Shoreline changes near river mouth: case study of Sant’Agata River (Reggio Calabria, Italy)

The analysis of shoreline changes is very important for coastal planning and management. In territories such as Calabria (Italy), characterized by signi ﬁ cant anthropogenic pressures and various eroded coasts, the knowledge of the shoreline changes, and the factors that in ﬂ uence them, is necessary for management and planning of coastal areas. In fact, shoreline position is one of the most important indicators of coastal dynamics. From this point of view recent advances in remote sensing and GIS techniques allow to estimate with great precision the shoreline changes over the years. The paper analyzes the shoreline changes near the mouth of the Sant ’ Agata River (Reggio Calabria, Italy), carried out through the comparison of various cartography data. Furthermore, the paper analyzes the main factors in ﬂ uencing the coastal dynamics in order to identify possible correlation between these factors and the shoreline changes. The analysis of these factors shows that, in this case study, the rainfall regime has a considerable in ﬂ uence on shoreline change. The methodology described in this paper is particularly useful for better understanding the factors that most in ﬂ uence the coastal balance and, therefore, is applicable to many contexts which are similar to the Sant ’ Agata river mouth.


Introduction
Coastal areas represent the transition zone between sea and land and are of particular importance for the presence of housing settlements. In fact, over 30% of the world (Syvitski, Vorosmarty, Kettner, & Green, 2005) and the Mediterranean (European Union, 2012) population live a short distance from the coastline. The demographic increases and the remarkable anthropization that characterized the second half of the last century have had a major impact on the equilibrium of the coastal areas and this process is visible both near the coastal cities and at the river mouths.
Amongst the human-induced factors (Manca, Pascucci, Deluca, Cossu, & Andreucci, 2013), the increase in anthropogenic pressure, observed in coastal areas over the last 50 years, has increased the vulnerability of the territory under the action of natural events such as floods (Fiori et al., 2014;Scionti et al., 2018), debris flow, storms and coastal flooding (Li, Yamazaki, Roeber, Cheung, & Chock, 2018), or a combination of these (Destro et al., 2018;Barbaro Petrucci et al., 2018). Furthermore, the construction of buildings, infrastructures, ports and coastal defence works are of particular importance. Other important factors are the construction of hydraulic structures interfering with fluvial dynamics such as levees, dams, inert drains from river beds and soil erosion by water (WSE).
Misdiagnosis of the factors listed above can lead to environmental disasters as in the case of Saline Joniche, near the Messina Strait (Barbaro, 2013), or in the case of Badolato, in the Calabrian Ionian coast (Miduri, Foti, & Puntorieri, 2017). Therefore, an accurate estimation of shoreline position and qualitative analysis of the causes of shoreline change (Li, Zhou, Zhang, & Kuang, 2014) is important for coastal zone planning and management (Barbaro, 2016). Moreover, for the Italian and Calabrian coasts, the importance of the aforementioned factors is accentuated by the high coastal development of both territories, over 7500 km of coast for Italy and over 700 km for Calabria, many of them in erosion (Barbaro, Foti, & Sicilia, 2014;Pranzini & Williams, 2013).
The paper, following the analysis of Barbaro, Bombino et al., (2018), describes a case study related to the mouth of the Sant'Agata River, where the analysis of the shoreline changes was carried out through the comparison of various cartography data. Furthermore, the paper analyzes the main factors influencing the coastal dynamics in order to identify possible correlation between these factors and the shoreline changes. In the following sections, after the description of the study area, will be analyzed: wave climate and longshore transport, and river sediment contribution, which depends on variations of hydraulic structures, rainfall time series, land cover and WSE. Finally, a cross-analysis of all these factors will be carried out, to understand the relative influence to shoreline changes.

Site description
The study area is located in the Southern part of Reggio Calabria, a city located in Italy in the southern part of the Calabria region near the Messina Strait (Figure 1), and it is characterized by the presence of both sea and mountains very close to each other and by the presence of "fiumare". These are typical rivers of southern Italy with torrential and irregular regime, characterized by extensive dry periods and with frequent events of sudden flood, generated by short and intense rainfall (Terranova, Antronico, Coscarelli, & Iaquinta, 2009). In Reggio Calabria there are more than 10 fiumare and, not far from the Sant'Agata, there are the Calopinace, Armo and Valanidi rivers, all enclosed in a few kilometers. The study area is heavily anthropized due to the presence, especially, of the airport, between Sant'Agata and Armo rivers, a sport center and various industrial activities (Figures 2-3) (Versaci, Minniti, Foti, Canale, & Barillà, 2018).

