Most hydrogeological instability phenomena have a recurrent nature: they have already occurred in the past, in the same sites, often with equal or higher intensity. Landslides are a case in point. Predicting landslide hazards requires knowledge of slope lithology, morphology and hydrogeology, as well as a retrospective analysis of prior slope movements and their scale in the investigated area. Moreover, identifying landslide hazards supposes static conditions at a given point in time. Thus, landslide risk assessment should not only define the levels of hazard and vulnerability of a given site, but also and above all anticipate their evolution throughout time, since the extent of geo-environmental risks cannot be projected indefinitely over time.
Risk conditions arising from landslides are directly correlated with their types and rates of evolution. From this standpoint, slope movements may be distinguished into moderately- to slow-moving landslides and fast-moving mud and debris flows.
The latter may be triggered by heavy rainfall, e.g. storms. One of these events recently involved the Ionian side of the Messina area. The disaster occurred in an area at high hydrogeological risk that had already been hit by landslides and floods in 2007. In particular, the origin and effects of the mud and debris flows were similar to those that ravaged the Sarno area in Campania in 1998 (De Vita, 2000, Gullà and Antronico, 2003). Under the classification commonly used in the international literature, these gravitational movements are defined as mud flows or debris flows, depending on whether their dominant fraction is fine or coarse (Varnes, 1978). As the definition of mud flows and debris flows often applies to mass transport along streambeds during high-water episodes, these phenomena may be specified as soil slips-mud flows and soil slips-debris flows, depending on their mechanism of initiation (Campbell, 1975, Ellen and Fleming, 1987). In both cases, these translational slides are extremely fast and triggered by very intense rainfall, when slopes reach critical rupture conditions and abruptly collapse.
These movements produce a huge amount of sediment, as well as linear erosion along ephemeral streams and first-order stream channels. As a result, measures should be taken to mitigate slope movements and minimise the erosional capability of channelled waters. Soil bioengineering may address both of these requirements, with functional and ecological advantages: live plant materials have a high anti-erosion potential, increase the shear strength of soil and are environment-friendly in the most diverse settings.
Gravitational movements and precipitation events
Intense and protracted precipitation often causes slope movements. The types and kinematics of rainfall-triggered landslides are different, depending on climate and geo-mechanical properties of soils in the sites involved. Climate factors, e.g. rainfall and temperature, affect the hydraulic and mechanical properties of outcropping soils, as well as the water budget of the local hydrographic network, with an impact on subsurface water filtration along slopes. Changes in these conditions in the short-medium term may activate shallow to averagely deep landslides.
From the viewpoint of the rainfall pattern in a given area, scenarios that typically trigger slope instabilities may be summarised as follows (Clarizia et al., 1996):
1)short spells of rain, i.e. very intense bursts preceded by significant cumulated rainfall;
2)short spells of rain, i.e. very intense bursts preceded by moderate cumulated rainfall;
3)hydrological years with high values of total rainfall, preceded by significant rainfall cumulated over multiple years;
4)hydrological years with high values of total rainfall, preceded by moderate rainfall cumulated over multiple years.
On the first assumption, diffuse shallow slope instabilities are very likely. Indeed, one of the most common and, at times, tragic consequences of extreme weather events is the activation of numerous shallow landslides (almost all of small scale) in eluvial-colluvial covers (De Luca et al., 1996; Johnson and Sitar, 1990). On the second assumption, shallow slope instabilities are usually isolated. In the latter two cases, deep to averagely deep slope instabilities may take place (Terranova and Gulla, 2002).
Soil lithology, dip and strike often make slopes prone to instability. Sometimes, extreme weather events are not the only triggers of slope instabilities but add to anthropogenic factors. All too often, lack of knowledge of geological and geomorphological issues has caused the evolution of precarious slope stabilities into overt slope instabilities. Identifying the geomorphological features of landslide bodies and interpreting them in terms of shallow deformations are very useful elements to gain insight into their types and possible evolution in space and time.
Based on the volume of their mobilised material, their kinematics and hazard level, instabilities may be distinguished into three categories, each with distinctive features: large-scale landslides, averagely deep instabilities and instabilities of regolith covers. The following is a description of the mechanisms of evolution of each category and of their causal factors, especially for shallow slope instabilities, whose unique features make them particularly insidious and hazardous in economic and social terms.
The geomorphology of slopes affected by deep landslides, known in the literature as Deep-Seated Gravitational Slope Deformations, has been investigated for decades. Their deformation mechanisms reflect the stratigraphic-structural and tectonic conditions favouring their onset, even on otherwise stable slopes. Thorough debates have led to define these landslides as the deformation of a mass not necessarily implying a well-defined deformation surface (Dramis and Sorriso-Valvo, 1994). Scale factors, which induce changes in the mechanical properties of the material, may have an impact on the deformation mechanism (Sorriso-Valvo and Tansi, 1996). Particular litho-stratigraphic and geological-structural settings, i.e. competent rock bodies overlying plastic lithotypes, are certainly predisposing factors for the evolution of landslides of this type, while the triggering factors are to be sought in the geodynamic setting, namely in the seismic stresses to which slopes may be exposed (Monteleone et al., 2010).
