First evaluation of the damage related to alluvial events in torrential catchments of Campania ( southern Italy ) , based on a historical database

This study presents an historical database of alluvial events in torrential catchments of Campania region, southern Italy. Detailed scrutiny and critical analysis of the existing literature, and of the data inventory available, allowed us to build 10 a robust database consisting of about 500 events. Being this study the first step of a longer project, aimed at eventually reaching an hazard analysis, information about time and site of occurrence are known for all the events. The outlet zone of torrential catchments (represented mainly by alluvial fans or fan deltas) are highly urbanized in Campania region, thus collecting information about past events could provide useful information on future events, in terms of damage, and of spatial and temporal occurrence as well. In section 1 we introduce the issue of alluvial events in Italy. Existing database and 15 published studies on hydrogeological events, in particular regarding Campania region, are presented in section 2, where we also discuss the importance of using the historical sources, and their limits and drawbacks. The geological and geomorphological settings of Campania region are introduced in section 3. Then, in section 4, we present our database by illustrating its general structure and the methodology used in collecting information. Statistical and data analysis carried out on the collected data are presented in section 5. Aimed at performing a complete hazard analysis, analysis on rainfall data 20 and the application of numerical models on alluvial events will be the future steps.


Introduction
Alluvial events in Italy are a frequent natural hazard, due to the peculiar orographic and climatic setting of many sectors of the country.The young orogen (the Apenninic Chain) in the peninsular area is characterized by high relief.In such a setting, torrential stream catchments are one the most widespread landscape unit, characterized by peculiar flood scenarios.During an alluvial event, a torrential stream catchment may be interested by different types of processes, including both fluvial processes (water flood, "debris floods/hyperconcentrated flow" and "debris flow", sensu Costa 1988) and gravity processes (i.e., landslides).As regards the fluvial processes, variations in the flow typology are mainly caused by the concentration of the sediments entrained in the flow.The sediments captured by the stream and transported downvalley generally come from the bed and banks of the torrent, and they may eventually contribute to build alluvial fans or fan deltas, depending on the geographical setting of the catchment (inland or coastal).Although each process has unique diagnostic effects and products, in nature there exists a continuum of flow conditions and sediment concentrations.Identification of the specific hydrological and geomorphic process is thus a central problem for the correct hazard recognition, since each process has different associated hazard characteristics (O'Brien and Julien 1985;Costa 1988;Fema 2000;Jakob 2005).Another important aspect Nat.Hazards Earth Syst.Sci. Discuss., doi:10.5194/nhess-2015-355, 2016 Manuscript under review for journal Nat.Hazards Earth Syst.Sci.Published: 19 January 2016 c Author(s) 2016.CC-BY 3.0 License. is that in torrential stream catchment a substantial difference of magnitude exists between the low water periods, including a seasonal lack of surface water in ephemeral streams, and flood events capable of heavily rearranging the river-bed topography.The low frequencies of important flow rates might cause the reduction of hazard perception and leads towards an incorrect land use and an unsafe management of the catchment.The collection of historical information of past events must be the starting point to deal with inhabitant security, land planning, and watershed management (D'Agostino, 2013).
As the first step in the process of hazard evaluation, we compiled a database with the following main aims: i) to identify over the whole region the most susceptible areas to alluvial events; ii) to discriminate, in the available literature and historical data, alluvial events in torrential stream catchment from flood on large river systems and from gravity processes, such as rapid earth or debris flow (Del Prete et al., 1998;Crosta and Negro, 2003;Guadagno et al., 2003Guadagno et al., , 2005;;Revellino et al., 2004;Zanchetta et al., 2004b;Di Crescenzo and Santo, 2005;Cascini et al., 2008b).

