**1. Introduction**

Landslides are considered, worldwide and in Italy, as one of the most important and frequent natural hazards [1–5] as their occurrence can directly impact humans, infrastructures, economic activities, and the social and environmental systems [6–8]. Landslides are a landscape modelling process inducing geomorphological changes on slopes in mountainous, hilly, and coastal areas. Their occurrence is generally controlled by predisposing factors (i.e., morphology, lithological and structural setting, vegetation cover, land use, climate, etc.) and triggering ones (e.g., heavy rainfall and snowfall events, snow melting, earthquakes, wildfires, human activity, etc.) [9–13]. Many of the triggering factors are only sufficient conditions for the occurrence of landslides, which are occasional and spasmodic. Therefore, it is essential to pay attention to predisposing factors in landslide analyses to set an organic correlation between climate regime, morphostructural/geological framework, and slope instability phenomena [14,15].

Many theories and methods have been proposed about the spatial relationship between landslides and causative factors [16–22] to perform landslide hazard assessment studies [23]. However, the type, extent, magnitude, and direction of the geomorphological processes and the location, abundance, activity, and frequency of landslides in a changing

**Citation:** Esposito, G.; Carabella, C.; Paglia, G.; Miccadei, E. Relationships between Morphostructural/ Geological Framework and Landslide Types: Historical Landslides in the Hilly Piedmont Area of Abruzzo Region (Central Italy). *Land* **2021**, *10*, 287. https://doi.org/10.3390/ land10030287

Academic Editor: Wojciech Zgłobicki

Received: 16 February 2021 Accepted: 8 March 2021 Published: 11 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

environment are still under debate. Establishing a relationship between climate change and its potential effects on the occurrence of landslides remains an open issue [24]. The role played by projected climate changes in modifying the response of single slopes or entire catchments, the frequency and extent of landslides, and the related variations in landslide hazard, remain to be discussed and understood [25–27]. Most of the current landslides in the Central Apennines are the reactivation by pre-existing ones, which have occurred in periods of climatic and geomorphological conditions different from those of the present. Most dormant slides and/or paleolandslides, in which the strength parameters are reduced to values close to the residual ones, can be reactivated and/or modified by natural causes, such as rainfall or snowmelt, as well as man-made disturbance [28,29].

Geomorphological mapping is a common and fundamental tool for the representation and the comprehension of the spatial and temporal development of landslides. Recent and new methods developed in the last decades have improved landslide analysis with multidisciplinary approaches including (i) morphometric analysis using very-high-resolution Digital Elevation Models (DEMs), (ii) interpretation and analysis of satellite images, including Synthetic Aperture Radar (SAR) images, and (iii) the use of new tools to facilitate field mapping [30–35]. Moreover, the investigation of geomorphological processes and dynamics, in different and complex morphostructural domains, became necessary for the assessment of the areas prone to landslides with reference to the predisposing and/or triggering factors.

According to national and regional inventories [36,37], the Abruzzo Region (Central Italy) is acknowledged as an area highly exposed to landslide hazards and risks. It is located in the central–eastern part of the Italian peninsula, and it is characterized by a landscape that is the result of a complex cyclic evolution that occurred in succeeding stages with the dominance either of morphostructural factors, linked to the conflicting tectonic activity (compressive, strike–slip, and extensional tectonics) and regional uplift, or morphosculptural factors, linked to drainage network linear down-cutting and slope gravity processes [14,38,39].

For developing the present study, the analysis of historical landslides was carried out following an integrated approach that incorporates literature and landslide inventory analysis, relationships between landslide types and lithological units, detailed photogeological analysis, and geomorphological field mapping. The paper focuses on selected slope instabilities to highlight the multitemporal geomorphological evolution and the interplay between morphostructural/geological framework and landslide dynamics in the hilly piedmont area of Abruzzo Region. The work shows an effective integrated approach in geomorphological studies for landslide hazard modelling at different spatial scales, readily available to interested stakeholders. Furthermore, it could provide a scientific basis for the implementation of sustainable territorial planning and loss-reduction measures in a changing environment.

## **2. Study Area**

The study area is located in the central–eastern part of the Italian peninsula along the hilly piedmont area of Abruzzo Region, between the Apennine chain and the coastal area (Figure 1a). It includes the lower part of the main SW–NE to W–E fluvial valleys (i.e., Vomano, Pescara, and Sangro rivers), and the small tributary catchments of the main rivers and those incising the coastal slopes.