Shoreline changes
The analysis of the shoreline changes was carried out through the comparison of various cartography data, which consists of: aerophotogrammetry provided by Italian Military Geographic Institute, orthophotos taken from the Open Data section of the National Geoportal, and satellite imagery provided by Google Earth.
The analysis was divided into three phases as follows. The first phase concerned the manual digitization of the shoreline, for each cartography data and using QGIS for aerophotogrammetry and orthophotos and using the spatial analysis tools of Google Earth Pro for satellite imagery. The second phase concerned the evaluation of the beach width at a transept positioned at the mouth of the Sant'Agata river . Finally, the last phase concerned the determination of shoreline rates of change using end point rate (EPR) and net shoreline movement (NSM) statistics (Table 1).
Regarding the identification of the correct shoreline, it should be noted that there are many reference lines representing the shoreline position (Boak & Turner,   2005). In the framework of a study based on photointerpretation, the choice and extraction of a common line used for different images is required. In this paper, due to the varying oceanographic conditions among the different cartographies, the reference line chosen was the wet/dry line. It has been shown that the wet/dry line closely approximates the High Water Line (HWL, Moore, 2000). Furthermore, in the area under examination the tidal excursion is of the order of tens of centimeters (Sannino, Carillo, Pisacane, & Naranjo, 2015) so the effects on the variation of the shoreline position are negligible. The digitalization of the shoreline was carried out on a scale of 1:1000 on QGIS and on a higher scale on Google Earth Pro. Therefore, the shoreline position has precision of the order of the meter and the shoreline changes have been approximated to the meter. This accuracy is in agreement with the aims of the paper, which concern the evaluate of the erosion and advancement trends, and not their precise quantification.
Regarding the evaluation of the beach width at the transept, to automate the calculation a function which can determine the distance from a given point was implemented on QGIS, using the "field calculator" in the program attribute table as the starting point. Furthermore, the transept originates from a fixed point, represented by the end point of the levee located in the hydraulic left of the Sant'Agata River.
From the analysis of the results shown in Table 1 and in Figures 4-8, it is possible to observe how advancement and erosion phases alternated from 1954 to today, from these data it is clear that: minimum width was observed in 2002 (32 m) while the maximum width was observed in 2015 (82 m).

Wave climate and longshore transport
The wave climate was analyzed starting from the wave data provided by the ABRC-MaCRO software,  developed by HR Wallingford Ltd. This software allowed us to obtain a time series of wave data, starting from the information available at the Met Office database. This database is composed of data reconstructed via the European Wave Model starting from wind field data. The time series obtained using this software extends from 16 October 1986 to 31 March 2006 and consists of 147467 sea states, for each of which significant height, mean and peak periods and direction are available. These data were grouped in sectors of 10°each and in classes with a significant height of 0.5 m each and a further grouping was made in time intervals (Table 2). These intervals were in agreement with the intervals emerged from the analysis of the shoreline changes. Starting with the time series, the following was calculated: frequency of occurrence of   sea state (hereinafter referred to as frequency) ( Figure  9), mean energy flux ( Figure 10) and longshore transport.
From the analysis of the results shown in Table 2 and in Figures 9-10, it is possible to observe that the study area is characterized by modest wave motion: indeed, less than 1% of the recorded sea states exceeds the threshold of 1.5 m and no sea level exceeds the threshold of 3 m. This result is due the morphology of the territory: the study area is located within the Strait of Messina, in an area characterized by small fetch, of the order of ten km. The frequency is concentrated in two main directions: one from the North-West and one from the South. Regarding the first direction, it is associated with a high frequency but has a low energy content due to the small fetches. Regarding the second direction, it is characterized by a lower frequency than the first but has much higher energy content due to exposure to the wave coming from the southern mouth of the Strait.
Longshore transport was estimated using the Tomasicchio, D'Alessandro, Barbaro, and Malara (2013) model (Table 3) and the study area was divided into two sections, one to the north and one to the south of the mouth, due to the different exposure to wave motion ( Figure 11). The southern section has a small beach, of the order of meters, while the northern section has greater width, on average between 20 and 30 m.
From the analysis of the results shown in Table 3 and in Figure 11, it is possible to observe that in the northern section the average value is about 20,000 m 3 /year while in the southern section the average value is about 11,000 m 3 /year, in both cases it is directed from South to North.

River sediment contribution
To study the river sediment contribution, variations of hydraulic structures, rainfall time series, land cover and WSE, estimated using the Gavrilovic (1959) model, were analyzed as for below. Preliminarily it was necessary to identify and morphometrically characterize the catchment area using QGIS and MapWindow and its Watershed Delineation plugin.  1986-2006 1986-1996 1996-1998 1998-2002 2002-2006

Morphometric characteristics of the Sant'Agata River basin
The identification and morphometric characterization phase of the catchment area was carried out by starting with the data available in the OpenData section of the Calabrian Geoportal (http://geoportale.regione.calab ria.it/). In particular, the DEM with square mesh of 5 m and the shapefile of the watercourses were used and the river basin (Figure 12), the perimeter, the main stream length, the maximum, minimum and the average heights, the average slope, the Horton order, the run-off time and the Gravelius index were calculated (Table 4). The basin has an elongated shape, has a considerable altitude difference between mountain and mouth sections, over 1600 m, a high slope, about 40%, and a modest run-off, less than 3 hours.

Hydraulic structures
The latest census of hydraulic structures in the Sant'Agata River dates back to 2010, the results have been implemented in the Archimede information system (Labate, 2010). This system is a section of the territorial information system of Calabria region civil protection.

