Abstract
From the geomorphological point of view, urbanization changes fluvial systems more drastically than any other single human activity. Its impact on rivers is of interest as it results in serious issues concerning geomorphic instability, ecosystem service reduction, and geo-hydrological risk increase. The present study provides a quantitative and qualitative assessment of landscape changes driven by urbanization, which occurred from the mid-19th century to the present day along the lower reaches of the Bisagno and Polcevera rivers and their tributaries in the Mediterranean coastal city of Genova (NW Italy). The bankfull channel morphological adjustments, the minor hydrographic network disappearance, the land use and land cover changes, and the current physical accessibility of river margins were investigated. This work is based on collection and review of historical data, photograph interpretation, GIS analysis, and field surveys. Urbanization essentially occurred between the mid-19th century and the late 1930s. The bankfull channel of the Bisagno and Polcevera rivers experienced large narrowing associated with widespread establishment of channelization structures. The minor hydrographic network essentially disappeared underground. After the 1930s, further minor anthropogenic interventions were implemented to accommodate urban expansion and consolidate channelization works; however, the overall geomorphological setting remained unchanged. The geomorphological evolution of the Bisagno and Polcevera rivers followed three different phases associated with the urban expansion process. Nowadays, the study area is completely urbanized, the fluvial stems flowing through the valley floor are totally disconnected from their adjacent areas, the main rivers show a relevant geomorphological stability induced by anthropogenic pressures, and the accessibility of river margins for people is rather limited. The outcomes from this study represent an essential knowledge base for technicians and policymakers to plan urban development and inform effective and sustainable management measures. Moreover, they provide insights into the medium- and short-term morphological evolution of urban rivers.
Keywords
Introduction
Many urban centres worldwide, especially in lowland areas, developed around rivers, as they provide essential resources for human life (Kummu et al., 2016; Wohl, 2004). River systems has always represented a storage of drinking water and irrigation water, a transportation and communication mean, a strategic defence from enemies, and a source of food, power, and materials (Haidvogl, 2018; Wang et al., 2022). In recent times, the less perceivable river-associated functions of biodiversity reserve and component of complex cycles, such as water, sediment, and nutrient cycles were described giving further relevance to fluvial systems and their ecosystem services (Fenoglio et al., 2010; Florsheim et al., 2008; Gruppuso et al., 2022; Kondolf, 1997; McCabe, 2011; Ward et al., 1999). In addition, rivers are an important driving force of culture and civilization (Wang and He, 2022). The first urban civilizations that developed in the Mediterranean area, in the Middle East, the Indian subcontinent and China were founded on rivers and their floods (Penny, 2010; Petts et al., 2002). Across Europe, the history of urban living dates back some 3000 years and is associated with the development of early towns that served as focal points for trade along inland waterways (Petts et al., 2002). Thereafter, cities have proliferated and have expanded with different spatial and temporal patterns across the world leading to significant changes in riverine landscapes (Gibling, 2018). Consequently, many of the world’s rivers are impacted by urbanization (Chin, 2006; Gregory, 2006; Gregory et al., 1992).
Urbanization affects a river in many ways (Grimm et al., 2008; Groffman et al., 2003; Walsh et al., 2005). It reduces water reserves because of abstraction and reduced infiltration (Flörke et al., 2018; Sun and Caldwell, 2015), causes water quality degradation due to urban run-off and wastewater drains (Burri et al., 2019; Strokal et al., 2021), reduces the ability to support wildlife due to the loss of riverine habitat complexity (Foley et al., 2005; Mcdonald et al., 2008), and originates hotspots for the introduction of non-native species (Gaertner et al., 2017; Santana Marques et al., 2020).
From the geomorphological point of view, urbanization changes fluvial systems more drastically than any other single human activity (Chin et al., 2022). It impacts on controlling factors of riverbed morphology causing geomorphic instability and consequent channel adjustments (Kondolf and Podolak, 2014; Wohl, 2004). Furthermore, it may result in relevant disturbances for riverine landscape up until the total disappearance of former landforms in river reaches that were embedded within urban areas (Cooper et al., 2018; Crutzen, 2002). Urbanization typically increases water discharge due to surface run-off from impervious areas, as well as frequency and magnitude of floods (Hollis, 1975; Mitsova, 2014; Ogden et al., 2011; Woltemade et al., 2020). The investigation of the different phases that characterize the urbanization process documented an increase in sediment supply during the build-out phase and a reduction in sediment supply after the completion of the development (Graf, 1975; Gupta, 1982; Nelson and Booth, 2002; Wolman, 1967; Yorke and Herb, 1978). In urban areas, rivers are generally channelized, diverted, and, in many cases, hidden underground (Gregory et al., 1992; Vietz et al., 2016b; Whitlow and Gregory, 1989). As a result, the connectivity between the riverbed and its adjacent floodplain is interrupted and the complexity of fluvial landforms and the associated habitats is lost.
The aforementioned urbanization-induced stresses make rivers less resilient to the effects of climate change and arose serious issues concerning biodiversity conservation, ecosystem service reduction, and geo-hydrological risk increase (Best and Darby, 2020).
In many cases, urbanization caused a progressive disconnection of the river from its surroundings not only from the physical point of view. Water courses hidden behind walls or underground often became open sewers and breeding ground for disease; in some cases they became inaccessible channels, and they were often forgotten (Kiechle, 2022; Petts et al., 2002; Rasool et al., 2022; Ridolfi et al., 2020; van Heel and van den Born, 2020). Thus, disconnection was also social and cultural, and led to the reduction in knowledge, awareness, and attachment of riverine communities with respect to their own river (Hillman et al., 2012; Kondolf and Pinto, 2017).
Recently, there is an increasing recognition of the need to restore and sustainably manage freshwater ecosystems associated with the increased recognition of the benefits, or services, that society derives from functional ecosystems (Findlay and Taylor, 2006; Gleick, 2000). In this light, a clear understanding of the temporal and spatial variations of fluvial landforms and processes is essential for successful management of rivers, urban reaches included (Chin, 2006; Winterbottom, 2000).