The Apennine chain area is characterized by a mountainous landscape (with reliefs up to 2900 m.a.s.l. high) interrupted by longitudinal and transversal valleys and wide intermontane basins (i.e., Fucino Plain, Sulmona Basin). It is made up of carbonate lithological sequences pertaining to different Meso-Cenozoic palaeogeographical domains. Carbonate shelf limestones, slope limestones, basin limestone, and marls represent the carbonate backbone of the main ridges of the Abruzzo Apennines, and allochthonous pelagic deposits are widespread in the southern sectors featuring a chaotic assemblage on clayey–marly–limestone units. The main tectonic features are represented by NW–SE to

N–S-oriented thrusts, which affected the chain from the Late Miocene to the Early Pliocene. Compressional tectonics was followed by strike–slip tectonics along mostly NW–SE to NNW–SSE-oriented faults that were poorly constrained in age and largely masked by later extensional tectonic events since the Early Pleistocene [40,41].

The hilly piedmont area is a low relief area (heights ranging from ~100 to 800 m.a.s.l.) characterized by a cuesta, mesa, and plateau landscape and a gently NE-dipping homocline, locally cut by fault systems (NW–SE, SW–NE) with low displacement [42–44]. Bedrock lithologies pertain to Neogene sandy-pelitic turbidites and Plio-Pleistocene marine clayey– sandy and conglomeratic deposits. The geological and structural setting is related to the Pliocene–Quaternary evolution of the Adriatic foredeep and the related regional uplifting processes. Since the Middle Pleistocene, the geomorphological evolution has primarily comprised the incision of major dip river valleys (WSW–ENE-oriented), characterized by fluvial deposits arranged in flights of at least four orders of terraces (Middle Pleistocene– Holocene) [44,45]. Quaternary continental deposits are widely present in the alluvial valleys, alluvial plains, and coastal slopes. They can be referred to fluvio-lacustrine, travertine, sandy shore, and eluvial–colluvial deposits (Figure 1b).

The geomorphological framework is mainly related to fluvial and slope processes. Fluvial processes affect the main rivers, alternating between channel incisions and flooding. The slope processes due to running water mostly affect the clayey and arenaceous-pelitic areas of piedmont and coastal sectors, generating minor landforms such as rills, gullies, and mudflows [46,47]. The area is extensively affected by different types of landslides (e.g., mostly rotational–translational slides, earth flows, rockfalls, complex slides), mostly characterizing the hilly piedmont and the chain area and, locally, the coastal area [3,48].

The present-day regional tectonic setting is dominated by extensional tectonics still active in the axial part of the chain, which is characterized by intense seismicity and strong historical earthquakes (up to M 7.0; [49]). The piedmont area is characterized by moderate uplifting and moderate seismicity, while the Adriatic Sea is affected by subsidence and by moderate compression and strike–slip related seismicity, as also documented by the recent seismicity [50] (Figure 1b).

Climatically, the study area belongs to temperate sub-littoral regime with scarce annual rainfall, mainly autumnal, and medium temperatures [51]. It is largely affected by the orographic setting, changing from a Mediterranean type with maritime influence along the coasts and the piedmont area to more continental-like in the inner sectors [52]. The hilly piedmont area is characterized by a maritime Mediterranean climate [53]. The average annual precipitation is 600–800 mm/year, with occasional heavy rainfall events (>100 mm/d and 30–40 mm/h). The mean annual temperature ranges between 12 and 16 ◦C in the coastal part of the region, with mild winters and hot summers, and from 8 to 12 ◦C in mountain areas, with more severe (low) temperatures, especially in the winter season [54,55].

**Figure 1.** (**a**) Location map of the Abruzzo Region in Central Italy; (**b**) geolithological map of the Abruzzo Region (modified from [56,57]). Legend: (1) eluvial–colluvial deposits; (2) sandy shore deposits; (3) recent fluvio-lacustrine deposits; (4) travertine deposits; (5) morainic deposits; (6) old fluvio-lacustrine deposits; (7) conglomeratic deposits; (8) clayey–sandy deposits; (9) sandy turbidites; (10) pelitic turbidites; (11) carbonate deposits in conglomeratic and calcarenitic facies; (12) allochthonous pelagic deposits; (13) carbonate ramp limestones; (14) basin limestones and marls; (15) slope limestone; (16) open carbonate shelf-edge limestones; (17) carbonate shelf limestones and dolomites. Seismicity derived from [49].

