Shoreline erosion due to anthropogenic pressure in Calabria (Italy)

ABSTRACT

In the second half of the twentieth century, after the Second World War, a considerable migration from inland to coastal areas was observed in many territories, especially in the Mediterranean. This process has caused the expansion of existing coastal settlements and the construction of numerous new towns, often substituting beaches and dune systems. All this has often altered the equilibrium conditions of several coastal ecosystems thereby triggering shoreline erosion processes. This issue is particularly important in territories such as Calabria, a region of southern Italy subject to considerable anthropogenic pressures and characterized by over 750 km of coast. The paper analyses the correlations between shoreline erosion and anthropogenic pressure in Calabria. For this purpose, historical and current cartographic Open Data, such as shapefiles, cartography, and satellite imagery, were analysed using QGIS. The analysis showed a strong correlation between the anthropogenic pressures, in terms of the expansion of the inhabited centres and, therefore, of the man-made areas near the shoreline, and the triggering of shoreline erosion processes. Finally, this analysis is of interest in the field of coastal area planning and management, and it is easily applicable and replicable as it is based on Open Data and free software.

KEYWORDS:

• Shoreline erosion
• anthropogenic pressure
• inhabited centres expansion
• man-made areas increase
• shoreline changes
• remote sensing

Introduction

Anthropogenic pressure is a most important issue, which currently affects many coastal territories (Alvarez-Cuesta et al., 2021a, 2021b; Bianco & García-Ayllón, 2021; H. Chen et al., 2022; Dias et al., 2013; Luijendijk et al., 2018; Martínez et al., 2020; Rodríguez-Santalla et al., 2021). This process has accelerated in the second half of the twentieth century, after the end of the Second World War, due to a considerable migration from inland to coastal areas which was observed in most coastal territories in Mediterranean areas (Barragán & de Andrés, 2015), and especially in Italy (Romano et al., 2017). Indeed, about 20% of the world population lives within 25 km of the coast (Rangel-Buitrago et al., 2018) and the number of coastal cities has quintupled in the last 70 years (Stronkhorst et al., 2018).

Anthropogenic pressure concerns both coastal and river territories and has often altered the equilibrium conditions of several river and coastal ecosystems triggering shoreline erosion processes (EUCC (European Union for Conservation of Coast), 1993; Komar, 2000; Van Rijn, 2011; Acciarri et al., 2016; Barbaro, 2016; Barbaro et al., 2017; Valsamidis & Reeve, 2017; Williams et al., 2017; Mavromatidi et al., 2018; Mentaschi et al., 2018; Anthony et al., 2019; Besset et al., 2019; Ozpolat & Demir, 2019; Tomasicchio et al., 2020; Bombino et al., 2022; Foti et al., 2022bb).

In coastal territories, the anthropic interventions have concerned mainly the expansion of existing coastal settlements and the construction of numerous new towns, often substituting beaches and dune systems (Aguilera et al., 2020; Barbaro et al., 2022; Brazner et al., 2018; Pagán et al., 2016; Pye, 1990; Yi et al., 2021; Zhai et al., 2020). Other important anthropic interventions concern the construction of port and coastal structures (Barbaro, 2013; D’Alessandro et al., 2011; Miduri et al., 2017; Prumm & Iglesias, 2016; Sarma, 2015). All this makes these areas more vulnerable in terms of the action of floods and sea storms (Barbaro et al., 2020; Fiori et al., 2014; Mori et al., 2021; Scionti et al., 2018), whose effects are amplified in the case of concurrent events (Barbaro et al., 2019b; Canale et al., 2021, 2020; Zellou & Rahali, 2019).

In the river territories, the anthropic interventions have mainly concerned the narrowing and concreting of river beds, the construction of dams and weirs, sediment withdrawal, and underground extractions that cause subsidence (Andredaki et al., 2014; Fortugno et al., 2017; Foti et al., 2020; Nguyen & Takewaka, 2020; Valderrama-Landeros & Flores-de-Santiago, 2019; Versaci et al., 2018; Walling, 2006; Zema et al., 2014).

In addition to anthropogenic pressure, various natural factors can also alter the natural equilibrium of coastal areas. The main natural factors include the wave climate (Almar et al., 2015; Bacino et al., 2019; Barbaro et al., 2013; Günaydın & Kabdaşlı, 2003; Ranasinghe et al., 2004; Thomas et al., 2011; Wang et al., 2020) and the sedimentary balance between longshore and river transport (O. A. Dada et al., 2018; Barbaro et al., 2019a, 2012, 2014; Foti et al., 2019; Li et al., 2014; Marchesiello et al., 2019; Ngowo et al., 2021; O.A. Dada et al., 2015).

This balance influences the shoreline position, which is the most important geoindicator of coastal evolution (Boak & Turner, 2005). Remote sensing and GIS (Geographical Information Systems) techniques allow us to identify accurately the shoreline position and to estimate the shoreline changes over the years (Aiello et al., 2015; Alesheikh et al., 2007; Anand et al., 2016; Apostolopoulos & Nikolakopoulos, 2021; Awad & El-Sayed, 2021; Bera & Maiti, 2019; Boumboulis et al., 2021; Cracknell Arthur, 2018; Fatima et al., 2018; Hashmi & Ahmad, 2018; Joevivek et al., 2018; Louati et al., 2015; Maiti & Bhattacharya, 2009; Matin & Hasan, 2021; Mills et al., 2005; Moore, 2000; Moussaid et al., 2015). The identification of the shoreline position can be done manually by photointerpretation techniques, or by automatic extraction algorithms (A.C. Teodoro et al., 2009; Asib et al., 2018; George et al., 2015; Hossain et al., 2021; Maglione et al., 2014; Pardo-Pascual et al., 2012; Sekovski et al., 2014; A. Teodoro et al., 2011).