Database on hydrogeological events (floods and landslides)
In any hazard and risk analysis the first step typically consists in collecting information about past events, starting from the concept that past events can provide information on future events, in terms of both spatial and temporal occurrence.Taking into account the loss of memory, it is important to reach the better knowledge possible about where hazardous events occurred in the past, what intensity they had, and what was the frequency of occurrence in the recent history.
Different types of information source can be considered to build a database on floods and debris flows.An high percentage of data and information typically is derived from collection and critical analysis of historical documents.In addition, this part of the work allows to get some hints about past events for which no evidence have been left in the field, or to confirm oral, often not exhaustive, documentary sources.
To provide some examples, the CNR IRPI Institute in Turin built a database on landslides, debris flows and stream floods in Northern Italy, collecting hundreds of thousands of records contained in published and unpublished documents and historical reports on natural damaging events over the last 500 years (Tropeano and Turconi, 2004).Further, this work highlighted the importance of using historical documents in the evaluation of natural hazards, as also stated by other authors in different parts of Italy (Calcaterra and Parise, 2001;Gringeri Pantano et al., 2002;Calcaterra et al., 2003).Other studies were focused on the triggering events, by providing a chronological description of 2.256 climatically triggered events in Switzerland, occurred between the years 563 and 1988 (Röthlisberger, 1991).
The AVI project (an Italian acronym for Aree Vulnerate in Italia, Areas Affected by Landslides or Floods in Italy) was commissioned by the Minister of Civil Protection to the National Group for Prevention of Hydrogeologic Hazards of the Italian National Research Council (CNR) with the aim to compile an inventory of information on areas historically affected by landslides and floods in Italy (Guzzetti et al., 1994).It is the most complete archive of landslides and floods produced in Italy, and is continuously updated.
In Italy, again, Brunetti et al. (2015), within the framework of a project with the National Department of Civil Protection, analyzed rainfall events that have resulted in shallow failures and debris flows to define national and regional thresholds for the possible occurrence of rainfall-induced shallow landslides and debris flows in the country.
As concerns Campania, several studies have been conducted in order to collect historical information about hydrogeological events.In the majority of the cases these databases refer to landslides or floods, or put together different types of events.In this framework a regional-scale database on alluvial events in torrential catchment does not exist.
The area of Sarno, Quindici and Bracigliano, hit by the catastrophic event on May 5, 1998, was particularly object of detailed studies: Mazzarella and Diodato (2002), on the basis of old documents, collected alluvial events that since 1794 affected the town of Sarno, also including in their work information about rainfall by means of an index of strength (Mazzarella and Diodato, 2002).On the other side of the same mountain, where the village of Quindici is located, Calcaterra et al. (2003) analysed past slope instability by means of historical and geological data to assess the landslide hazard.A careful scrutiny of the historical information was made by the authors, who provided a critical evaluation of the examined sources in order to rank their reliability.Calcaterra et al. (2003) highlighted the importance of combined historical-geological investigations, since only the cross-analysis between historical research and geological, geomorphologic and engineeringgeological approaches may help to get a good knowledge on both the spatial and temporal distribution of the events.
Other authors have performed sub-regional or regional studies in Campania on landslides and floods (Migale and Milone 1998;Esposito et al., 2003, Di Crescenzo andSanto, 2005).As regard the alluvial events in the torrential catchments of the region, Alessio et al. (2013) built a database on floods occurred in the Somma-Vesuvius area based on historical (19-20th century) and geomorphological data: they listed 87 flood events, and retrieved spatial and temporal information, hourly and daily rainfall recorded and numbers of victims.In addition, they also calculated geomorphological and planimetric parameters of the catchments affected by the flood events.
Further works dealt with the observed or documented damage (Porfido et al., 2013), or with morphological and morphometric parameters aimed at a susceptibility analysis (Santangelo et al., 2012).A geomorphological and morphometric analysis of 102 basin/fan systems, located along the border slopes of the carbonate massifs of Southern Apennines, was carried out by Scorpio (2011).In this study, in order to analyze the susceptibility distribution to alluvial fan flooding in the region, morphometric features of the basin/fan system were used to classify the fans in terms of transport process (debris flow or water flood).
Eventually, many papers are addressed to specific study areas, such as the Teglia catchment (Santo et al., 2015), the Sala Consilina area in the Maddalena Mountain Ridge (Santangelo et al., 2011), and Santa Maria a Vico and Arienzo in the Caserta Mountains (Di Crescenzo et al., 2013).After the November 10, 2009, event in the northern sector of the island of Ischia, Santo et al. (2012) performed a flood hazard, through collection of historical information about landslides and floods (Del Prete and Mele, 2006), combined with geomorphological and rainfall analysis.Esposito et al. (2011) studied published and unpublished historical sources along the coastal areas of Campania, with greater detail in the Amalfi Coast.In the latter paper the authors also presented a level of quality of the available sources, similarly to what done by Calcaterra and Parise (2001).