The evolution of deep-seated gravitational slope deformations is usually very slow, although forms of brittle slope rupture, e.g. scarps and trenches, observed in different settings are very likely to swiftly accelerate deformation rates.
Averagely deep instabilities
These instabilities mostly include complex landslides, such as soil slips, mud and debris flows. In this case, slopes frequently have niches, scarps and counterslope areas that are typically associated with rotational slides and more or less elongated mud and debris flows. Length, width and thickness of these landslide bodies are closely related to the morphometric characteristics of the slopes.
In particular, rotational slides have a curvilinear rupture surface, along which an unstable mass with homogeneous rheological properties rotates; the force generated by the foot of the landslide moving upslope counterbalances the one of the mobilised mass, naturally halting its movement and thus stabilising the slope.
Translational slides may instead occur in lithotypes belonging to a rhythmic succession of soils with different physico-mechanical properties. Their movement follows existing discontinuity planes. Subvertical fracture systems play a crucial role in this type of instabilities, as they represent predisposing factors for the collapse of more or less conspicuous masses.
These instabilities are typical of the “Langhe” area of Piedmont, where they involve late Miocene formations made up of alternating marly and arenaceous-sandy lithotypes; in these cases, their maximum thickness almost never exceeds 6 m, while their frequency is high (Forlati and Campus, 1996).
Another example is the Vajont landslide, the paradigm of an anthropogenic disaster. In October 1963, a mass of over 260 million m3 of rocks and debris detached from the northern slope of Mount Toc and moved at high speed along a paleo-surface, plunging into the artificial dam built along the homonymous stream (Semenza and Ghirotti, 1998). Although the ancient landslide body had been identified by geomorphological studies, the experts of the time failed to realise the true nature and level of hazard of what they were investigating, owing to the geological complexity of the area and the unique features of the valley. The Vajont disaster, which is still the focus of scientific debate and reflections, epitomises that understanding nature and its mechanisms of evolution is the main goal of those working in the field of geological sciences applied to land and the environment.
Instabilities of regolith covers
These phenomena consist in the sliding of soils making up the eluvial-colluvial covers of the bedrock, which generally remains stable. They usually originate from heavy rainfall and propagate along streams of mountains and/or hills, often causing major damage to human or industrial settlements located near alluvial fans. Some of the features of these instabilities are as follows: i) they have a very fast evolution; ii) they are almost always related to intense precipitation; iii) they are frequently associated with high solid transport along streams owing to strongly concentrated rainfall, small size of hydrographic basins, steep slopes and availability of mobilisable material in streambeds and along slopes.
Fires, intense erosion and strongly concentrated rainfall enhance shallow landslide hazards. Continuous urbanisation and construction of roads in mountainous and hilly areas make land and the environment increasingly vulnerable. Inevitably, the combination of these two trends progressively increases the geo-environmental risks associated with these slope instabilities (Fig. 1). The main predisposing factor is the occurrence of weathered and/or reworked covers, resting on more or less undisturbed bedrock and with sharply different rheological properties. In this regard, the nature of the original rock appears to be poorly significant (Campbell, 1975, Govi and Sorzana, 1980). Another equally important causal factor is the sharp differentiation between the cover and the bedrock in terms of permeability and shear strength (Clarizia et al., 1996). The thickness of the layers involved depends on their hydraulic conductivity vs. rainfall intensity. Generally, shallow layers-more permeable than underlying ones owing to roots and cavities due to various forms of biological activity-reach temporary saturation conditions upon very intense rainfall.
Slopes reach critical rupture conditions when, as a consequence of particularly heavy rainfall, the shallower layers of their covers become saturated. Therefore, neutral positive pressures arise in areas where permeability vertically changes, e.g. at the contact between the cover and the bedrock or near clayey interlayers. This significantly reduces the stability of the slope, which may reach critical rupture conditions (Govi and Sorzana, 1980, Cancelli and Nova, 1985, Bellini and Martini, 1997). These instability conditions are reached almost simultaneously in areas whose covers have homogeneous permeability and thickness and where rainfall has similar trends in terms of intensity and duration.
The surfaces involved by the individual phenomena range from few m2 to some hundreds of m2. Consequently, the solid volumes mobilised by each phenomenon are small, because the depth of the glide surfaces is generally below 1 m. Conversely, the concentration of the phenomena per unit of surface area ranges from some landslides per km2 to more than 100 landslides per km2, depending on rainfall characteristics.
