the sea dynamics when extreme waves occur, we lean towards the most prudential value, considering Stot =1.08 m.
The average level (H = 0.34 m), corresponding to the astronomical and barometric components, is similar to the calculated component, equal to 0.38 m. The maximum uplift of wave crests (L
v) with respect to the mean sea level may be calculated as follows:
7. Discussion
Coastal risk in the Mediterranean area is rising due to growing population, commerce and infrastructure in coastal areas, and the impacts of global sea level rise and warming, associated with increased frequency of sea storms. Major coastal hazards are represented by storm surges and the coastal flooding caused by severe tropical or extra-tropical storms. Storms in the Mediterranean Sea are frequent especially during the winter months and their impact on coastal infrastructures and environment is highly hazardous.
The present study analyzes the extreme sea storm event that occurred on 28 December 2020 in the Gulf of Naples. The exceptional nature of this storm is well represented by the maximum evaluated Run Up, which in the order of magnitude for the events with a return time equal to 100 years [
43].
The event observed on 28 December 2020 was mainly characterized by the sea conditions, including the wind induced wave motion. The overall maximum wind speed measured by the meteorological monitoring network, although intense, was not exceptional, as it can be usually observed at least once in a year. The sea roughness, instead, was of extraordinary intensity. Particularly, the wave crests height was remarkable as the storm occurred in phase with the astronomical tidal maximum in sigitial phase. The damage on the coastal infrastructures was consequently very high.
The results of our study suggest that Castel dell’Ovo western bastion played a major role by interfering with the wave motion in front of Via Partenope, thus causing waves 10–15% higher than normal when the incoming direction is between SSW to SW, as recorded on 28 December 2020. This research also highlights the importance of integrating the analysis of the entire spectrum of factors that control the entity of the storm surge phenomenon. The surge calculated as a consequence of all the factors was 1.08 m. This value, under conditions of unlimited fetch and for equal duration of the wind event, corresponds to a wave height generated by a wind of greater intensity. The analysis of the present wind event through Gröen-Dorrestein nomogram, characterized by a duration of 14 h and an average wind speed (15 m s−1), suggests that a greater wave height of 1.08 m, for the same duration, is generated by a higher wind event with an average intensity of 17.5 m s−1 (+20%).
Our results confirm the results of modeling studies on directional storm surge in enclosed seas [
49] and wind and wave setup contributions to extreme sea levels at a tropical high island [
50]. Usually, surge height depends on the cosine of the angle between the wind direction and the major axis of the narrow gulf. Even though a longer fetch will produce higher storm surge, the exact relationship is complex and dependent on the local physiography, as observed for case studies of the Red Sea and the Adriatic Sea [
49]. Numerical simulations of wind waves, storm-surge, and wave setup indicate that the storm track and the local morphology led to local differences in extreme sea levels, in the order of 1 m along Samoa coastline [
50].
In our case a relevant contribution was given by tidal rise of local sea level. Other studies have also confirmed the importance of the timing of the sea storm landfall relative to local high tide, as observed in storm surge simulations made for Sandy Hurricane which struck New York area in 2012 [
51]. The present study also suggests that, in the case of artificial cost (e.g., when a seawall occupies protects the coastline), wave breaking occurs at the toe of the structure, and wave height needs to be considered independently from the runup [
52]. In these cases, flood risk may be underestimated if storm surge height alone is considered, and wave characteristics need to be incorporated in order to adequately model magnitude and extent of the potential flood [
52].
The results of the UAS survey provided detailed information on the coastal impact of the storm and a detailed 3D reconstruction of the seawall sector heavily damaged during the event [
53]. The 3D topographic data confirmed the occurrence of erosive processes along the seawall during the sea storm resulting in a generalized instability of the calcareous rock blocks at the base of the seawall.
The detailed topographic profiles (
Figure 23) clearly show that the combined effects of seawall reduction and local scale effects on wave heights favored the final flooding of the coastal stretch of Via Partenope and the seawall during the storm event.
8. Conclusions
The risk related to extreme marine events is significantly high along the coastal area of the Gulf of Naples, Italy. On 28 December 2020, the area has been hit by an extreme sea storm, with severe consequences along most areas of Naples’ coastline. In the waterfront area of Via Partenope, Naples, the waves overrun the roadway, causing extensive damage on the coastal seawall, roads, and touristic structures.
In this study, we analyze the meteorological evolution of the sea storm and its effects on the waterfront, suggesting that Castel dell’Ovo western bastion and tuff cliffs have enhanced the local-scale wave magnitude and associate coastal damage, due to induced reflective processes on the waves approaching the coastline. Satellite meteorological data were integrated by accurate estimates of meteorological parameters obtained from weather sensor networks and UAS digital photogrammetry data processed with the SfM technique.
The analysis of the event of 28 December 2020, suggests that our ability in predicting the effects on the coast of a sea storm may be improved by considering all local scale factors, including the duration and intensity of wind events, the tidal component, and wave diffraction and reflection phenomena. For this scope, the application of remote sensing integrated systems, such as UAS surveys and digital photogrammetry, represents an effective support in the assessment of damages, resilience, and vulnerability and the evaluation of adequate risk mitigation measures to be adopted for coastal zone management. The outcomes of the present study also provide a contribute for specialists in construction/design engineering towards the planning of protection structures along the coasts.