Use of historical documents: strengths and weaknesses
Archival and historical data are typically descriptive, written by non-technicians, and rarely they do contain quantitative information on significant elements such as type of process, volume of sediments, runout distance or water height (Stoffel et al., 2013).Nevertheless, they represent an invaluable source of information.
Availability of historical documents, and the possibility to examine them, is a consequence of the history of the country, and of its socio-economic conditions as well.As stated by D'Agostino (2013), a debris flow during wars, famines, or in time of plagues probably does not represent a social priority; thus, the historical analyses could be affected by biases originating from different sensibility to similar events during different historical ages.Obviously, countries of ancient civilization offer a longer period of records, and greater possibilities to find interesting information.At the same way the urbanization history may influence the historical analysis, as a recent urbanized area may records only recent events because there were not witnesses in the past.The quality of historical information is also related to the involved area and the severity of the event.If a heavy event hits only a catchment or municipality, it is likely to have precise temporal and spatial information.On the other hand.if the event affected a larger area, the information will likely be more generic.Frequently, errors in transcription can result in attributing wrong time of occurrence to an event.It is therefore fundamental to compare all the available information sources.As for location, it may occur that the source reports the locality name of the property being hit, or the name of a street that does not exist anymore, which makes difficult (and sometimes impossible) to exactly locate the site.
As concerns damage to people, in most cases the reported information are very generic, such as "some victims", or in other cases the source provides the total number of victims per event but not for the individual municipalities.The precise estimate of the number of victims is thus not easy.In some cases divergences exist in the numbers of casualties reported by different sources for the same event.The differences can be related to several reasons, including the fact that the exact number is typically available only at the end of the search and rescue operations, from a few days to weeks after the event.During this period, newspapers and even official reports may provide different and changing data (Salvati et al., 2010).
Typically, information sources mainly document the most remarkable events for number of deaths and damages caused.In the last decades, however, with the intensifications of land use, information about smaller events has become available (Stoffel et al., 2013), and might be relevant and useful in order to develop a better hazard assessment.

Study area
The study area is the Campania region (Southern Italy) extending from the Tyrrhenian Sea to the Southern Apennine Chain for about 13500 kmq.The orographic setting is characterized by the presence of a central mountain ridge made up mainly of Mesozoic carbonates, elongated for more than 200 km in a NW-SE direction, with maximum peaks reaching 2000 meters a.s.l.On the western side the chain is bounded by a deep (up to 2 km) coastal graben originated by Plio-Quaternary extensional tectonics, which were filled by marine/transitional sedimentary successions, and are now occupied by large and flat coastal plains (Ascione et al., 2008).During the late Quaternary a strong volcanic activity was registered in the Campania plain coastal graben with the growth of the Somma-Vesuvius and the Phlegraean Fields volcanoes (Romano et al., 1994).The landscape of the western portion of Campania is thus characterized by a wide flat area with isolated volcanic reliefs and islands.On the eastern side of the region the carbonate ridges pass to hilly landscapes of lower elevation widespread geomorphologic unit along the flanks of the main carbonate ridges, and the slopes of the volcanoes as well, where they have also the higher longitudinal gradients.In the remnant hilly part of the region stream catchments have lower mean longitudinal profiles, and wide alluvial plains linked to perennial river systems prevail.A simplified geological scheme of Campania region is shown in Figure 2.
The general climate is humid temperate with mean annual precipitation ranging from 1000 to 2000 mm.In short, the main situations responsible for abundant rains over the region are generally north-westerly or westerly winds bringing eastwardmoving cyclonic depressions.Due to the rugged topography of the region, heavy convective precipitation often results in flash floods, with concomitant widespread floods and landslides.