There is a wide literature worldwide concerning the morphological evolution of rivers over the last centuries and decades (e.g., Batalla et al., 2018; Gilvear and Winterbottom, 1992; Gurnell et al., 2009; Heitmuller, 2014; Kanhaiya et al., 2019; Liébault and Piégay, 2002; Michalková et al., 2011; Wyzga, 1996). Relatively fewer investigations were conducted specifically focusing on rivers affected by urbanization (Chin, 2006; Chin and Gregory, 2005). Most of them examined the influence of urbanization of catchments on downstream reaches (Ashmore et al., 2023; Booth, 1990; Brookes et al., 2005; Chin, 2006; Chin et al., 2022; da Luz and Rodrigues, 2015; Ebisemiju, 1989; Graf, 1975; Gregory et al., 1992; Leopold et al., 2005; Whitlow and Gregory, 1989; Wolman, 1967; Yorke and Herb, 1978; Yousefi et al., 2019). In contrast, few works considered specifically riverbed changes in reaches embedded within urban areas, as well as urbanization-induced landscape changes, in terms of landforms and land use, close to the riverbed and in former riverbed zones (Hohensinner et al., 2013; Leopold et al., 2005; Liu et al., 2014; Serrano Notivoli et al., 2017; Wang et al., 2019; Yousefi et al., 2018; Zerboni et al., 2021).
A substantial list of available studies investigating Italian rivers is reported in Mandarino (2022). Recently, an increasing interest in urban geomorphology, fostered by the AIGEO’s (Italian Association of Physical Geography and Geomorphology) working group dealing with this topic, resulted in significant progresses in geomorphological research on the urban environment (Ascione et al., 2021; Brandolini et al., 2020, 2021; Cappadonia et al., 2020; Del Monte et al., 2016; Guerra et al., 2021; Mandarino et al., 2021a; Pelfini et al., 2021; Porta et al., 2021; Roccati et al., 2021). However, very few works focused specifically on urban rivers (Forno et al., 2024; Foti et al., 2022; Mandarino et al., 2021b; Roccati et al., 2019) and, similarly, urbanization-induced changes in landforms and land use, which affected riverine areas and former riverbed zones, were scarcely investigated at the national scale (Mandarino et al., 2019a, 2020, 2021b, 2023; De Musso et al., 2020; Scorpio and Piégay, 2021).
Genova is one of the largest and most important coastal cities in Italy. (i) Its hazardous location at the outlet of small and steep coastal catchments, (ii) its meteo-hydrological features, and (iii) the widespread urbanization over slopes, valley floors, and coast make this city very prone to be affected by geo-hydrological events associated with intense rainfall. The metropolitan area of Genova was already investigated in numerous studies focusing on climate (Acquaotta et al., 2018, 2019; Piana and Faccini, 2023; Silvestro et al., 2012), geomorphology (Faccini et al., 2021a; Mandarino et al., 2021c; Paliaga et al., 2019b; Terrone et al., 2021), land use and land cover evolution (Faccini et al., 2016; Piana et al., 2019), past floods (Faccini et al., 2021b; Paliaga et al., 2020), geo-hydrological hazard (Cevasco et al., 2010; Faccini et al., 2018; Paliaga et al., 2019a; Piana et al., 2019; Rosso and Rulli, 2002; Rulli and Rosso, 2002), and structural works for flood protection (Redaelli and Bringiotti, 2019, 2019, 2019). However, a detailed analysis of the landscape changes that occurred along rivers is still missing.
Against this background, this paper provides a quantitative and qualitative assessment of landscape changes driven by urbanization, which occurred from the mid-19th century to the present day along the lower reaches of the Bisagno and Polcevera rivers and their tributaries in the Mediterranean coastal city of Genova. These rivers were completely embedded within the continuous urban and industrial fabric, and anthropogenic landforming activity had a profound impact on the landscape up until the total erasure of some former landforms. This work is based on collection and review of historical data, photograph interpretation, GIS analysis, and field surveys. The objectives of this research are (i) to define the effect of urbanization on river morphology, (ii) to investigate quantitatively the landscape changes over the valley floor in terms of land use and landforms, (iii) to provide useful data to inform effective and sustainable management measures to mitigate hydrogeomorphic risk and restore the fluvial environment, and (iv) to define an essential knowledge base for dissemination aiming to increase the citizens’ knowledge and awareness of the investigated urban rivers.
Study area
Genova is a coastal city located at the northern apex of the western Mediterranean Sea, in NW Italy (Figure 1). Its development was strictly related to navigation and maritime trade. As the capital of the Republic of Genova, namely, an independent state, this city was one of the major trading powers in the Mediterranean Sea and Black Sea during the first half of the second millennium and it was among the most relevant financial centre in Europe in the 16th and 17th centuries (Balzaretti, 2018; Bitossi, 2018; Khvalkov, 2017). The present-day municipality spreads over 240 km2 and along a 35 km long stretch of coast, and presents a large conurbation essentially covering the coastal area, the valley floors, and the lower parts of slopes completely. The first permanent settlement dates back to the pre-Roman period, whereas, the first traces of human occupation are associated with a small settlement located at the Bisagno River mouth and ascribable to the Neolithic (Arobba et al., 2018). The oldest part of the city is in a sort of natural morphological amphitheatre, the so-called Porto Antico Bay, situated between the downstream-most parts of the Polcevera and Bisagno catchments (Brandolini et al., 2018; Faccini et al., 2021a). The former catchment covers an area of 139 km2; the latter spreads over 97.7 km2. The maximum elevation is 1113 m and 1037 m, respectively, a few kilometres distance from the coast. Thus, these catchments are characterized by reliefs higher than 1000 m rapidly descending towards the Ligurian Sea and drained by numerous steep and short creeks. Genova is located at the junction between the Alps and the Apennines (Capponi et al., 2009; Molli et al., 2010, and references therein). The western Polcevera catchment corresponds to the Sestri-Voltaggio Zone and presents tectonometamorphic units with carbonate rocks, shales, schists, and metaophiolites (Molli et al., 2010). To the east, these units are in contact with very low-grade flysch units, essentially siltstones and shales mixed with sandstones, that dominate the central and eastern part of the Polcevera catchment. Marly-limestones outcrop at the eastern main divide of the Polcevera catchment and in most of the Bisagno catchment; this last presents also shales and slates locally. Fluvial valleys developed according to the main regional lineaments (Fanucci and Nosengo, 1977; Ferraris et al., 2012), steep slopes and narrow valleys are displayed where marly-limestones outcrop and over the Sestri Voltaggio Zone, and, in contrast, gentler slopes and wider valleys characterize the areas with a clayey bedrock. The downstream-most sector of the main valleys presents small and narrow alluvial plains. Location of the study area. The Bisagno and Polcevera catchments are depicted in (a) while the investigated valley floors (black meshed area in (a)) are shown in (b). The overall Bisagno and Polcevera study reaches (b) are referred to as BSR (including B1, B2, and B3) and PSR (including P1, P2, and P3), respectively. Basemap in (b): the most recent Google Earth image. For interpretation of the references to colours in this figure legend, refer to the online version of this article.