From this point of view, a key element is the cartography data available, such as historical cartographies, aerial photos, orthophotos, satellite images and UAV and Lidar (Light Detection and Ranging or Laser Imaging Detection and Ranging) surveys (Ayadi et al., 2016; Braga et al., 2013; Gonçalves et al., 2019; Mao et al., 2021; Le Mauff et al., 2018; Nicolae-Lerma et al., 2019; Palazzo et al., 2012; Papakonstantinou et al., 2016; Robertson et al., 2004). Of these data sources, until the 1970s the historical cartographies were the primary source of information and, currently, they are still widely used to carry out historical analyses, coupled with remote sensing data (Ganasri & Ramesh, 2016; John & Scott, 2010; Xuejie & Damen, 2010).

On the other hand, the use of satellite images began in 1972 with the Landsat program, which currently is in its ninth mission launched in September 2021. Satellite images can be classified according to the resolution into: high resolution (0.5–5 m), such as IKONOS-2, Worldview-2, RapidEye, Corona, Pleiades (Bayram et al., 2004; Duarte et al., 2018; Maglione et al., 2015; Pantanahiran, 2019; Vassilakis et al., 2016); medium resolution (5–20 m), such as Sentinel-2,SPOT, IRS 1 C/1D and TERRA-ASTER (Cabezas-Rabadán et al., 2019; Chen & Chang, 2009; Ratna & Bijker, 2019; Sheik & Chandrasekar, 2013; Zifeng et al., 2018); and low resolution (>20 m), such as IRS – LISS-3 and Landsat (C. C. Chen et al., 2019; Kumaravel et al., 2013; Mitra et al., 2017; Nan, 2018; Yan et al., 2019).

Satellite images have been widely used in recent times for three main reasons: most of them are available free, they interface well with GIS and are provided in infrared spectral bands, therefore the water-land interface is well defined (Aladwani, 2022; Ferreira et al., 2021; Hagenaars et al., 2018; Konlechner et al., 2020; Qiao et al., 2018; Toure et al., 2019; Vos et al., 2019). Many free satellite data are available on the United States Geological Survey (USGS) website, on the Copernicus portal, and on the Google Earth software. The last one provides high resolution freely available satellite imagery taken in different time periods (Prasita, 2015; Paravolidakis et al., 2016; Mörner and Klein, 2017; Martínez et al., 2018; Senevirathna et al., 2018).

The paper analyses the correlations between shoreline erosion and anthropogenic pressure in Calabria. This is an interesting case study due to its geomorphological peculiarities and due to the considerable anthropogenic pressure caused by uncontrolled urban sprawl (Aceto et al., 2016; Fiorini et al., 2019; Cantasano et al., 2020; Foti et al., 2022aa,b). The analysis is divided into the following phases: classification, quantification, and comparison between current and 1950s man-made areas; analysis of the shoreline changes and analysis of the possible correlations between the increase in man-made areas and the triggering of shoreline erosion processes. For this purpose, historical and current cartographic data of the Open Data sections of the Italian and Calabrian Geoportals and of Google Earth, such as shapefiles, cartography, and satellite imagery, were analysed using QGIS.

Study area description

Calabria is a region of southern Italy which represents the tip of the famous Italian boot. It is located exactly in the centre of the Mediterranean Sea, with latitude between 38 and 40° N and with longitude between 15 and 17° E, and is enclosed by two seas, the Tyrrhenian, and the Ionian, by the Strait of Messina and by the Gulf of Taranto (Figure 1). Each of them has different climatic characteristics and different fetch extensions. Regarding climatic characteristics, the Ionian coasts are mainly exposed to the winds of Scirocco, South-East, and Grecale, North-East, while the Tyrrhenian coasts are mainly exposed to the winds of the Mistral, North-West. Regarding the fetch extensions, it varies from a few tens of kilometres within the Strait of Messina and the Gulf of Taranto to several hundred kilometres along various directions in the Ionian and Tyrrhenian seas. These differences cause a high variability of weather and sea conditions between the various coastal areas that influence coastal dynamics. The main difference is that the average and frequent wave conditions are slightly higher in the Ionian coast, while the exceptional wave conditions are much greater in the Tyrrhenian coast. In addition, in the Tyrrhenian coast the intense wave conditions are concentrated along a few directions, coming mainly from the north-west, and there are no secondary and tertiary sectors. In contrast, in the Ionian coast, the intense wave conditions can come from different directions, varying between north-east and south-east, and in different locations, there are secondary and tertiary sectors. From the seasonal point of view, in autumn in most of the Calabrian coasts, except for the Gulf of Taranto and the southern Ionian coast, the maximum values of significant wave height are higher than in winter. In addition, the danger of the Tyrrhenian Sea concerning wave climate is generally limited to autumn and winter, while in the Ionian Sea, it is extended to the whole year (Foti et al., 2022). Furthermore, Calabria is a microtidal environment so tidal excursion is negligible (Sannino et al., 2015).

Figure 1. The Calabrian region (shown with red polygon), in the center of the Mediterranean Sea.

