P. Mazzanti and F.V. De Blasio
Abstract The morphology of subaqueous landslide deposits is seldom analyzed quantitatively or at least semi-quantitatively with regard to the dynamics of the flow.
However, the peculiar morphology of the deposits can reveal information on the mechanics of propagation of the mass movement and on the mechanism of emplacement.
Horseshoe-shaped deposits and oriented blocks are two peculiar morphologies of subaqueous landslide which have been identified in this paper. Both morphologies have been described in a qualitative way by considering the shape of the deposit and the geomorphological features of the areas of emplacement. Furthermore, these morphologies have been analyzed in a quantitative way by means of mathematical and numerical approaches. The close relationship between these morphologies and the landslide mechanism of propagation and emplacement has been thus preliminary demonstrated by accounting for the interaction between moving mass and the ambient fluid.
Keywords Subaqueous landslides • landslide dynamics • landslide deposits • horseshoe-shaped deposits • oriented blocks
The dynamics of subaqueous landslides is still poorly understood. One of the most remarkable features is the low frictional resistance met by subaqueous landslides, which implies that water has the effect of enhancing the flow (De Blasio et al. 2006b).
P. Mazzanti ( )
Dipartimento di Scienze della Terra, Università di Roma “Sapienza”,
P.le Aldo Moro 5, 00185, Roma, Italy e-mail: paolo.mazzanti@uniroma1.it
F.V. De Blasio ( )
Department of Geosciences, University of Oslo/ P.O. Box 1047 Blindern, N-0316 Oslo,
Norway & International Centre of Geohazards, P.O. Box 3930 Ulleval Stadion,
N-0806 Oslo, Norway e-mail: f.v.d.blasio@geo.uio.no
D.C. Mosher et al. (eds.), Submarine Mass Movements and Their Consequences,
Advances in Natural and Technological Hazards Research, Vol 28,
© Springer Science + Business Media B.V. 2010
141

142 P. Mazzanti and F.V. De Blasio
Mechanisms like hydroplaning (Mohrig et al. 1998), pore pressure augmentation
(Gee et al. 1999) or sediment dilution at the base (De Blasio et al. 2005) have been suggested as mechanisms for the increase in the mobility, but a deep understanding of the fluid dynamics of subaqueous landslides is still lacking. Laboratory experiments may assist in the development of ideas (Mohrig et al. 1998; Ilstad et al. 2004a, b;
Breien et al. 2007), but they bring to conclusions not readily applicable to real events because of scaling problems.
An alternative approach in the understanding of subaqueous landslides dynamics is the analysis of the morphology of the final deposits to extract information about the mechanism of propagation and emplacement (see Mazzanti
2008; De Blasio and Elverhøi 2009). In this paper we examine some peculiar landslide deposits which can be recognized in open, closed or semi-closed basins, like lakes or fjords. In particular, we consider: (1) horseshoe-shaped deposits, and (2) oriented blocks. After a brief geomorphological description, we attempt a correlation of these peculiar morphologies to the landslide dynamics of emplacement.
Horseshoe-shaped deposits are a typical morphology observed in certain types of subaqueous and coastal landslides. These morphologies appear like fairly regular semi-circular rims higher than the rear part. Deposits characterized by such a morphology have been recognized in several subaqueous landslides in open, closed or semi-closed basins. In Fig. 1 we show some examples of horseshoe-shaped deposits at Lake Tahoe (Gardner et al. 2000), Norwegian Fjords (Blikra et al. 2006) and
Lake Albano (Mazzanti et al. 2007).
Even if some frontal and lateral ridges have also been recognized in subaerial rock-avalanches, it is only in subaqueous landslide deposits that they appear to be very regular and perfectly semi-circular. This evidence suggests a direct and strong relationship between morphology of landslide deposits, landslide propagation and the ambient fluid. In addition to surrounding water, a primary role seems to be played by the landslide path; these morphologies, in fact, are typical of mass movements in areas characterized by a sudden change in the slope gradient (i.e. from a relatively steep slope to an extremely flat area).
Furthermore, these deposits are characterized by a very short runout (distance between the source area and the final part of the deposit) if compared to usual subaerial (Legros 2002) and subaqueous landslides (Gardner et al. 2000; Blikra et al. 2006; Mazzanti et al. 2007). Such morphologies are possibly the result of the landslide mechanism of emplacement which is strongly affected by the landslide dynamics.