As to the average gradient of the slopes most susceptible to shallow landslides, it varies from 16° to 45°, more frequently from 25° to 35° (Govi et al., 1985, Johnson and Sitar, 1990, Polloni et al., 1996), whereas the detachment niche often lies near natural or anthropogenic slope ruptures, as well as small depressions.
If the slopes are very steep and their materials have a high water content, the velocities of these movements are estimated at several metres per second (Hutchinson, 1988); tests on physical models conducted in Japan suggest velocities of up to 15 m s− 1 on 35° slopes (Fukuoka and Yoshida, 1977).
Contribution of soil bioengineering to slope stabilization
Evolutionary models of landslide-prone slopes are of paramount importance: i) to assess the level of hazard and estimate the risk in the investigated area; ii) to plan the most appropriate remediation measures; and iii) to better define the use of the investigated area upon land planning. Soil bioengineering may provide a major contribution in the most diverse settings, from consolidation of unstable slopes to restoration and rehabilitation of sites of high hydrogeological instability, by using live plant materials, possibly in combination with timber, rock, geotextiles or metal structures.
Vegetation can drain the soil by consuming large amounts of water via evapotranspiration: if the soil is less wet, the internal friction of its particles becomes higher, increasing slope stability. Moreover, plant roots can hold and reinforce the soil.
Storms bring about not only gravitational movements, but also intense and concentrated erosion (mostly linear) along streams and slopes denudated of their covers by gravitational movements; as is obvious, this process causes an over-accumulation of sediment onto valley floors (Fig. 2).
Any restoration and consolidation measure based on soil bioengineering techniques should thus consider this additional factor of hydrogeological instability, by discriminating the stream zone from the one of the landslide crown and the zone of bare rock from the one of deposition of debris. Each of these zones calls for suitable stabilisation systems, such as rock and timber retaining walls or gabions, fascines, vegetative or grass covers or live crib walls.
However, before repairing landslides, slopes should be reprofiled by excavation and deposition of the resulting material onto other areas, so as to optimise their morphology; unquestionably, the controlling parameter for these measures is the final slope gradient, which is turn related to the geotechnical properties of the soil.
Among the various solutions offered by soil bioengineering, revegetation is certainly the one that best protects the soil from surface erosion due to battering rain and surface runoff, although a non-negligible role in soil stabilisation is also played by the consolidating action of plant root systems and by the drainage effect due to water absorption and transpiration.
On slopes prone to shallow landslides, live crib walls have not only a high environmental sustainability but provide better long-term stability than dry timber crib walls; moreover, the longer the cuttings planted into the soil, the higher the deep consolidating effect. Geosynthetics enhance the stabilising action of plant materials and can be used for various purposes, e.g. soil armouring, drainage, protection from erosion and support for initial plant growth.
Additionally, surface draining systems may capture surface runoff waters and channel them, along well-defined or preferential pathways, towards natural discharge areas. Deep drainage techniques (draining trenches, microslotted subhorizontal pipes, filtering wedges etc.) may lower groundwater elevation and contribute to mitigating and/or removing major destabilising factors on slopes susceptible to averagely deep landslides.
The assessment of geo-environmental hazards has aroused growing interest in the past few decades in response to two needs: i) ensuring an acceptable level of risk to both new and existing human and industrial settlements; and ii) protecting sites of social, artistic, cultural and economic value from catastrophic events. It is worth pointing out that any human action tends to change the rhythm and sequence of natural events in the surrounding environment and that these changes are more and more frequently irreversible. Hence, it is imperative to plan land uses on the basis of geological and morphological considerations rather than administrative ones. Furthermore, local communities should be made aware of geo-environmental hazards, as indifference to these issues causes, more and more often, not only considerable damage to property and infrastructure but also losses of human lives.
Given their high frequency and rapid onset, landslides involving eluvial-colluvial covers are among the instabilities that may directly or indirectly result into significant geo-environmental issues. Their level of hazard is related less to the volume of soil mobilised by each phenomenon than to the fact that, under particular circumstances, they may rapidly evolve into mud or debris flows channelling into stream channels at high speed.
The effects of the catastrophic flood on the Ionian side of the Messina area in October 2009 emphasises the need for focusing on these instabilities when developing soil defence and land-use plans. Indeed, their unique features make them particularly insidious and hazardous in economic and social terms. At present, although their time of onset is fairly predictable, critical rainfall thresholds for triggering them are still hardly definable. Thus, although weather alerts are systematically launched every autumn, the overall socio-economic toll of these disasters is increasingly high, owing to the lack of adequate prevention and land planning and management policies. Hence, the need arises for putting in place effective programmes of reforestation/revegetation, by leveraging the potential of soil bioengineering techniques.
Source: Hydrogeological hazards and weather events: Triggering and evolution of shallow landslides
Authors: Salvatore Monteleone, Maria Sabatino