Rainfall data
There are 2 gauges in the Sant'Agata River basin and in its neighboring areas ( Figure 13). Table 5 shows the registration period, the number of years available, the elevation, the weight, calculated using the Thiessen polygon method (ASCE, 1996;Fiedler, 2003), and the average rainfall and temperature for each gauge. Table 6 shows the average annual rainfall values for each gauge from a sufficiently large period before 1954 to today, divided into time intervals consistent with those identified in the Figure 11. Southern and northern sections in which the study area was divided. paragraph on the shoreline changes. Table 6 also shows the percentage variations, compared to the average rainfall, for each gauge and for each time interval. From the analysis of the results shown in Tables 5-6, it is possible to observe that in each interval the percentage variations are of an agreed sign for both stations. Moreover, the greatest positive variation for both stations were observed in the same period, 2007-2012, with about 16% more for Reggio Calabria gauge and 7% more for Cardeto gauge. Also, the major negative variations for both stations were observed in the same period, 2006-2007, with about 10.5% less for Reggio Calabria gauge and about 17% less for Cardeto gauge. Other significant variations were observed in the 2012-2017 period, between 5 and 5.5% positive for both stations, and in the 1998-2002 period, between 8 and 9% negative for both stations.

Land cover
The land cover data used was from the Corine Land Cover project and relates to the years 1990 to 2012. This was divided into three periods, being 1990-2000, 2000-2006 and 2006-2012, and was freely available on the government agency website "Istituto Superiore per la Protezione e la Ricerca Ambientale (ISPRA)". It should be noted that data before 1990 and after 2012 is not available. Comparing these periods, it is possible to observe that only negligible percentages of the entire surface of the basin have undergone changes (Figure 14).

Soil erosion by water (WSE)
Over recent decades, WSE has become a severe and extended issue affecting all European countries, although to varying degrees. The European Mediterranean countries are particularly prone to erosion, because they are subject to prolonged dry periods followed by heavy erosive rains falling on steep slopes characterized by fragile soils (Grimm, Jones, & Montanarella, 2002, 2003Van der Knijff, Jones, & Montanarella, 1999). In particular, natural conditions and the human impact have made Calabria in southern Italy particularly prone to intense WSE (Terranova et al., 2009). For this reason, river sediment contribution was evaluated using the Gavrilovic (1959) model, which is based on an analytical equation to determine the annual volume of detached soil due to surface erosion. This equation depends on the average yearly precipitation, the average yearly temperature, the drainage area, the average slope of the basin and some coefficients related to the soil protection (a function of the type of vegetation cover), the erodibility (a function of type of rock), the erosion, and the stream network development (a function of the type of basin erosion). The annual river sediment transport averages 20,000 m 3 /year.

Discussion and conclusions
The analysis of shoreline changes is very important for coastal planning and management. This paper describes a case study on the analysis of shoreline changes near the mouth of the Sant'Agata river, which analyzed the Figure 13. Active gauges in Sant'Agata River basin and in its neighboring areas.  possible correlations between the shoreline changes and the main factors that influence it. In particular, the wave climate, the longshore transport and the contribution of river sediments were analyzed, influenced by changes in hydraulic structures, precipitation regime, land cover and WSE. This last factor has been analyzed because the Mediterranean European countries, including the Calabria region, are particularly exposed to the WSE. The analysis of possible correlations between the shoreline changes and the main factors that influence it was carried out using an empirical decomposition method, which is one of the classic models of time series analysis. The method was applied to each factor examined, to compare the evolutionary trend of this factor with the evolutionary trend of the shoreline to assess if there is a correlation between them. This analysis was restricted mainly to the interval 1986-2006, in which data of all the parameters examined are available. In particular, it is possible to observe that river transport is, on average, of the same amount of longshore transport south of the mouth, and in this section the beach is a few meters wide, while north of the mouth is larger, and in this section the beach is between 20 and 30 m wide.
Moreover, due to the morphology of the territory, both the wave motion and, consequently, the longshore transport are modest. During the period 1986-1996, where the shoreline has progressed slightly, longshore transport has assumed the highest values. In the period 1996-1998, where the shoreline advanced at a speed of 3.5 m/year, longshore transport has assumed the lowest values. In the period 1998-2002, where the shoreline has eroded at a speed of over 8 m/year, longshore transport has assumed lower than average values. In the period 2002-2006, in which the shoreline advanced at a speed of 11 m/year, on the other hand, longshore transport has assumed values above the previous interval, but still below average values. Therefore, there seems to be no correlation between these parameters.
Regarding the contribution of river sediments, the influence of hydraulic structures is limited because most of them were built before the period analyzed and the changes to the land cover affected a negligible portion of the basin. Furthermore, changes in precipitation have always been consistent with shifts in the shoreline in all periods analyzed. Therefore, in this case study, there seems to be a direct correlation between precipitation and changes in the coastline. This is consistent with the hydrological characteristics of most of the Calabrian rivers and, in general, of many Mediterranean rivers in which the WSE plays an important role.
The methodology for analyzing shoreline changes, as described in this article, is particularly useful to better understand the factors that most influence the coastal sediment balance and, therefore, is applicable to many contexts that are similar to the mouth of the Sant'Agata river.