Peculiar Morphologies of Subaqueous Landslide Deposits 143
Fig. 1 Horseshoe-shaped deposits at: (a–c) Albano Lake (Italy) (Mazzanti et al. 2007); (b)
Tafjord (Norway) (Blikra et al. 2006); (d) Lake Tahoe (USA) (Gardner et al. 2000). Horseshoeshaped deposits are enclosed in dotted white ellipses
The main differences in the dynamics of subaerial and subaqueous landslides are related to some physical processes like: the buoyancy effect, the drag and lift forces exerted on the moving body and peculiar mechanisms like hydroplaning (Mohrig et al. 1998; De Blasio et al. 2004). A body moving in a fluid, like a landslide in water or an airplane in air, is affected by both drag and lift forces. The drag force works parallel and in opposition to the direction of movement and can be divided into two components: the front drag (also called form drag, which acts on the frontal area of the body, and, thus, is strongly dependent on the object form) and the skin drag
(which mainly depends on the viscous friction between the fluid and the upper and lower surface of the moving mass). Instead, the lift force is perpendicular to the direction of travel or, in other words, normal to the drag force and to the flow direction.

144 P. Mazzanti and F.V. De Blasio
The importance of the lift force in subaqueous landslides is demonstrated by the phenomenon of hydroplaning (Mohrig et al. 1998; De Blasio et al. 2004). The lift force D (as well as the drag force) is dependent on the square of the mass velocity
U (De Blasio et al. 2006a)
D =
1
2 r
W
C U A (1) where r w
is the density of water, U is the velocity of the mass, A is the reference area of the moving mass (corresponding to the projection area of the object on a plane perpendicular to the direction of motion) and C
L
is the dimensionless lift coefficient. At high values of velocity, the lift force could become greater than the pressure load exerted by the mass (De Blasio et al. 2006a) and thus the frontal part of the mass is lifted up.
Laboratory experiments performed by Ilstad et al. (2004a) show that the lifting up of the frontal part of the gravity flow under particular conditions can be drastic enough to result in a “flip-back”. This phenomenon consists of a significant rising of the frontal part of the flow which is then flipped back by the front drag force
(significantly increased at that point due to the thickening of the mass head). Such an increase of the front drag force also results in a sudden reduction of velocity.
The mechanisms discussed above occur mainly in the case of cohesive flows (low sand/clay ratios) in the experiments by Ilstad et al. (2004b). We suggest here that the flip-back phenomenon could explain the horseshoe-shaped deposits discussed above. Assuming that all the frontal part of the landslide is affected by the flip-back mechanism, the result would be a frontal ridge with a semi-circular shape (the mass in a non confined and flat area moves radially from the source with a maximum velocity in the flow direction and decreasing velocities in the direction of spreading; see Fig. 2). The occurrence of the flip-back could also explain the limited runout of these deposits as a consequence of a drastic increase of the front drag force which is able to quickly stop the mass and therefore freeze the subaqueous landslide.
To estimate whether the front flipping may occur with the flow velocities expected for small subaqueous landslides, we adopt a model for the front as a straight parallelepiped of width W, length L and thickness T. The condition for flipping to occur is that the torque exerted by the drag force be greater than the one due to gravity. This condition leads to a minimum velocity for flipping which is independent of L and W
U
FL
≈ 2
Δ r
W r g T
C
D cos ϑ sin 2 ϑ
(2) where C
D
, J are the front drag coefficient and the angle of inclination of the front with respect to the horizontal. Evidently, an angle J different from zero is necessary to initiate flipping, a condition that must occur frequently considering that the flow of the subaqueous landslide is irregular and the landslide is never perfectly horizontal. With values
Δ r r
W
= 1; L = m ϑ = 45 ; C
D
= 0.5; T = 3 m (3)