The climate is overall referred to as Mediterranean (Fratianni and Acquaotta, 2017). However, differences exist between the coastal area of Genova and the inland valleys (Acquaotta et al., 2019; Bignami and Biagi, 2018). Dry and hot summer and mild winter characterize the coastal sector, while more temperate conditions, warm summer, and cold and wet winter characterize the inlands. Referring to the period 1991–2010, the annual mean temperature is 12–13°C in the uplands and around 16°C at the coast (ARPAL, 2013). The annual mean rainfall ranges between little under 1400 and just over 1800 mm in the uplands, and from 1100 to 1300 at the coast, with most of rainfall occurring in autumn and the minimum in summer (ARPAL, 2013; Faccini et al., 2015). These differences are associated with the physical setting of the Genova Gulf, the orographic uplift, and the atmospheric circulation prevailing over the Ligurian Sea conveying moist air masses from the sea to the relief around Genova. These factors result in frequent intense rainfall events and high annual rainfall values occasionally >3000 mm. Combined with widespread urbanization, they make the Polcevera and Bisagno catchments, as well as the city of Genova, highly prone to severe floods (Brandolini et al., 2012; Brandolini and Terranova, 1994; Faccini et al., 2015, 2018; Silvestro et al., 2012). In recent decades, a net increase in rainfall intensity occurred, since the rainy days have decreased and the total annual rainfall has basically remained constant (Acquaotta et al., 2018, 2019). This variation may be interpreted as a possible evidence of a shift towards a more extreme climate regime (IPCC, 2022; Tamm et al., 2023).
The Polcevera and Bisagno catchments are heavily urbanized in their middle and lower parts, particularly in the main valley bottoms. The upper part and, in general, the steepest slopes are prevalently covered by woods while cultivated areas and little villages are somewhat scattered (Brandolini et al., 2012; Piana et al., 2019). After the Second World War (WW2), severe depopulation occurred in the mountainous portions of these basins due to significant social and economic changes. This led to the abandonment of numerous agricultural areas and the consequent increase in forest cover (Faccini et al., 2016; Piana et al., 2019). Except for the main valley bottom, both catchments did not undergo large urbanization upstream of the study reaches.
This research focuses on the hydrographic network draining the valley floor of the lower Polcevera and Bisagno valleys and on the valley floor itself. The Polcevera and Bisagno lower reaches, 7.2 km and 5.5 km long, respectively, and the lowland reaches of their tributaries were examined. In the first half of the 19th century, within an agricultural and periurban landscape, these water courses were already conditioned by anthropogenic interventions, such as bank protections and diversions, aimed to defend cultivated fields and settlements, and use water for irrigation and waterpower. Later, urbanization profoundly impacted the whole study area. The overall investigated valley floor covers 483 ha in the Polcevera catchment and 252 ha in the Bisagno catchment. These two valley floors are bordered by steep slopes that are largely urbanized and mostly drained by first order creeks, and the outlets of the main tributary valleys.
Materials and methods
A multi-temporal analysis based on historical information collection and review, map and aerial photograph interpretation, field observations, and quantitative investigations in a GIS environment was performed. The study area corresponds to the valley floor. The riverbed planform features and morphological dynamics were examined focusing on both the study reach as a whole and smaller single reaches for more detailed investigation. As defined in previous research (Brierley and Fryirs, 2013; Gurnell et al., 2016), these last are river portions currently characterized by homogeneous hydrogeomorphic features and are generally <5 km in length. The valley floor was split up according to the aforementioned reaches, and (i) the open-air minor hydrographic network (MHN) extent over time, (ii) the land use and land cover of the valley floor over time, and (iii) the present-day accessibility of river margins were quantitatively examined following the spatial scales described before. All GIS analyses were performed by using QGIS (QGIS.org, 2023).
Data sources
Summary of data sources considered in this study.
*Retrieved from the geoportal of the Municipality of Genova; ** only small parts of the study area are covered by 1934 maps, therefore 1939 was considered as the reference year for all the 1930s data in the overall analysis. WMS: web map service.
Field surveys enabled the collection of additional information on the present-day geomorphological conditions of the riverbed and the valley floor, as well as the validation of data derived from photograph interpretation. Furthermore, large bibliographic research was conducted to gather historical information from archival documents and contemporary publications.
Analysis of morphological adjustments along the Bisagno and Polcevera rivers
Morphological features and morphological adjustments along the Bisagno and Polcevera rivers were investigated quantitatively to detect changes in morphology and dynamics over the period considered.
Digitization of bankfull channel and valley floor
The riverbed was manually mapped as a polygon on each dataset (Table 1) based on photointerpretation and field evidence. It was defined as the bankfull channel, namely the area occupied by base-flow channels, bars, islands, and other possible in-channel vegetation units, according to the “geomorphic-unit-oriented” definition of bankfull channel reported in Belletti et al. (2017), Rinaldi et al. (2015a, 2017), and Zhang and Fryirs (2024). In urban reaches characterized by continuous or almost continuous artificial banks, the bankfull channel boundaries corresponded to the artificial bank edges. Similarly, the valley floor, namely, the sub-flat surface located between the left and right slope toes, was manually mapped as a polygon on each dataset.
Quantitative assessment of bankfull planform adjustments
The bankfull channel centerline (BCC), namely, the line of points equidistant from the banks, was derived automatically using skeletonization and smoothing techniques (Block, 2014; Cencetti et al., 2017; Mandarino et al., 2019b). The bankfull channel area (BCA), namely, the planform extent of the riverbed, and the length of the bankfull channel centerline (BCL) were computed; their ratio provided the average bankfull channel width (BCW) (Clerici et al., 2015; Surian et al., 2008). The bankfull channel width variation index (ΔBCW) was computed through the formula
Given that the general direction of the riverbed remained relatively stable during the considered period, the sinuosity index (SI) was computed at reach scale by dividing the BCL by the constant length of the straight line connecting the endpoints of the reach, to compare the values referred to different years (Rinaldi et al., 2016a, and references therein). The braiding index (BI) and the anastomosing index (AI) were calculated by averaging the count of baseflow channels separated by bars and islands, respectively, intersecting equally-spaced transects located along and perpendicular to the BCC at the reach scale (Rinaldi et al., 2016a, and references therein). The distance between transects was 70 m, for Bisagno, and 80 m, for Polcevera, corresponding to the most recent BCW of the overall study reach. Furthermore, the confinement degree (CD) (Brierley and Fryirs, 2013; Rinaldi et al., 2013, 2016a), defined as the percentage of river banks not directly in contact with the plain but with hillslopes or ancient terraces, over the total length of the two banks, and the confinement index (CI) (Rinaldi et al., 2013), defined as the ratio between the floodplain width (riverbed included) and the channel width, were computed. The channel pattern was then defined according to the classification scheme developed by Gurnell et al. (2016) and Rinaldi et al. (2016a).