#### **3. Materials and Methods**

Landslide analysis was achieved through an integrated approach based on the combination of literature data, landslide inventory analysis, statistical analysis of the relationships between landslide types and lithological units, detailed photogeological analysis, and geomorphological field mapping, supported by multidisciplinary analysis and GIS– based techniques.

#### *3.1. Landslide Inventory Maps and Database Analysis*

Landslide inventories and databases represent an important tool to document the extent of landslide phenomena in a region, to investigate the distribution, types, pattern, recurrence, and statistics of slope failures, to determine landslide susceptibility, hazard, and risk, and to study the evolution of landscapes dominated by mass-wasting processes [58].

A preliminary GIS-based analysis was performed to store, organize, and manage available data recorded in four different databases and catalogues, briefly described as follows. The IFFI database (Italian Landslide Inventory—[59,60]) supplies a detailed picture of the distribution of landslide phenomena within Italy. As of today, the IFFI database holds 620,793 landslide phenomena, covering an area of approximately 23,000 km2, which is equivalent to 7.9% of the Italian territory [37]; for the Abruzzo Region, the database is updated to 2007. The compilation of the catalogue was structured in several phases: (i) collection of bibliographic cartographic data useful to identify areas subject to landslides; (ii) verification by aerial photo interpretation and cartographic transposition; (iii) verification through field-based analysis; (iv) digitization. A total of 6557 events (categorized as rockfalls, lateral spreading, complex landslides, translational and rotational slides, debris flows, earth flows, DSGSDs, and soil creep areas) were included in the inventory used in this study. The CEDIT catalogue (Italian catalogue of earthquake-induced ground failures—[61]) includes more than 150 earthquakes and almost 2000 earthquake-induced effects, which involved almost 1100 localities; the catalogue is updated to the 2016–2017 Central Italy seismic sequence [62,63]. The catalogue implies detailed research of historical documents and reports as well as of already published scientific papers. The analysis of reported seismically induced effects infers that most of them are landslides, which account, alone, for about half of the total (44%). Among all these earthquake-induced landslides, only seven events are located in the hilly piedmont area, and they were selected, recognized, and integrated into the analysis in terms of georeferenced location and detailed information. The EEE catalogue (Earthquake Environmental Effects catalogue—[64]) is aimed to collect in a standard format the wealth of information of environmental/geological effects induced by a seismic event; the catalogue contains tables that include information at site of each EEE, including detailed characteristics on the type of earthquake. The database is updated to the 2016–2017 Central Italy seismic sequence. Among all the documented seismic-induced effects, only landslides (six events falling within the study area) were selected and included in the analysis. The FraneItalia catalogue [65] contains information retrieved from online news sources (especially Google Alerts and Italian Civil Protection press reviews) on landslides that occurred in Italy. It contains all the landslide events reported since 2010 (January 2010–December 2017), not only the ones that caused direct consequences to people or major damage; it is structured as a geo-referenced open-access database containing information on a variety of landslide features and consequences. For this study, all the landslides (162 events falling in the study area) for which it was possible to univocally define the location and the type of movement were selected and included in the inventory.

Available data (i.e., georeferenced location and detailed information) from the abovementioned catalogues were merged to completely define the landslides' spatial distribution over the Abruzzo Region (Figure 2).

**Figure 2.** Landslide spatial distribution over the Abruzzo Region. This graphical representation includes the georeferenced location of rockfalls, landslides (lateral spreading, complex landslides, translational and rotational slides), debris flows, earth flows, DSGSDs, and soil creep areas. This general labelling derives from all historical documents, technical reports, and detailed information included in available inventories and databases, such as the Italian Landslide Inventory (IFFI) catalogue [60]; the Italian earthquake-induced ground failures (CEDIT) catalogue [61]; the Earthquake Environmental Effects (EEE) catalogue [64]; the FraneItalia catalogue [65]. The black line represents the study area.

> Even if the landslide spatial distribution over the Abruzzo Region is related to rockfalls, landslides (lateral spreading, complex landslides, translational and rotational slides), debris flows, earth flows, DSGSDs, and soil creep area, landslides located in the study area and used for this analysis were categorized and selected according to the type of movement into four categories: rotational and translational slides, complex landslides, earth flows, and rockfalls. This specific labelling was followed to highlight the most characterizing and frequent mass movement types, according to geological–structural setting, location and

abundance of landslides. Then, the spatial distribution of each category was evaluated through the creation of density maps, generated using the QGIS (version 3.10, 2019, "A Coruña") HeatMaps (Kernel Density) tool, which calculates a magnitude-per-unit (1 km2) area from a point or polyline features using a kernel function to fit a smoothly tapered surface to each point. Landslide density maps generally show a synoptic view of landslide distribution for large regions or entire nations in order to portray the first-order overview of landslide abundance. Density is a clearly definable and easily comprehended quantitative measure of the spatial distribution of slope failures. These maps derive from the georeferenced location of each initiation point of landslides (defined as the center of the main headscarp) and assume that landslide density is continuous in space, which may not be the case everywhere.