Sources used in data collection
The  Guzzetti et al., 1994).All these data were located as precisely as possible, and included in the web gis database.
The second step involved the critical analysis of each input data aiming at discriminating between alluvial events and landslides, as in many reports these different types of processes are put together under the general classification of "hydrogeological events".In particular, we took into account the following event typologies: 1. Flood events in alluvial plain; 2. Flood events in torrential stream catchment; 3. Landslides; 4. Mixed and/or doubt cases.
All the events that may be easily discriminated based on general description and geographic location were inserted in the database.For the mixed and doubt cases a second analysis was carried out, in order to classify them in one of the four classes, or to definitely exclude them from the database.
Once all the events were grouped in different typologies, we moved to exclude from our analysis the classes 1 and 3, focusing our attention on the class 2, which represented the 56% of the collected data (Fig. 3), and included a list of more than 500 events of alluvial events occurred in torrential stream catchments.

Catchment typologies
The torrential stream catchments of Campania most affected by alluvial events in the last decades are mainly located in the carbonate and volcanic settings (Palma et al., 2009;Santangelo et al., 2012;Santo et al., 2012Santo et al., , 2015;;Alessio et al., 2013).As regards their geographic location they may have both an inland outlet (generally represented by a fan or by a well defined foothill area; Santangelo et al., 2012), or a coastal outlet, generally represented by a fan delta (Esposito et al., 2011;Santo et al., 2012).They also show similar morphometric conditions (Santangelo, 2012;Alessio et al., 2013) which can be summed up as follows: -Limited catchment area (from few km 2 to 10 km 2 ); -High relief energy (from hundreds of meters up to 1000 m); -High slopes gradient (generally greater than 35°); -High mean gradient of feeder channel (greater than 15°); -Low concentration time (from 30 minutes to some hours).
As the sediments captured by the stream and transported down-valley generally come from the bed and banks of the torrent, we tried to discriminate among different catchment typologies, basing on the following parameters: -Bedrock of the catchment (carbonate/volcanic); -Presence/absence of detrital cover and, in case of presence, its origin (weathered bedrock, soil, pyroclastic cover, etc.); -Type of outlet zone (alluvial fan or fan delta in coastal area).
Thus, the collected events were eventually grouped, based upon their occurrence in the following five classes: 1. Carbonate catchment with pyroclastic cover; 2. Carbonate catchment without pyroclastic cover; 3. Carbonate catchment with pyroclastic cover and outlet to the sea; 4. Volcanic catchment; 5. Volcanic catchment with outlet to the sea.

Structure of the database, and collected information
The database contains about 500 events and is being continually updated.Each event is identified by an ID, and by the geographical coordinates (events are located by using Google Earth©).The catchment type for each event, according to the five classes defined, is also indicated in the database.Information about time of occurrence is mandatory to include an event in the catalogue.To evaluate the temporal degree of accuracy, five classes were introduced.Undoubtedly, going back to the past the accuracy decreases to the lowest level (annual accuracy).The highest value of temporal accuracy, on the other hand, is assigned when the availability of information on the time of occurrence is complete (hour, day, month and year of occurrence).As concerns the trigger, the rain gauge closest to the site of the event, along with the rainfall data, is also included for the events for which daily or hourly rainfall data were accessible.Particular attention has been given in the catalogue to damage to people and infrastructures.The documented events caused heavy damages to the society, primarily involving buildings, infrastructures and lifelines (roads, pipelines, etc.).Damage to the population includes number of victims, injured, homeless and missing people.In several cases the information source does not allow to quantify damage to people, since the reported information are typically generic, such as "some victims" or "several victims".Thus, aimed at not loosing any useful data, information about damage are reported also as text, as from the original sources.Further, additional useful information can be included in the field "notes".
The database is synthetically shown in Table 1, which contains the main information for each event, as date and site of occurrence, damage and victims.