From a morphological point of view, Calabria is characterized by hills and mountains, while the flat lands are less than 10% of the entire territory. The main massifs are Pollino, Sila and Aspromonte, all with a maximum altitude in the order of 2000 m. The main coastal plains are that of Sibari, on the Ionian coast in the Gulf of Taranto, and those of Lamezia Terme and Gioia Tauro, both on the Tyrrhenian coast. Its narrow and elongated shape means that it has about 750 km of coastline, with an alternation of beaches, mainly sandy and pebbly, and high coasts, with the main headlands being those of Capo Rizzuto, on the Ionian coast, and of Capo Vaticano, on the Tyrrhenian coast.

Furthermore, Calabrian rivers (locally called “fiumare”; Sabato & Tropeano, 2014; Sorriso-Valvo & Terranova, 2006) are generally characterized by a torrential and irregular hydrological regime, with extensive dry periods and frequent sudden floods caused by short and intense rainfall. Also, many of these rivers have very wide beds with coarse grain size. This combination of hydrological and granulometric characteristics causes high solid transport, and the relative variations can alter the coastal dynamics and the shoreline evolution near the river mouths, as shown by Barbaro et al. (2019a) and by Foti et al. (2019) in the case studies of mouths of Petrace River and Sant’Agata River, respectively.

Methodology

The main aim of the proposed methodology is to analyse the correlations between coastal eroded areas and anthropogenic pressure in Calabria. The methodology, developed using QGIS software, was divided into four phases and was applied at the municipality level. In detail, due to its considerable coastal length, Calabria has 116 coastal municipalities, 71 of them are on the Ionian coast, 3 of them are on the Strait of Messina and 42 of them are on the Tyrrhenian coast. Among these municipalities, those with negligible coastal length, less than 1 km, and those located on promontories with no beach except for small pocket beaches, were not included in the analysis. Therefore, 11 municipalities were excluded, 8 of them on the Ionian coast and the other 3 on the Tyrrhenian coast. Therefore, the analysed municipalities comprised 63 on the Ionian coast, 3 on the Strait of Messina, and 39 on the Tyrrhenian coast.

The four phases were:

1. Acquisition of historical and current cartographic data available, such as shapefiles, cartography, and satellite imagery.

2. Classification, quantification and comparison between past and present man-made areas.

3. Analysis of the shoreline changes.

4. Analysis of the possible correlations between the increase in man-made areas and the triggering of shoreline erosion processes.

The input data of the first phase were the shapefiles of current municipal boundaries and the shapefiles of the shoreline and of the inhabited centres, both dated 1954 and both digitized based on CASMEZ, “Cassa del Mezzogiorno”, cartography of 1954 in scale 1:10,000. All these data were taken from the Open Data section of the Calabrian Geoportal (http://geoportale.regione.calabria.it/opendata, accessed on 15 February 2022). Another input data was the shapefile of the Corine Land Cover fourth level of 2018, available in the Open Data section of the Italian Higher Institute for Environmental Protection and Research (https://groupware.sinanet.isprambiente.it/uso-copertura-e-consumo-di-suolo/library/copertura-del-suolo/corine-land-cover/corine-land-cover-2018-iv-livello, accessed on 15 February 2022), where the current inhabited centres are highlighted. This shapefile was obtained through photointerpretation of satellite images in the Land thematic area of the Copernicus program and was characterized by a scale of 1:100,000, a minimum cartographic unit for the coverage of 25 hectares and a minimum width of the linear elements of 100 metres. Finally, the last input data were the most recent Google satellite imagery, provided by Google Earth Pro and all relating to the years between 2019 and 2021, depending on the examined location. In detail, in most cases the images were relative to 2021. Exceptions were the northern part of the Gulf of Taranto, the central-southern part of the Ionian Sea, and the northern part of the Tyrrhenian Sea where the images were mostly relative to 2019 and, in a few cases, to 2020. These images may be affected by uncertainties regarding their horizontal accuracy (McRoberts, 2010; Nikolakopoulos & Dimitropoulos, 2017; Potere, 2008). Indeed, Google Earth positional accuracy is not fixed but varies over time. This may be referred to the process of updating Google Earth by replacing the original images with better resolution images. Research carried out by Potere (2008) in 109 cities around the world showed a root-mean-squared error (RMSE) in the order of a few tens of metres. Benker et al. (2011) obtained a RMSE of over 2.5 m in the Big Bend region (Texas, USA). Mohammed et al. (2013) obtained a RMSE of about 1.6 m in Khartoum (Sudan). Farah and Algarni (2014) obtained a RMSE of over 2 m in Riyadh (Saudi Arabia). Pulighe et al. (2016) obtained an overall horizontal positional accuracy close to 1 m in Rome (Italy). Goudarzi and Landry (2017) observed a horizontal accuracy between 0.1 and 2.7 m in Montreal (Canada). Therefore, CASMEZ cartography, Corine Land Cover fourth level, and Google Earth satellite imagery have different accuracies. These data are subject to uncertainties during their overlap on QGIS, carried out in the second phase. To limit the uncertainties, some control points were used, corresponding to fixed points present in all the data such as buildings, roads etc. In this way, the uncertainties were contained within the order of a metre and these values are in accordance with the objectives of this paper, as described below.