Finally, the channel abandonment index (CAI), which indicates the amount of abandoned riverbed area relative to the oldest riverbed, and the channel formation index (CFI), which reveals the extent of acquired riverbed area relative to the most recent riverbed, were computed to describe channel planform displacements over time (Mandarino et al., 2021b). They were defined (i) for each consecutive period, namely, any two adjacent sets of imagery, and (ii) for the whole investigated time span, with each couple of data referring to as time 1 (t1), the oldest, and time 2 (t2), the most recent. These parameters were computed using the following equations
Morphological dynamic index
In order to evaluate the morphological dynamics of the Bisagno and Polcevera rivers at the reach scale, and consequently, indirectly assessing its hydrogeomorphic hazard conditions, the Morphological Dynamic Index (MDI) was computed (Rinaldi et al., 2015b). This tool belongs to the IDRAIM system (Rinaldi et al., 2015b, 2017), which is a methodological framework for hydro-morphological evaluation, analysis, and monitoring of water courses, whose development was promoted and funded by the Italian national Institute for Environmental Protection and Research (ISPRA).
The MDI defines and classifies the degree of channel dynamics associated with progressive changes occurring over a relatively long time scale, excluding the possible responses to extreme flood events (Rinaldi et al., 2015b; 2016b). This index is based on a scoring system presenting a set of 11 indicators grouped into three main components (Rinaldi et al., 2015b): (i) morphology and processes, which focuses on the riverbed features and current (i.e., on a time scale of the last 10–15 years) processes; (ii) artificiality, which refers to the presence of structures conditioning channel dynamics; and (iii) channel adjustments, which is related to morphological changes occurring over a relatively long time period (i.e., about the last 100 years). Most indicators are described into five classes ranging from class A, representing negligible dynamics, to class E, indicating the highest level of morphological dynamics; the scores are progressively higher from A to E. The MDI is defined as the ratio between the sum of the scores of all the indicators and the maximum score that could be reached when all appropriate indicators are in the highest class. Therefore, MDI ranges from 0 (minimum morphological dynamics) to 1 (maximum morphological dynamics), and five distinct classes were established as follows: (i) very low, 0.0≤MDI<0.2; (ii) low, 0.2≤MDI<0.4; (iii) medium, 0.4≤MDI<0.6; (iv) high, 0.6≤MDI<0.8; and (v) very high, 0.8≤MDI≤1.0.
Minor hydrographic network mapping
The minor hydrographic network (MHN) draining the valley floor, intended as the combination of unculverted creeks, ditches and channels, Polcevera and Bisagno rivers excluded, was manually mapped as a line on 1848, 1878, and 1930s maps, and 2018 orthophotos. Changes in MHN length over time (∆MHN) were assessed to investigate the culverting process.
Land use and land cover mapping
The land use and land cover (LULC) related to the valley floor was manually mapped on the 1848, 1878, and 1930s maps. The LULC vector layer at 1:10,000 scale, dated back to 2019, and freely-available at the Regione Liguria Geoportal (https://geoportal.regione.liguria.it/) was considered to depict the most recent LULC condition. According to the first level of LULC classification established within the frame of the Corine Land Cover inventory (Bossard et al., 2000; Feranec et al., 2016), and due to the impossibility of defining more detailed categories on historical maps with an adequate level of accuracy, four LULC classes were considered: (i) artificial area, (ii) agricultural area, (iii) natural and semi-natural area, and (iv) water body. In this study, the water body class includes the bankfull channels, as previously defined, of the Bisagno and Polcevera rivers, and their tributaries.
Accessibility assessment
The physical accessibility to the bankfull margins of the Bisagno and Polcevera rivers was assessed as a proxy to investigate the level of possible social connection between the river and citizens (Kondolf and Pinto, 2017), assuming that an accessible riverfront (i) is essential for allowing activities along the margins themselves and (ii) leads people to the river and to realize its presence, its relevance in terms of ecosystem services, and the related risks. In particular, focusing on the longitudinal connectivity (Kondolf and Pinto, 2017) and following the approaches developed by Che et al. (2012) and Hermida et al. (2019), the accessibility of the river margins was investigated by classifying them into five categories according to the type of road: (i) river margin with vehicular and pedestrian path; (ii) river margin with vehicular path; (iii) river margin with pedestrian path; (iv) river margin with railway path; and (v) river margin with no path.
Results
The study reach on the Bisagno (BSR) and that on the Polcevera (PSR) were each divided into three reaches (Figure 1(b)). Their limits correspond to (i) in-channel structures (i.e., the limits between B1 and B2, B2 and B3, P1 and P2, and P2 and P3), such as weirs and bed-revetments, (ii) tributary confluences (i.e., the limit between P2 and P3), (iii) changes in channel morphology (the limit between B2 and B3), and (iv) changes in channel width (i.e., the limit between P1 and P2). Figure 2 shows the variations in the extent of the bankfull channel and the MHN over the examined period. Bankfull channel and minor hydrographic network changes over time over the valley floor of the Bisagno (a) and Polcevera (b) rivers. The light blue polygons (lines) represent the bankfull channel (unculverted minor hydrographic network); the red dotted line indicates the perimeter of the valley floor, highlighting reach limits. For interpretation of the references to colours in this figure legend, refer to the online version of this article.