Data analysis
The database on alluvial events in torrential catchments of Campania contains about 500 events at the time we write (September 2015).The oldest event occurred in 1540, and affected the town of Amalfi.Temporal distribution of the events on the territory (Fig. 4) reveals that most of the municipalities have been affected more than once during the time period covered by the data included in the database (from 1540 to nowadays).Despite there are several catchments with similar geological and geomorphological characteristics, many villages did not record any event, which is probably related to lack of inhabited areas.More than 60% of the events occurred during the last 50 years, but this outcome is likely related to the higher availability of information sources, to the numerous scientific studies carried out in the last decades (mostly as an effect of the Sarno-Quindici catastrophic event in 1998), and to the growing attention toward geological hazards from the society.Five grades of temporal accuracy have been defined in order to classify the different level of knowledge of time of occurrence of the events.Events for which only the year of occurrence is known have a low accuracy, while a high accuracy is assigned when hour, day, month and year of occurrence are known.The histogram in Fig. 5 indicates that most of the collected events have a middle-to-high accuracy, meaning that day, month and year of occurrence are known.The accuracy degree decreases for the oldest events, as well as also for the smaller events.
The recurrence of alluvial events in the same area was also investigated.As shown in Figure 6, among the 86 municipalities that have been damaged by torrential flooding, 16 recorded more than 10 events.At a greater detail 9 out of 16 were located in the outlet zone of coastal carbonate catchments (Sorrento peninsula -Lattari Mts.ridge), whilst the others are in the piedmont areas of inland carbonate massifs (Picentini and Matese Mts.) or in the volcanic area of Somma-Vesuvius.In all these cases the recurrence of the events is very low, ranging from 39 to 3 years, with a mean value of 15(Fig.7).If only the most damaging events are taken into account the recurrence time increases to 50.
As regards the type of involved catchments, Figure 8 shows the events distribution on the territory, classified according to the aforementioned five classes.Most of the events took place in carbonate catchments with pyroclastic cover, both with and without an outlet to the sea (Figs. 8 and 9a).The widespread presence of carbonate catchments, and the high urbanization as well, affected the information source: as above mentioned, higher urbanization means typically greater availability of data, due to a higher attention toward occurrence of natural hazards.Distinguishing between different catchment types is important in order to discern the different bed and banks materials available, that the surge could entrap and transport downvalley.In a carbonate catchment the surge could take gravel without matrix or with low matrix (Fig. 10a), whilst in carbonate catchments with pyroclastic cover medium and coarse gravels with an high percentage of matrix are available (Fig. 10b).In the volcanic catchments, on the other hand, due to greater erodibility of the material, it is possible to find mainly silts and clays on bed and banks (Fig. 10c).
The monthly distribution of the events is quite variable in the different types of catchments.Figure 9b shows that all the events reach the peak in October, and that Autumn is the season with the highest frequency.After the dry period, heavy rainfall can generate sudden and high discharge, and the water runoff can carry downstream sediments accumulated as a result of the erosion processes.
On the basis of the collected data, in the carbonate catchments with pyroclastic cover the lower number of events is registered in April, or generally during the spring, if there is an outlet to the sea.In the carbonate catchments without cover during the spring season (March-May) there are no events, and the lower number of alluvial events is in January.As regards the volcanic catchments, February and March are the months with the smallest number of events.If the catchment has an outlet to the sea during the spring (March-May) no event has been recorded, and November-December are the months with the lower numbers of events.
As regards damage to people, about 18% of the events caused at least one victim.Figure 11 depicts their distribution in the region.The most dangerous events interested the province of Salerno, affecting the carbonate catchments with pyroclastic cover, both with and without an outlet to the sea.This means that carbonate catchments with pyroclastic cover (class 1) are the most hazardous.All the events with more than 100 victims took place in October; for the most damaging events the total amount of deaths was caused by landslides, debris flows and floods in floodplains.As a consequence, it is not easy to evaluate the victims caused only by alluvial events in torrential catchments.In many situations it was difficult to assign the precise number of casualties, since the information appear to be generic, under forms as "some victims", or "few victims".
Further, the attribution of the number of casualties to each municipality was very difficult, since typically the source provides the total number of victims per event, not distinguishing for the different villages.This was the case, for instance, for the events in 1581 (700-1000 victims), in 1899 (approximately 100 victims), in 1924 (approximately 100 victims), and in 1910 (over 200 victims) that have been the most harmful recorded in the history.Furthermore, as mentioned before, differences in the data reported by different sources for the same event had to be registered.
Taking into account the total amount of collected events, the database contains few cases with injured, homeless and missing people.As regards homeless, for instance, the number is too small when compared with the numbers of events that caused complete disruption of the buildings.This is probably an hint for evaluating that the documented data actually in some ways underrate the reality.
Alluvial events in torrential catchments caused severe damages to the society, primarily involving building, lifelines and infrastructures.In particular, roads and private buildings are the most affected categories.The most dangerous catchments types are carbonate catchments with and without pyroclastic cover.Table 2 reports the category damaged by each event included in the database.In Figure 12 some examples of alluvial events in different torrential catchment types are reported.
Frequently they cause severe economic losses to society.