In the second phase, the shapefiles of the inhabited centres of 1954 were superimposed on the Corine Land Cover fourth level of 2018 and on the most recent Google satellite imagery to classify, quantify and compare the current and past extension of inhabited centres. The criterion adopted to classify inhabited centres concerned their distance from the shoreline. In detail, three classes were identified: the first class includes the inhabited centres less than 100 m away from the shoreline, the second class includes the inhabited centres that are between 100 and 1000 m from the shoreline, and the third class includes the inhabited centres over 1000 m away from the shoreline.

The third phase involved the comparison between the 1954 shorelines and the shorelines manually digitized based on the most recent Google satellite images, all related to the years between 2019 and 2021, depending on the examined location as described above. The analysis was carried out and aggregated at the municipality level, only in municipalities where there has been an increase in man-made areas compared to the 1950s. The manual digitalization of the most recent shorelines was carried out on an eye altitude of 200 m, corresponding to a scale greater than 1:1000, on Google Earth Pro using its spatial analysis tools. The baselines were placed at fixed points such as promenades, roads, and structures and, where these fixed points are very distant from the beach, the baselines were placed at the foot of dune systems. After digitization had occurred, shorelines and baselines were saved as kml files, then imported and saved on QGIS as shapefiles. The projection coordinate system used was the WGS84/UTM zone 33 N (EPSG 32633). All the analysed files have been saved and processed with this system.

This phase is generally characterized by various uncertainties, especially in georeferencing and orthorectification processes, in the identification of the wet/dry line or other similar lines, and regarding the impact of sea storms (Boak & Turner, 2005; Hapke et al., 2010). In this case, uncertainties can be quantified according to Del Rio and Garcia (2013). In detail, the following errors were estimated: physical, scanning and georeferencing. The reference line chosen was the wet/dry line, the cartography data is all related to the summer period and no storm conditions were observed in any of the data, so the effects of seasonal variation and of individual sea storms on shoreline change are negligible. Also, the physical component of the error was estimated using the formula of Allan et al. (2003), which depends on the average and maximum values of the tide height and on the beach slope. To estimate the tide excursions, the recordings of the tide gauges of Crotone and Reggio Calabria were analysed, and the Tide Tables of the Italian Marine Hydrographic Institute (2020) and the scientific papers were consulted, especially Sannino et al. (2015). The average tide height values do not exceed 50 cm, the minimum tide height values do not exceed −70 cm and the maximum tide height values do not exceed 80 cm. The beach slope was estimated using the QGIS Profile tool plugin based on the 1 m side square mesh LIDAR DTMs available on the Italian Geoportal (http://www.pcn.minambiente.it/mattm/, accessed on 15 February 2022), and the values of the examined locations varied between 1 and 15%. So, the estimated error was between 1 and 15 m assuming maximum tide height conditions, and was between 1 and 14 m assuming minimum tide height conditions. It should be noted that these are very precautionary values, as the times of the satellite image was not known and, consequently, it is not possible to know the tide conditions at these times. Also, the scanning error is less than one metre as the scale is greater than 1:1000. On the other hand, regarding the georeferencing error, the use of baselines as control points contained the error within a few tens of cm. Finally, since the aim of the paper is the identification of the shoreline changes, but not their precise quantification, an accuracy in the order of one metre was considered for estimating the shoreline position and the shoreline changes.

In the last phase, a cross-statistical analysis of the results of the previous phases was carried out to verify whether there is a correlation between the increase in man-made areas, in terms of the expansion of inhabited centres, and the triggering of shoreline erosion processes. Firstly, for each municipality the maximum shoreline retreat value was considered, not counting retreat in the order of a few metres. Furthermore, the coastal erosion intensity was classified according to a scale with four classes: slight erosion, for maximum retreat of up to 20 m; moderate erosion, for maximum retreat of between 20 and 50 m; intense erosions, for maximum retreat of between 50 and 100 m; severe erosion, for maximum retreat exceeding 100 m. Similarly, for each municipality the maximum value of the shoreline advancement was considered, not counting advancement in the order of a few metres. To estimate the shoreline advancements and retreats in each municipality, some transects were traced where the greatest shoreline displacements were observed.

Results

The results were aggregated at the municipality level and are summarized in Tables 1–3 and Figures 2–5. In detail, Table 1 shows a summary of the municipality with inhabited centers in the 1950s and today, classified according to the distance from the shoreline, and a summary of the inhabited centers area in the 1950s and today. Table 2 shows a summary of the comparison between the municipality with inhabited centers in the 1950s and today, classified according to the distance from the shoreline, and a summary of the comparison between the inhabited centers area in the 1950s and today. Table 3 shows a summary of the classification of coastal erosion processes, according to their intensity.

Figure 2. On the left municipality with inhabited centers in the 1950s, to the right municipality with inhabited centers today, both classified according to the distance from the shoreline.

Figure 3. Inhabited centers in the 1950s, classified according to the distance from the shoreline.

Figure 4. Inhabited centers today, classified according to the distance from the shoreline.

Figure 5. Classification of shoreline erosion processes, according to their intensity.

Table 1. Summary of the comparison between the municipality with inhabited centers in the 1950s and today, classified according to the distance from the shoreline, and summary of the comparison between the inhabited centers area in the 1950s and today.

Inhabited centers	Municipality		Area
	1950s	Today	1950s	Today
< 100 m from the shoreline	32 (30%)	83 (79%)	14.3	256.3
between 100 and 1000 m from the shoreline	9 (9%)	16 (16%)
over 1000 m from the shoreline	64 (61%)	6 (5%)

Table 2. Summary of the number of municipalities where shoreline retreat or advancement have been observed, or that shoreline changes are due to other causes, classified according to the distance from the shoreline.