Bankfull planform adjustments
The Bisagno River reaches experienced changes in BCW characterized by a common trend of progressive decrease from 1848 to 1939, while no variations occurred in the period 1939–2018 (Figure 3(a)). As for the Polcevera (Figure 3(a)), reaches P2 and P3, along with the PSR, showed a marked reduction in BCW between 1848 and 1878. Then, narrowing never >11 m was observed in all reaches and in all considered periods, except for P1 where BCW slightly increased in the period 1939–2018. The ΔBCW further stresses the aforementioned trends related to the width variation over time (Figure 3(b)). No positive ΔBCW were registered and the only widening process was documented in reach P1 during the period 1939–2018. Bankfull channel width (BCW) (a) and bankfull channel width variation (∆BCW) (b) through time at the reach scale (B1, B2, B3, P1, P2, and P3) and for the whole study reach (BSR and PSR).
The downstream-most reaches experienced an increase in BCL, +15% in B3 and +9% in P3, from 1939 to 2018 (Figure 4(a)). BCL remained substantially constant in the other reaches and periods, except for P2 where a reduction by 11% was registered between 1848 and 1878. The overall change in BCL was negligible for the PSR and +4% for the BSR in the period 1848–2018. Since BCA is the product of BCW and BCL, and BCL changes very little, BCA follows the trend of BCW, except for B3 and P3, where a slight BCA increase was documented in the most recent period (Figure 4(b)). Referring to the whole investigated period, the overall BCA reduction was 45% for the Bisagno and 65% for the Polcevera. Bankfull channel planform features over time. Bankfull channel length (BCL), Bankfull channel area (BCA), sinuosity index (SI), and braiding index (BI) through time at the reach scale and for the whole study reach along the Bisagno (B1, B2, B3, and BSR) and Polcevera (P1, P2, P3, and PSR) rivers. SI and BI are dimensionless parameters. Legends in (d) are, respectively, referred to all the above plots. In (b) the outlier referred to the BCA of the whole study reach (PSR) in 1848 is 1,618 ha. The grey filling in (c) and (d) highlights that SI and BI cannot be <1.
The SI overall assumed very low values (Figure 4(c)). In Bisagno it was constantly 1; in Polcevera it ranged between 1.0 and 1.2. Braiding phenomena progressively disappeared in both Bisagno and Polcevera (Figure 4(d)). The BI relative to the Bisagno decreased from 1.2–1.5 to 1 up to the 1939; no data were available for B3 in 1939 and 2018. Similarly, a progressive reduction in BI from 1.5–2, in 1848, to 1, in 2018, was observed along the Polcevera River. The AI was 1 in all reaches over time; thus, no islands were mapped.
Bankfull channel confinement in 1848 and 2018.
CI: confinement index; CD: confinement degree; PC: partly confined channel; U: unconfined channel. CI is a dimensionless parameter.
Bankfull channel pattern through time.
W: wandering; B: braided; SAB: sinuous with alternate bars; HM: highly modified; StAB: straight with alternate bars; U: undefinable (because it is culverted). Channel pattern in parentheses indicate the overall planform setting of highly modified (HM) reaches.
CAI and CFI highlighted that most bankfull channel was abandoned in the period 1848–1878 in Polcevera reaches, and in the period 1878–1939 in Bisagno reaches, against an overall absence of bankfull channel acquisition (Figure 5). These trends revealed that a non-negligible planform migration occurred in Bisagno (slight migration) and Polcevera (moderate migration) reaches in periods 1878–1939 and 1848–1878, respectively. This migration was mainly associated with the documented bankfull channel narrowing and did not correspond to an overall bankfull channel displacement. The channel abandonment index (CAI) and channel formation index (CFI) referred to each investigated period and to the period 1848–2018. Letters on bars indicate the qualitative class for metrics: VL: very low (≤15%); L: low (>15% and ≤35%); M: moderate (>35% and ≤50%); H: high (>50% and ≤75%); VH: very high (>75%). Letters in italic style indicate the bankfull channel planform migration based on the combination of CAI and CFI. I: intense; M: moderate; S: slight; N: no/very slight. The symbol ** after letter indicates that the bankfull channel migration process is mainly associated with the abandonment of bankfull channel area.
Morphological dynamic index
Morphological dynamic index at the reach scale and for the overall study reaches.
*The procedure for MDI assessment would lead indicator M2 in class C (erodible non-cohesive alluvial banks ≤66% of total bank length, or cohesive ≤90%), even if erodible non-cohesive alluvial banks are ≤10% of total bank length (reflecting class A) due to bank protections, because there is no vegetation on banks. This rule is associated with the potential higher proneness of bare banks to be eroded (Abernethy and Rutherfurd, 1998; Beeson and Doyle, 1995; Wynn and Mostaghimi, 2006). However, in urban areas where the riverbed is totally channelized up to the bank edge, like the study reaches are, the application of the aforementioned rule would lead to overestimate the potential dynamics. For this reason, class A was defined for M2 in all reaches.
Letters A, C, D, and E represent the indicator class (Rinaldi et al., 2015b); na: not applied. Indicators M1 to M6 refer to morphology and processes, A1 and A2 to artificiality, and CA1 to CA3 to channel adjustments. MDI score expresses the ration between the sum of the scores associated with indicators, and the maximum score that could be reached when all appropriate indicators are in the highest class.
Ultimately, all reaches, along with the overall BSR and PSR, showed low values of MDI, except for B3 which is characterized by a very low MDI. The only differences in MDI among reaches are associated with the presence of in-channel transversal structures and bed revetments (i.e., indicators M3 and A2).
Minor hydrographic network
In the Bisagno valley floor, the extent of the unculverted MHN was 4.35 km in 1848, 2.93 km in 1878, 0.35 km in 1939, and 0.26 km in 2018. In Polcevera valley floor these values were 14.79 km, 7.4 km, 2.22 km, and 1.9 km. Figure 6 shows ΔMHN for single reaches and the whole study reaches. Only negative values were observed, with an evident reduction in MHN values up to 1939 (Figure 2). Between 1939 and 2018 values <6% and <12% were documented over the Bisagno and Polcevera valley floors, respectively. In the period 1878–1939 the unculverted minor hydrographic network completely disappeared in reaches B2 and B3. Variation in minor hydrographic network (MHN) length over time at the reach scale (B1, B2, B3, P1, P2, and P3) and for the whole study reach (BSR and PSR).