#### *3.2. Statistical Analysis of the Relationships between Landslides and Lithological Units*

Lithology shows a grea<sup>t</sup> influence on landslide development since different lithological units may be affected by different landslide types. Moreover, soil cover deposits, mostly exposed to weathering, may influence land permeability and the landslide type, as known from thematic literature [66,67].

In order to stress the role played by lithological units on the development of landslides and build up a statistical relationship with the spatial distribution of landslide type, a vector lithological map (previously categorized into 17 lithological units according to the sedimentation environment and the lithological features of the outcrops) was spatially overlapped with the landslide distribution layer, derived from the selected inventories and databases.

A GIS-based overlay between the georeferenced location of the initiation points of landslides (defined as the center of the main headscarp) and lithological units was performed to understand the influence of lithologies on landslides. This correlation was carried out for different types of landslides (rotational and translational slides, complex landslides, earth flows, and rockfalls) recorded in the hilly piedmont area.

#### *3.3. Detailed Multitemporal and Multidisciplinary Analysis*

Multitemporal and multidisciplinary analyses were performed to outline the mass movement types and evolution mechanisms that characterize the different morphostructural domains of the study area. Selected case studies (one for landslide type; about rockfalls, according to a moderate to low spatial distribution, no landslide events have been identified as clearly representative of this mass movement type in the study area, so no case study was reported) have undergone several main movements from the 18th century onwards. These are intended to be representative of the most characterizing and frequent mass movement type, showing significant features useful for understanding the relationships between landslide types, lithologies, and morphostructural setting.

Multitemporal geomorphological analysis was based on detailed analysis of historical maps and literature data, stereoscopic air-photo interpretation, and field mapping. Airphoto interpretation was performed using 1:33,000, 1:20,000, 1:13,000, and 1:5000 scale stereoscopic air-photos (Flight GAI 1954, Flight CASMEZ 1974, Flight Abruzzo Region 1981–1987, and Flight Abruzzo Region 2018–2019), 1:5000-scale orthophoto color images (Flight Abruzzo Region 2010), and Google Earth imagery; this analysis was also supported using high-resolution Digital Elevation Models (DEMs). Field mapping was carried out at an appropriate scale (1:5000–1:10,000), according to international guidelines [68], Italian geomorphological guidelines [69] and the thematic literature concerning geomorphological mapping, fieldbased and numerical analysis [70–73]. It was focused on the definition of lithological and morphostructural features, superficial deposit cover, and the type and distribution of geomorphological landforms with reference to the main landslides affecting the study area.

Rainfall data analysis was carried out to outline the distribution of the climatic parameters and conditions in the hilly piedmont area. The analysis was based on a rainfall

dataset obtained from a network of 51 gauges (data provided by the Functional Center and Hydrographic Office of the Abruzzo Region, Pescara, Italy). Using the ArcGIS Kernel Interpolation function, the variation of the distribution of rainfall in the study area was derived for a 65-year time record (1950–2015).

To support the geomorphological dynamic of the area and improve the knowledge of spatial and temporal evolution of landslides, an interferometric analysis (InSAR) was implemented. The approach used is the so-called Persistent Scatterers Interferometry (PSInSAR), which is based on the information achieved by pixels of the SAR images characterized by high coherence over long time intervals [74]. Generally, constructed structures, such as buildings, bridges, dams, railways, pylons, or natural elements, such as outcropping rocks or homogeneous terrain areas, can represent good Persistent Scatterers (PSs). However, these techniques are also affected by some limitations. First, because only objects which are good "radar reflectors" can be analyzed, they cannot attain information over highly vegetated areas. This aspect is not secondary, as landslides often involve non-urban areas [75]. For the present study, we performed analyses of past displacements using data-stacks from the ESA archive ranging in the period 1992–2010. Specifically, Envisat data were selected from the 2003–2010 period, providing quantitative data (i.e., the detection of targets affected by displacements) about displacement information present in both the ascending and descending geometries.