Inhabited centers	Total	Shoreline retreat	Shoreline advancement	Shoreline changes due to other causes
< 100 m 1950s and < 100 m today	32	20	4	8
100–1000 m 1950s and < 100 m today	9	5	1	3
> 1000 m 1950s and < 100 m today	44	36	2	6
> 1000 m 1950s and 100–1000 m today	14	3	9	2
> 1000 m 1950s and > 1000 m today	6	2	3	1
Total	105	66	19	20

Table 3. Summary of the classification of shoreline erosion processes, according to their intensity.

Inhabited centers	Slight	Moderate	Intense	Severe
< 100 m 1950s and < 100 m today	1	9	8	2
100–1000 m 1950s and < 100 m today			4	1
> 1000 m 1950s and < 100 m today	3	14	15	4
> 1000 m 1950s and 100–1000 m today		2	1
> 1000 m 1950s and > 1000 m today		1	1
Total	4	26	29	7

The analysis showed that in the 1950s most of the inhabited centres were located at a considerable distance from the coast. Indeed, of the 105 analysed municipalities, in just 32 the inhabited centres were located near the shoreline, less than 100 m, covering a total area of less than 15 km² while in 9 municipalities the inhabited centres were located between 100 and 1000 m from the shoreline, and in all the others 64 municipalities the inhabited centres were located at over 1000 m from the shoreline or on heights. Downstream of these 64 inhabited centres there were only inhabited areas or scattered houses, which in 38 municipalities were located less than 100 m from the shoreline, while in the other 26 municipalities inhabited areas or scattered houses were located between 100 and 1000 m from the shoreline. Inhabited areas and scattered houses have been identified in accordance with the quantitative definitions of the Italian Institute of Statistics (ISTAT), which classifies them according to the distance between neighbouring houses. In detail, if this distance is less than 30 then areas are defined as inhabited, otherwise they are defined as scattered houses. Inhabited areas and scattered houses differ from inhabited centres due to the lack of services, such as schools and offices, and of meeting places, such as squares and gardens. Therefore, in this paper inhabited areas and scattered houses have been analysed together. The inhabited centres located near the shoreline were distributed almost evenly between the Ionian and Tyrrhenian coasts. In contrast, most of the inhabited centres located over 1000 m from the shoreline, or situated at a height, were on the Ionian coast. Indeed, of the 32 inhabited centres close to the shoreline, 17 were on the Ionian coast, 2 were on the Strait of Messina, and 13 were on the Tyrrhenian coast. In contrast, 42 of the 64 inhabited centres located over 1000 m from the shoreline or situated at a height were on the Ionian coast and the other 22 were on the Tyrrhenian coast, constituting more than half of the inhabited centres of this part of the coast.

Currently, most of the inhabited centres are located near the shoreline. Indeed, in 83 municipalities out of 105, the inhabited centres are located less than 100 m from the shoreline covering a total area of over 250 km², sixteen times higher than in the 1950s, while 16 inhabited centres were located between 100 and 1000 m from the shoreline and just 6 inhabited centres were located over 1000 m from the shoreline or on heights. Downstream of these 6 inhabited centres there were only inhabited areas or scattered houses, which in 4 municipalities were located less than 100 m from the shoreline, while in the other 2 municipalities the scattered houses were located between 100 and 1000 m from the shoreline. Of the 83 inhabited centres close to the shoreline, 46 were on the Ionian coast, 3 were on the Strait of Messina and 34 were on the Tyrrhenian coast. In contrast, 14 of the 16 inhabited centres located between 100 and 1000 m from the shoreline were on the Ionian coast and the other 2 were on the Tyrrhenian coast, while the inhabited centres located over 1000 m from the shoreline or at a height were equally divided between the Ionian and Tyrrhenian.

Grouping the results by time shows that the number of inhabited centres near the shoreline has considerably grown from 32 in the 1950s, equal to 30% of the Calabrian inhabited centres, up to 83 today, equal to 79% of the Calabrian inhabited centres. The number of inhabited centres located at a distance between 100 and 1000 m from the shoreline has also grown from 9 in the 1950s, equal to 9% of the Calabrian inhabited centres, up to 16 today, equal to 16% of the Calabrian inhabited centres. In contrast, the number of inhabited centres located over 1000 m from the shoreline or on the hills has significantly decreased from 64 in the 1950s, equal to 61% of the Calabrian inhabited centres, up to 6 today, equal to 5% of the Calabrian inhabited centres.

Grouping the results by area shows that on the Ionian coast the inhabited centres near the shoreline have increased from 17 in the 1950s, equal to 27% of the Ionian inhabited centres and to 16% of the Calabrian inhabited centres, up to 46 today, equal to 73% of the Ionian inhabited centres and to 44% of the Calabrian inhabited centres. In contrast, on the Tyrrhenian coast the inhabited centres near the shoreline increased from 13 in the 1950s, equal to 33% of the Tyrrhenian inhabited centres and to 12% of the Calabrian inhabited centres, up to 34 today, equal to 87% of the Tyrrhenian inhabited centres and to 32% of the Calabrian inhabited centres.