Land use and land cover changes
The results concerning the LULC evolution highlight a progressive disappearance of agricultural areas and watercourse areas, accompanied by a generalized increase in artificial areas (Figures 7 and 8). In the Bisagno valley bottom, the largest reduction in water body surface occurred between 1878 and 1939. In contrast, in the Polcevera valley bottom, this happened between 1848 and 1878. In the late 1930s, the artificial area amounted to 84.3% and 84.4% along the Bisagno and Polcevera, respectively, and the rest was essentially occupied by water bodies. The natural and semi-natural areas were only found in B3 and P3, and corresponded exclusively to the beach located close to the mouth of the Bisagno and Polcevera rivers. Land use and land cover of the Bisagno valley floor in 1848, 1878, 1939, and 2018 at the reach scale (B1, B2, and B3) and for the whole study reach (BSR). For interpretation of the references to colours in this figure legend, refer to the online version of this article. Land use and land cover of the Polcevera valley floor in 1848, 1878, 1939, and 2018 at the reach scale (P1, P2, and P3) and for the whole study reach (PSR). For interpretation of the references to colours in this figure legend, refer to the online version of this article.

Considering the overall BSR, the artificial area over the valley floor increased by 124% in 1878, 353% in 1939, and 387% in 2018 compared to its extent at the beginning of the study period. Similarly, referring to PSR, the artificial area increased by 315% in 1878, 651% in 1939, and 780% in 2018. The overall reduction in water course area amounted to a total of −17% in 1878, –58% in 1939, and −66% in 2018 over the Bisagno valley floor, and −61% in 1878, −65% in 1939, and −69% in 2018 over the Polcevera valley floor.
Accessibility
Figure 9 displays the accessibility of the river margins according to the type of road. No pedestrian paths are located along the margins of the Bisagno and Polcevera rivers. As for the former, accessibility is allowed through vehicular and pedestrian paths and vehicular paths. The highest values are associated with the first category in reaches B1 and B2, and in BSR. As for the latter, most of accessibility (>50%) is associated with vehicular paths in all reaches. The remnant (18.6% referring to PSR) is related to vehicular and pedestrian paths and railway paths. The most inaccessible reaches are B3 and P3. Accessibility of the river margins according to the type of road at the reach scale (B1, B2, B3, P1, P2, and P3) and for the whole study reach (BSR and PSR) along the Bisagno and Polcevera rivers.
Discussion
Effects of urbanization on river morphology
The differences in water body area variation, bankfull planform features, and ∆MHN observed over time between the Bisagno and Polcevera valley floors are associated with different anthropogenic interventions, intended as different type and timing of urbanization affecting the investigated areas (Figures 7 and 8).
The valley floor of the Bisagno River experienced a gradual urbanization, namely, construction of road infrastructures, facilities, and residential buildings, from the mid-19th century to the late 1930s (Piana et al., 2019; Rosso, 2014). This affected the riverbed area especially from the late 19th century onwards, causing relevant bankfull channel narrowing associated with widespread channelization works (Figure 10). Additional urban expansion in the 1930s resulted in the almost-complete (89%) culverting of the reach B3. New culverts (i.e., 17% of reach B1 and 14% of reach B2) were also constructed in the late 1980s and early 1990s, further reducing the water body area. Landscape changes over time along the Bisagno River. Reach B1 in the late 19th century (a), in approximately the second half of the 1920s (b), and today. The white triangle in (a), (b), and (c) indicates the same cemetery corner. The white arrow in (b) indicates the urban expansion and the related riverbed narrowing occurred at the slope toe in the early20th century. In (d), the coverage of the downstream-most reach occurred in the 1930s is displayed. The present-day riverbed is shown in (e) and (f). The former shows the culvert close to the stadium and the latter the culvert of the downstream-most reach; here the river is below the road located in the middle of the photograph and the river mouth can be observed in the distance at the end of the aforementioned road. The white dotted arrows indicate the flow direction. Image sources: “C’era una volta Genova” group (a, b), Andrea Mandarino (c, e, and f), “Antica Foce” group (d). For interpretation of the references to colours in this figure legend, refer to the online version of this article.
Differently form the aforementioned Bisagno case, a single-urbanization event impacted severely on the Polcevera River, namely, the railway line construction in 1854. It resulted in narrowing and straightening the Polcevera bankfull channel through channelization works (Brandolini et al., 1994), and entailed the anthropogenic cut-off of a river bend in reach P2 (Figure 11). Here, a new, narrower (70–75 m wide, compared to the previous 150–200 m) riverbed was excavated and displaced about 300 m towards the west; the abandoned channel was filled in with debris dug from the railway tunnels and was used as a preferential settlement for urban expansion. Moreover, this riverbed straightening resulted in further reclaimed lands soon after used for urbanization. Thus, the railway embankment allowed a large expansion of artificial area, primarily linked to industrial growth, and then integrated with residential buildings and other structures and facilities, to be carried out. The reduction in sinuosity, length and area documented during the first period (Figure 4) is associated with this change in river path, which caused also an increase in channel slope and, consequently, channel incision. The simultaneous in-channel extractions of sediments for construction purposes, including the building of railway embankments, exacerbated erosion problems in the area (Leale and Tomaselli, 2012). The present-day riverbed path was essentially already defined in the 1878 maps. After 1878 further minor anthropogenic interventions were implemented repeatedly to accommodate urban expansion and consolidate channelization works. The Polcevera River in 1848 (a) and 2018 (b) in reach P2. The anthropogenic cut-off of the river bend is evident and highlighted by the 1848 bankfull channel (white polygon) superimposed on the 2018 image. The white dotted arrows indicate the flow direction. For interpretation of the references to colours in this figure legend, refer to the online version of this article.
Thus, the banks of both the Bisagno and Polcevera rivers were overall established by building high embankments and walls, acting as both bank protection structure and, locally, as levee, which stabilized the bankfull channel location. Behind these structures, fillings were carried out to shape a levelled and higher ground level than before (Figure 12). As a result, the adjacent plain was completely disconnected from the riverbed and became an anthropogenic recent terrace. No floodable areas close to the bankfull channel were planned. In the 20th century, consolidation weirs, bed sills, and bed revetments were constructed along both the Bisagno and Polcevera rivers to prevent erosive phenomena typically induced by channelization (Ashmore et al., 2023; Hohensinner et al., 2018; Simon and Rinaldi, 2006). Channelization works resulted in a highly-modified channel pattern, since (i) the notable geomorphic stability, as highlighted by the MDI and documented even in case of high-magnitude floods (e.g., in 1970, 1992, 2011, and 2014), and (ii) the geomorphic units of reaches cannot reflect any ‘natural’ boundary conditions (Rinaldi et al., 2016a). Despite this, a kind of morphological classification for the HM reaches was established (Table 3) with the sole purpose of characterizing the in-channel forms. It is noteworthy that the frequent in-channel interventions of bankfull reprofiling and vegetation removal (Comiti et al., 2011) performed over the last decades by means of bulldozers furtherly prevented the spontaneous development of in-channel forms. Moreover, they represented a limit in field survey activity as they obliterated the field evidence of ongoing geomorphological processes. The Bisagno River in 1926 (reach B1). This photograph displays the channelization process that resulted in a relevant riverbed narrowing. The area behind bank protections in the foreground was infilled and, thus, the ground-level increased. The well-evident bridge in the distance displays 6 arches (like it is today) of the original 18. The white dotted arrow indicates the flow direction. Image source: Municipal Archive.