Final remarks
By analysing the existing literature, we selected alluvial events occurred in torrential catchments of Campania, a land that has repeatedly been affected by severe flooding which caused serious damage and fatalities.
We collected temporal and spatial information on about 500 alluvial events, thus building the first specific database concerning this typology.To this aim, a critical scrutiny of the existing literature was performed, to provide a degree of reliability to the collected information.We also defined the accuracy related to the temporal information available, and for most of the collected events the accuracy resulted to be middle-high, meaning that day, month and year of occurrence are known.
In order to reconstruct flooding scenarios we defined different catchment types on the base of the main geological (bedrock lithology and presence/absence of detrital cover) and geomorphological parameters (type of outlet zone).This differentiation may be useful to understand the type of transported bed load (coarse vs fine grained) and to characterize the deposition area.
We collected also information about damage to people and society.Most of the events took place in carbonate catchments with pyroclastic cover, both with and without an outlet to the sea.Among the 86 municipalities that were damaged by torrential flooding, 16 recorded more than 10 events.In Campania the recurrence time of alluvial events is very low, ranging from few years for damages to buildings and infrastructures up to some decades in the case of events with victims.The widespread presence of carbonate catchments in the region and the high urbanization of the area affected the information The study has also shown that a significant number of catchments were interested by floods with high recurrence time The loss of historical memories of these events is for sure at the origin of an increase in the risk conditions.
The collection of this database represents the first step toward a full hazard analysis; furthermore, by analyzing triggering rainfall events and concentration time, it could also contribute to the development and implementation of specific early 310 warning systems.Eventually, the application of numerical models on torrential floods could be useful to identify the flooded areas depending on the different peak discharge and the different bed material that can be entrained in the flow.

Figure 1 (
modified after D'Agostino, 2013) illustrates the links between the time scale of the event (single, multiple years period, all documentable time) and the scale area involved (single catchment, part of a mountain valley, a region).Depending on both the factors, different levels of historical research have to be developed: specific, medium scale and extended historical research.
Nat. Hazards Earth Syst.Sci.Discuss., doi:10.5194/nhess-2015-355,2016   Manuscript under review for journal Nat.Hazards Earth Syst.Sci.Published: 19 January 2016 c Author(s) 2016.CC-BY 3.0 License.sources, and these represent the main weaknesses of the historical documents.Notwithstanding volcanic areas are densely populated, alluvial events in carbonate catchments with pyroclastic cover caused more damage to people and society than in 305 volcanic catchments.

Figure 1 :
Figure 1: Relation between area and time scale for the research accuracy, modified after D'Agostino (2013).

Figure 2 .
Figure 2. Location and geological setting of the study areas.Key: 1) Mesozoic carbonate massifs; 2) Cenozoic hilly terrigenous areas; 3) Quaternary volcanic areas; 4) Quaternary intermountain catchments and coastal plains.The broken line indicates the boundaries of Campania region.

Figure 3 :
Figure 3: Event typologies collected during the database building.

Figure 4 :
Figure 4: Temporal distribution of alluvial events in Campania.

Figure 5 :
Figure 5: Temporal accuracy distribution of the collected events in Campania.The histogram shows the distribution in five classes of temporal accuracy: L low, ML Middle-Low, M Middle, MH Middle-High, H High.

Figure 6 :
Figure 6: Recurrence of alluvial events in the municipalities.

Figure 7 :
Figure 7: Recurrence time of the events in the municipalities that have recorded more than 10events.

Figure 8 :
Figure 8: Representation of the events on the territory, according to the catchment class.

Figure 9 :
Figure 9: a) percentage of events in each catchment class; b).monthly distributions of the events in the different catchment classes.Catchment classes: 1) Carbonate catchment with pyroclastic cover, 2) Carbonate catchment without pyroclastic cover, 3) Carbonate catchment with pyroclastic cover and outlet to the sea, 4) Volcanic catchment, 5) Volcanic catchment with outlet to the sea.

Figure 11 :
Figure 11: Distribution of victims recorded in the territory.