Shoreline changes were analysed in 85 out of 105 municipalities. The 20 municipalities where the shoreline changes are due to other causes with respect to the expansion of the inhabited centres and, therefore, to the man-made areas increase, such as erosion at river mouths or the construction of ports, were excluded from the comparison.

Generally, erosive processes prevail in the 85 analysed municipalities. Indeed, in 66 there has been a shoreline retreat while in the other 19 there has been a shoreline advancement, equal to 78% and 22% of the analysed municipalities, respectively. Most of the municipalities where there is a shoreline retreat, 61 out of 66 total equal to 72% of the analysed municipalities, are related to inhabited centres currently located near the shoreline. Of these 61 inhabited centres where there is a shoreline retreat, 41 are newly built settlements and the other 20 municipalities are related to inhabited centres already present in the 1950s that have subsequently expanded. Of these 41 inhabited centres, in the 1950s 36 of them were located over 1000 m from the shoreline and 5 of them were located at a distance between 100 and 1000 m from the shoreline. Regarding the remaining 5 municipalities where there is a shoreline retreat, 3 of them are related to inhabited centres currently located between 100 and 1000 m from the shoreline, and 2 of them are related to inhabited centres currently located over 1000 m from the shoreline.

On the other hand, most of the municipalities where there is a shoreline advancement, 12 out of 19, are related to inhabited centres that have expanded since the 1950s but currently are still distant from the shoreline. Indeed, in 9 of these 12 municipalities the inhabited centres have expanded from being over 1000 m away from the shoreline in the 1950s to a distance of between 100 and 1000 m from the shoreline today, and in another 3 municipalities the inhabited centres are currently still more than 1000 m away from the shoreline. In addition, another 4 municipalities of 19 are related to inhabited centres close to the shoreline as early as the 1950s. Therefore, only 3 cases of shoreline advancement out of 19 are related to inhabited centres near the shoreline and have been built recently.

In the 66 municipalities where shoreline retreats are observed, the intensity of erosive processes was also evaluated according to the scale described above, highlighting that in most of the municipalities, 55 out of 66, there was moderate or intense erosion while in only 4 municipalities was there slight erosion and in 7 municipalities there was severe erosion. However, severe erosion has been observed only in municipalities related to inhabited centres currently located near the shoreline and these municipalities are all on the Tyrrhenian coast. The maximum shoreline advancement in the whole analysed interval was observed in Curinga, on the Tyrrhenian coast, with a value of 120 m, and another significant shoreline advancement was observed at Rocca Imperiale and Cassano allo Ionio, on the Ionian coast, with values of 90 and 70 m respectively, and at Grisolia, on the Tyrrhenian coast, with a value of 80 m. On the other hand, the maximum shoreline retreat in the whole analysed interval was observed at Bonifati, with a value of 200 m, and other significant shoreline retreats were observed at Amantea and Guardia Piemontese, with values of 160 and 150 m respectively. All these inhabited centres were on the Tyrrhenian coast.

Discussion

The results described in the previous section show that in the last 70 years in Calabria there has been a considerable increase both in the number and in the extension of inhabited centres near the shoreline. Furthermore, shoreline erosion processes currently prevail with respect to shoreline advancement. Most of the municipalities where there are erosion processes are related to inhabited centres currently located near the shoreline but not present in the 1950s. On the other hand, most of the municipalities where there is a shoreline advancement are related to inhabited centres that have expanded since the 1950s but currently are still distant from the shoreline. The analysis carried out highlighted that most of the erosive processes occur in the case of inhabited centres located less than 100 m from the shoreline. This distance can therefore be taken as a critical distance.

The results were analysed considering a single anthropogenic driving factor, that is, the increase of the inhabited centres and, therefore, of the man-made areas. Thus, the coastal areas where the shoreline changes are caused by other causes with respect to the expansion of the inhabited centres, such as erosion at river mouths or the construction of ports, were excluded from the comparison for a total of 20 out of 105 municipalities. In 7 of these excluded municipalities there are ports, while in the other cases there are inhabited centres near the river mouths. Regarding the latter case, it is an issue common to various Calabrian rivers due to their geomorphological and climatic peculiarities described above. Indeed, at the mouths of about 20 Calabrian rivers, maximum shoreline retreat values exceeding 100 m are observed, with a maximum value exceeding 300 m observed at the mouth of the Mesima River on the Tyrrhenian coast (Foti et al., 2022).

The analysis highlighted that the anthropization process affected the Tyrrhenian coast more than the Ionian one. This result can be related to the morphological peculiarities of the territory. Indeed, the northern Calabrian Tyrrhenian coast, from Falerna to the Basilicata region, is characterized by a mountainous relief very close to the coast with few flat coastal areas. Therefore, the inhabited centres have expanded close to the coast. On the other hand, on the Ionian coast there is generally a greater distance between the coast and the reliefs, so several inhabited centres have been built away from the coast, often behind the existing dunes.