Both terminal stretch B3 and P3 showed elongation because of the construction of large sea embankments, undertaken for urban expansion purpose, which extended the river mouths towards the sea. This landscape change and localized excavations of the slope toe at the plain margin (Figure 10(b) and 13) widened the valley floor. The combination of bankfull channel narrowing, which led the riverbed itself far from slopes, along with the enlargement of the valley floor, resulted in the documented changes in confinement class (Table 2). The Polcevera River and its adjacent areas in 1848 (a) and 2018 (b) in reach P3. This comparison highlights the change in valley floor extent associated with the excavation of slopes. Contour lines in (a) show the presence of a relief that is missing in (b). The valley floor margins referred to 1848 (white dotted line) and 2018 (white continuous line) are depicted in (b). The white dotted arrows indicate the flow direction. For interpretation of the references to colours in this figure legend, refer to the online version of this article.
In this study, the comparison of current and past elevation differences between the riverbed and adjacent areas could not be used as a reliable indicator of changes in bed level over time. This is because the riverbed elevation was stabilized by the aforementioned transversal structures, and widespread anthropogenic ground level changes occurred across the valley floor, as shown in Figure 12 and discussed by Piana et al. (2019). This issue, along with the lack of systematic and long-term data, made the definition of both recent and historical in-channel vertical adjustments trends rather challenging and deferred to further investigation. However, it can be hypothesized that during the second half of the 19th century until stabilization with barriers, the Bisagno and Polcevera rivers experienced a condition of sediment deficit (i.e., channel degradation) most likely associated with (i) in-channel sediment quarrying and the aforementioned channelization within the investigated reaches and upstream of them (Leale and Tomaselli, 2012), and (ii) the increase of forest cover (Faccini et al., 2016) and torrent control works (at the catchment scale), according to what was documented at the Italian national scale during the 20th century (Preciso et al., 2012; Scorpio et al., 2015; Scorpio and Piégay, 2021). It’s not excluded that bed-level was partially recovered in certain reaches through the use of transverse structures.
The MHN consisted of ditches, artificial channels, and the main tributaries before urbanization. Then, they were progressively enclosed in concrete channels and hidden underground, preventing the lateral connectivity with the plain. Furthermore, roads and rainwater drainage systems introduced extensions of this network, resulting in an anthropogenic increase of drainage density (Booth, 1991; Chin, 2006; Whitlow and Gregory, 1989).
As the city encompassed the rivers, they remained largely disconnected from the city itself in terms of accessibility for citizens, especially referring to the Polcevera River. Today, pedestrians can walk along some portions of the Bisagno and Polcevera riverbeds only along a vehicle road with a footpath. The absence of areas and paths allowing citizens to meet and do activities along river margins, along with the complete disappearance of some reaches underground, resulted in a progressive detachment of the population from the watercourse (Wantzen, 2022), which led to the loss of knowledge and awareness of the river systems (Botzen et al., 2009; Dempsey et al., 2018).
Geomorphological evolution and comparison with other rivers
Three phases of morphological evolution were identified for the Bisagno and Polcevera rivers: (i) pre-urbanization phase, namely, up until the mid-19th century, in which some constraints already existed but the riverbed morphology and dynamics were able to develop spontaneously; (ii) urbanization phase, namely, from the mid-19th century to the late 1930s, in which the largest riverbed adjustments associated with channelization occurred; and (iii) post-urbanization phase, namely, from the late 1930s to the present day, in which planform stability dominated and some minor channel adjustments were due to further anthropogenic interventions. Since there was no large urbanization at the catchment scale upstream of the study reaches (Piana et al., 2019), urbanization at the reach scale was the main driver of morphological changes rather than urbanization-induced changes in water and sediment fluxes on which most literature focusses (see the introduction). The aforementioned morphological trend is totally different from what was commonly observed along Italian rivers over the last two centuries, namely, two historical phases of channel degradation (late 19th century–1950s and 1950s–1990s) followed by an emerging trend to recovery (Comiti et al., 2011; Mandarino, 2022; Scorpio et al., 2015; Surian et al., 2008, 2009). The few available studies that specifically focused on urban reaches highlighted that the riverbed morphological changes were strictly associated with the phases of urban expansion, as documented on the Po River and its tributaries in Torino (Forno et al., 2024), the Tanaro River in Alessandria (Mandarino et al., 2021a), the Centa, Arroscia, and Neva rivers in Albenga (Mandarino et al., 2021b), the Entella River in Chiavari and Lavagna (Roccati et al., 2019), and several rivers located in Calabria (Foti et al., 2022). In these cases, urban-induced riverbed changes consisted of narrowing, straightening, bank stabilization, and, sometimes, culverting and bed stabilization. It is noteworthy that such anthropogenic pressures also affected non-urban rivers, as in the case of the Adda (Turconi et al., 2023), Adige (Scorpio et al., 2018), and Scrivia (Mandarino et al., 2019a) rivers, where a completely stabilized riverbed is encompassed within an agricultural landscape.
Similar (in type but not in timing) geomorphic changes to those documented along the Bisagno and Polcevera rivers, were observed in other cases worldwide. For example, Hohensinner et al. (2013) described the Danube riverscape changes, essentially including channel narrowing and stabilization, associated with the development of Vienna. In Suide County (China), more than 1/3 of the area affected by urban expansion from 1990 to 2017 belonged to the Dali and Wuding rivers (Wang et al., 2019). Riverbed area decrease and channelization due to urbanization were documented also in the central sector of the Pyrenees in the period 1940–2012 (Serrano Notivoli et al., 2017) and in Wuhan (China) after the 1950s (Liu et al., 2014). The Los Angeles River was channelized in the mid-19th century to protect the city from floods, and today the river is made almost entirely of concrete (Gumprecht, 2001). The Faustino and Meio rivers (Brazil) were almost completely filled and replaced by buildings and roads in the 1980s (Silva de Souza et al., 2021).