The anthropization process of the Calabrian Tyrrhenian coast occurred in a disorganized way, without adequate planning and management, also causing the destruction of natural environments and ecosystems (Cantasano et al., 2020). Therefore, much previous research about anthropogenic pressure and shoreline changes in Calabria has focused mainly on the Calabrian Tyrrhenian coast. In particular, Ietto (2001) analysed the evolution of the Tyrrhenian coasts in the second half of the last century. Bellotti et al. (2009) analysed the shoreline evolution of Belvedere Marittimo beginning in 1873. D’Alessandro et al. (2011) analysed a case study of design and management aspects of a coastal protection system between Paola and San Lucido. Ietto et al. (2012) analysed shoreline evolution between Capo Suvero and Gizzeria. Ietto et al. (2018) defined a new Coastal Erosion Risk Assessment Indicator and applied it to the entire Calabrian Tyrrhenian coast from Capo Vaticano to the North. Other important previous research on shoreline changes in Calabria include Foti et al. (2022a), (2022b)). One of these studies describes and classifies the shoreline evolutionary trends at different time scales along the Calabrian coasts in over 50 sample areas with different morphological and anthropogenic characteristics, such as the presence of inhabited centres, scattered houses, ports, coastal defence works, pocket beaches, dune systems, and river mouths. The main result is that the sample areas in the erosion classes prevail over those in the advancement class for very long-term, long-term, and middle-term time intervals and this result may be related to the considerable anthropogenic pressures that occurred in the second half of the last century. The other research, in contrast, evaluates the effects of anthropogenic pressures on the Calabrian dune systems. Therefore, this study, unlike all previous research, analyses the effects of anthropogenic pressure on the shoreline evolution along the entire Calabrian coast considering each coastal municipality, not only along its Tyrrhenian coast or in specific locations. Furthermore, it was not limited to specific issues such as dune systems. In addition, this research strengthens that of Foti et al. (2022b) highlighting the correlation between anthropogenic pressure and shoreline erosion.

Similar research has been carried out in other countries, especially with regards to irregular urban sprawl without an adequate regulatory plan as in this case study, such as China and Brazil. For example, Cai et al. (2022) observed that human-made coastlines account for nearly 70% of the total Chinese coastline and these effects have been especially visible in the last 40 years during the rapid development of China’s economy. Zhu et al. (2022) analysed the effects of intensive human development activities on 15 counties and districts along the coast of the Pearl River Estuary (China). Lins-de-Barros (2017) analysed the case of Região dos Lagos (Brazil), where the urban population grew significantly on the coast between 2000 and 2010, to develop a methodology that integrates physical, socioeconomic and ecosystem dimensions of coastal vulnerability into a useful tool for coastal zone management. Finally, some sample areas were analysed in detail below. The first four sample areas are related to municipalities where shoreline advancements are currently observed compared to the 1950s, while the other sample areas are related to municipalities where shoreline retreats are currently observed compared to the 1950s. The large number of examples is due to the peculiarities of each of them. In detail, the first example concerns a tourist village, newly built entirely behind a coastal dune (Marina di Sibari). The second example concerns the expansion of an inhabited centre existing in the 1950s near the sea but behind a coastal dune (Botricello). The third example concerns the case of greater shoreline advancement compared to the 1950s (Curinga) where there are no houses up to over 1000 m from the shoreline. The fourth example concerns Grisolia, which is one of the few municipalities on the northern Tyrrhenian coast where the inhabited centre is far from the coast and where the shoreline has advanced compared to the 1950s. In contrast, the other examples are related to cases of strong anthropization with high shoreline retreats which, in the case of Bonifati, are greater in Calabria. In these cases, roads and infrastructures have often been built close to or substituting the beach, with areas where the beach is totally eroded or is present only due to the construction of numerous coastal defence works.

The first sample area is Marina di Sibari, a tourist village in the municipality of Cassano allo Ionio in the Gulf of Taranto with an area of approximately 1 km². The village was built entirely after the 1950s, more than 250 m away from the shoreline behind a dune system. In the coastal stretch in front of the village, the current shoreline continues to advance as compared to that of the 1950s, with a maximum value of 70 m (Figure 6).

Figure 6. Marina di Sibari. Large panel: shoreline of 1954 (red line) with background Google satellite image of October 2019. Small panel: overlap between 1954 CASMEZ cartography and Google satellite image of October 2019.

Botricello is a town on the Ionian coast which in the 1950s had an area of just 0.1 km² and was located about 1500 m from the shoreline, with some scattered houses no less than 250 m from the shoreline. Currently, the inhabited centre has an area of about 2 km² but has expanded exclusively behind a dune system. In the coastal stretch in front of the inhabited centre, the current shoreline continues to advance as compared the 1950s, with a maximum value of 65 m (Figure 7).

Figure 7. Botricello. Large panel: shoreline of 1954 (red line) with background Google satellite image of May 2021. Small panel: overlap between 1954 CASMEZ cartography and Google satellite image of May 2021.

Curinga is a town on the Tyrrhenian coast which in the 1950s was located on a hill about 8 km from the shoreline, with some scattered houses no less than 1000 m from the shoreline. Subsequently, the Acconia hamlet was built downstream of Curinga but is located at over 2 km from the shoreline, far behind a large dune system. In the coastal stretch of this municipality, the current shoreline continues to advance as compared to the 1950s, with a maximum value of 120 m (Figure 8).

Figure 8. Curinga. Large panel: shoreline of 1954 (red line) with background Google satellite image of September 2021. Small panel: overlap between 1954 CASMEZ cartography and Google satellite image of September 2021.

Grisolia is a town on the Tyrrhenian coast which in the 1950s was located on a hill over 3 km from the shoreline, with some scattered houses no less than 200 m from the shoreline. Currently, the inhabited centre was developed downstream at more than 1 km from the shoreline. In the coastal stretch in front of the inhabited centre, the current shoreline continues to advance as compared the 1950s, with a maximum value of 80 m (Figure 9).

Figure 9. Grisolia. Large panel: shoreline of 1954 (red line) with background Google satellite image of October 2019. Small panel: overlap between 1954 CASMEZ cartography and Google satellite image of October 2019.

Sangineto is a town on the Tyrrhenian coast which in the 1950s was located on a hill over 5 km from the shoreline, with some scattered houses no less than 100 m from the shoreline. Currently, the inhabited centre has an area of about 0.7 km² and it has almost completely expanded close to the beach, even less than 50 m from the shoreline of the 1950s. The coastal stretch is heavily eroded compared to the 1950s, with a maximum shoreline retreat of 125 m and the beach is totally eroded in several parts. Therefore, various coastal defence works have been built (Figure 10).

Figure 10. Sangineto. Large panel: shoreline of 1954 (red line) with background Google satellite image of October 2019. Small panel: overlap between 1954 CASMEZ cartography and Google satellite image of October 2019.

A similar condition to Sangineto can be observed in Belvedere, a town on the Tyrrhenian coast just north of Sangineto. In the area shown in Figure 11 it can be observed that in the 1950s there were only a few scattered houses. In the second half of the last century the inhabited centre expanded considerably near the shoreline, with buildings built less than 50 m from the 1950s shoreline. Currently, the coastal stretch is heavily eroded compared to the 1950s, with a maximum shoreline retreat of 120 m and the beach is totally eroded in several parts. Therefore, various coastal defence works have been built and the only parts where the shoreline is slightly advanced compared to the 1950s are those upstream of a groyne.

Figure 11. Belvedere. Large panel: shoreline of 1954 (red line) with background Google satellite image of October 2019. Small panel: overlap between 1954 CASMEZ cartography and Google satellite image of October 2019.

Another similar condition to Sangineto can be observed in Bonifati, a town on the Tyrrhenian coast just south of Sangineto. In the area shown in Figure 12 it can be observed that in the 1950s there were only a few scattered houses. In the second half of the last century the inhabited centre expanded considerably near the shoreline, with buildings built less than 50 m from the 1950s shoreline. Currently, the coastal stretch is heavily eroded compared to the 1950s, with a maximum shoreline retreat of 200 m and the beach is totally eroded in several parts. Therefore, various coastal defence works have been built.

Figure 12. Bonifati. Large panel: shoreline of 1954 (red line) with background Google satellite image of October 2019. Small panel: overlap between 1954 CASMEZ cartography and Google satellite image of October 2019.

Tortora is a town on the Tyrrhenian coast which in the 1950s was located on a hill over 5 km from the shoreline, with some scattered houses no less than 150 m from the shoreline. Currently, the inhabited centre has strongly expanded and has an area of about 2 km² and it has almost completely expanded close to the beach, even less than 100 m from the shoreline of the 1950s. The coastal stretch is heavily eroded compared to the 1950s, with a maximum shoreline retreat of 100 m and there are parts where the beach is totally eroded. (Figure 13).

Figure 13. Tortora. Large panel: shoreline of 1954 (red line) with background Google satellite image of October 2019. Small panel: overlap between 1954 CASMEZ cartography and Google satellite image of October 2019.

Conclusions

The paper described and analysed the correlations between shoreline erosion processes and anthropogenic pressure in Calabria, in terms of expansion of the inhabited centres and, therefore, of the man-made areas near the shoreline. The choice of Calabria as a case study is related to its geomorphological, climatic, and anthropic peculiarities.

The analysis was carried out using QGIS by comparing historical cartographic sources, relating to the 1950s immediately after the end of the Second World War and before the phase of expansion of the inhabited centres that characterized most of the coastal territories, especially in Mediterranean areas and in the study area, and current cartographic sources.

This analysis highlights that in the last 70 years in Calabria there has been a considerable increase both in the number and in the extension of the inhabited centres near the shoreline, which have increased from 32 in the 1950s to 83 today and from about 15 km² in the 1950s to more than 250 km² today, respectively. Also, there is a strong correlation between the anthropogenic pressures, in terms of the expansion of the inhabited centres and, therefore, of the man-made areas near the shoreline, and the triggering of shoreline erosion processes. Indeed, when comparing the most recent shoreline with that of the 1950s, most of the shoreline erosion processes are observed where new inhabited centres have been built close to the shoreline. On the other hand, most of the shoreline advancements are observed where currently the inhabited centres are distant from the shoreline. In detail, most of the erosive processes occur in the case of inhabited centres located less than 100 m from the shoreline. So, this distance can be taken as a critical distance. Furthermore, the greatest effects of anthropogenic pressure in terms of shoreline erosive processes are observed in the northern Tyrrhenian coast. Indeed, in this area the inhabited centres have expanded very close to the shoreline, both due to the short distance from the mountain and due to disorganized and unplanned urban growth.

This analysis is easily applicable and replicable as it is based on Open Data and free software such as QGIS. Through Open Data, some operational and economic limitations of the past have been overcome, mainly concerning costs, acquisition times and accessibility of these data. Indeed, Open Data are generally available in easily downloadable and editable formats and have no limitations on sharing and use, such as teaching, research, professional purposes etc. Furthermore, Open Data allows the interoperability of systems and the creation of free access platforms that can be used by researchers, decision makers and stakeholders. So, this analysis is of interest both in the field of scientific research and in the field of planning and management of coastal areas and their related defence interventions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was funded by the Public Works Department of the Calabria Region.