Given the above, the Urban Stream Syndrome (Walsh et al., 2005), typically addressing urbanization-induced hydrogeomorphic changes in river reaches downstream of urban areas (Vietz et al., 2016b), can also apply to river reaches encompassed within urban areas. In the former case, the geomorphic response of rivers to catchment urbanization includes channel deepening, widening, and instability (Vietz et al., 2016b, and references therein). In contrast, in the latter case, which includes the investigated reaches, rivers generally undergo narrowing and straightening during urbanization, followed by stability due to channelization works preventing any geomorphic response to urbanization of riverine areas.
Implications for river management
The urban expansion that occurred over the valley floor of the Bisagno and Polcevera rivers, and overall over their catchments, is an evident cause of the increased flood risk (Faccini et al., 2015, 2016), as largely documented elsewhere (e.g., Cao et al., 2022; Cooke, 1984; Tellman et al., 2021; Zhang et al., 2008). Moreover, the recent increase in frequency of intense rainfall events most likely associated with global warming (Gaume et al., 2009; Hall et al., 2014; IPCC, 2022; Najibi and Devineni, 2018) worsened the flood hazard condition (Acquaotta et al., 2019).
Against this background, and considering that urban rivers are undoubtedly the most degraded river systems (Francis, 2014), what kind of management should be implemented? Urban rivers offer opportunities for recreational land use as well as problems in control and maintenance (Wolman, 1967). Therefore, management methods need to be chosen to suit the characteristics of different areas (Chin and Gregory, 2009). Wolman (1967) and Brookes (1988) recommended the need for a holistic approach, placing the urban drainage in its catchment context. House and Ellis (1994) advocated a combination of organizational and institutional cooperation and planning, as well as the implementation of sound and sustainable engineering and ecological principles.
Here, a program for geomorphic monitoring of the main riverbeds and the MHN is essential to gather data that form the basis for developing management strategies. These last should aim to mitigate flood risk and recover the fluvial environment, not only from the geomorphological point of view.
Re-establishing the pre-urbanization landscape conditions is impossible as it would need social, political, and economic support far different from what it might have today. However, small-scale river rehabilitation interventions, even those with a social focus aimed at enhancing waterfront aesthetics, merit consideration (Guimarães et al., 2021; Kurth et al., 2015; Pinto and Kondolf, 2020; Scorpio et al., 2018), especially under urban renewal. Brownfields may represent effective possibilities for land acquisition aiming to realize both multi-use riparian spaces (Prior, 2016) and stormwater control measures (SCMs; Vietz et al., 2016a), as the case of Alicante (Hernández and Morote, 2019). Moreover, small and dispersed SCMs over the valley floor should be implemented to limit runoff input to riverbeds, such as tanks that use stormwaters for domestic and industrial purposes (Vietz et al., 2016a).
Recently, two underground spillways are being built, one for the Bisagno River and one for one of its tributaries, to divert part of the floods directly to the sea through a new channel. Furthermore, during the past few years, the covering of the final stretch of the Bisagno River has been reconstructed to achieve a slight increase in the maximum conveyable discharge.
In such an urbanized context, this is not enough. Any further soil consumption has to be stopped, no further vulnerable elements have to be located in floodable areas, and flood warning systems need for maintenance and enhancement. The intent should be to reduce vulnerability by increasing the community resilience to future hazardous events. In this perspective, increasing the population’s knowledge and awareness of river systems is crucial. Improved, preferably high-quality accessibility would be desirable to establish a new close relationship between the river and the riverside community.
Conclusions
This research investigated the geomorphological changes experienced by rivers that were encompassed within urban areas. Urbanization essentially occurred between the mid-19th century and the late 1930s, causing significant alterations of (i) the morphology and dynamics of the main rivers, (ii) the morphology and dynamics of the MHN draining the valley floor, and (iii) the overall landscape of the valley floor itself. The bankfull channel of the Bisagno and Polcevera rivers experienced large narrowing associated with widespread establishment of channelization structures. The MHN essentially disappeared underground. After the 1930s, further minor anthropogenic interventions were implemented repeatedly to accommodate urban expansion and consolidate channelization works; however, the overall geomorphological setting of both the main riverbeds and the MHN remained unchanged. The geomorphological evolution of the Bisagno and Polcevera rivers followed three different phases associated with the urban expansion process. Nowadays, the study area is completely urbanized, the fluvial stems flowing through the valley floor are totally disconnected from their adjacent areas, the main rivers show a relevant geomorphological stability induced by anthropogenic pressures, and the accessibility of river margins for people is rather limited.
The aforementioned processes arose serious issues in terms of flood risk. Future management should be focused on mitigating flood risk, improving the fluvial environment, and establishing a new positive relationship between rivers and citizens. In this perspective, the outcomes from this study represent an essential knowledge base for technicians and policymakers to plan urban development and inform effective and sustainable management measures. Moreover, they provide insights into the medium- and short-term morphological evolution of river reaches encompassed within urban areas. The approach implemented in this research could represent a reference scheme for further analysis on urban rivers. The metrics considered allowed the geomorphological features and dynamics of the bankfull channel to be investigated in detail; furthermore, this case represents, to our knowledge, the first MDI application to urban rivers. Dissemination of the results is essential to increase the citizens’ knowledge and awareness of river systems. A large-scale replication of this research in other urban rivers would enlarge the spectrum of available information for river management.
Footnotes
Acknowledgements
The authors thank the two anonymous reviewers for their valuable comments that improved the quality of the manuscript.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: (i) the DM 1062/2021 FSE REACTEU PON Project—Green (CUP N. D31B21008270007 University of Genova), (ii) the Ministero dell’Università e della Ricerca of Italy in the frame of PRIN 2022 project (funded by European Union – Next Generation EU) – Project title: GEOTRes – Geoheritage threating and resilience: mapping the impact of geomorphic and human processes in sensitive morphoclimatic environments (Prot. 2022NYS98X; CUP: D53D23004810006; Principal investigator : R. Azzoni; Responsible of Genova University Unit: P. Brandolini), and (iii) the Research funds of the University of Genova (FRA2022-Brandolini 100022-2022-PB-FRA_001 and FRA2023-Mandarino 100022-2023-AM-FRA_001).