Peculiar Morphologies of Subaqueous Landslide Deposits
Fig. 2 Sketch showing the emplacement of the frontal ridge
145 we find a flipping velocity of the order 13 m/s, a value that looks reachable also according to numerical simulations of similar events (Mazzanti 2008). The uncertainties in these estimates are probably not crucial for the final conclusions. Values of the drag coefficient might be greater than that used here, but probably not >1.
It is also interesting to estimate the deceleration to which the landslide is subjected when the front is raised from an angle J = 0 to the maximum level J = 90°. If the flipping front is detached from the rest of the body, the deceleration is of the order of
1 r
2 r solid a
(
− 1 − T − 1
)
(4) where a accounts for the added mass (in this rough calculation we take this constant equal to one) and with U = 13 m/s it is found that | D a| ≈ 6.7 m s −2 , nearly as high as the gravity acceleration. Such a high value also implies a prompt stoppage of the landslide front. If the front is coupled to the rest of the landslide body, the mass that enters the calculation is higher because it contains a part of the landslide mass.
The resulting deceleration is consequently smaller. Nevertheless, these estimates show that the raising of the front likely leads to a substantial braking of the landslide.
The body will continue to move inertially, underriding the front that topples backwards.
Because the front of the landslide is semi-circular, the toppling material will have a horseshoe-shaped morphology (Fig. 2).
An alternative explanation for the horseshoe-shaped morphology is suggested by
Hampton et al. (1996) and confirmed by Schnellmann et al. (2005). Several seismic profiles collected in Lake Lucerne (Switzerland) allowed them to recognize several

146 P. Mazzanti and F.V. De Blasio fold-and-thrust belt structures in soft sediments, induced by the superimposition of subaqueous landslides resulting in a bulge in the frontal part of the deposit.
We suggest here that such mechanisms should produce, if observed in plain view, a horseshoe-shaped morphology around the main landslide deposit. In this case, in contrast to the flip-back hypothesis, the frontal semi-circular ridge results from the deformation of lacustrine sediments. Therefore, it is a consequence of the emplacement of the landslide mass, and not of the intrinsic dynamics of the landslide.
However, in both explanations, the rate and mode of emplacement of the landslide can be considered the main reason for the development of frontal ridges. Further seismic reflection investigations in ridge zones will probably shed light, by looking at the sub-bottom geometry of the deposit, on the most probable model.
Isolated blocks are not uncommon in subaqueous landslide areas. Whereas some outrunner blocks may travel alone far away from the main accumulation of underwater landslides (Longva et al. 2003; Prior et al. 1982a, b, 1984; Kuijpers et al. 2001;
Nielsen and Kuijpers 2004), others may be transported as rafted blocks by the flowing landslide, probably transformed into a debris flow (Lastras et al. 2005; De Blasio and
Elverhøi 2009). Rafted blocks are typically much smaller than the original landslide, with maximum recorded volumes of 10 6 m 3 (Vanneste et al. 2009) and exhibit an extraordinary mobility, travelling for kilometres along gradients as low as 1° or even less. Blocks show in some cases a very low aspect ratio (De Blasio et al. 2006a).
An interesting characteristic of some subaqueous blocks is the orientation of the main axis with respect to the flow direction. Most outrunner blocks usually have the main axis perpendicular to the flow direction (De Blasio et al. 2006a), which is consistent with the fluid mechanics of isolated bodies moving in a fluid (Batchelor
1967). The same conclusion holds for blocks rafted by mass flows, the main difference being that for rafted blocks the interaction force with the mass flow is more important than that with water. Thus, rafted blocks are oriented by the action of the viscous mass flow rather than by water. Axial orientation parallel to the flow is expected only for the Stokes regime of very low Reynolds numbers (Rhodes 1998), a condition hardly met for large-scale blocks. However, as shown in Fig. 3, the outrunner blocks at Lake Albano (Mazzanti et al. 2007) show a higher value of the height/length ratio and, in contrast to the cases described by De Blasio et al.
(2006a) the maximum axis is sub-parallel to the direction of flow. As an example, at Lake Albano we can recognize blocks oriented with the main axis in the direction of flow having a planar dimension of nearly 30 × 10 m, a maximum height of about
3–4 m and which have covered distances up to 280 m in a very flat area (Fig. 3;
Mazzanti et al. 2007).
In the following, a possible mechanism for block orientation is investigated.
This mechanism can also shed light on the dynamics of subaqueous mass flows.
We suggest that a velocity gradient within the mass flow may turn the blocks with

Peculiar Morphologies of Subaqueous Landslide Deposits 147
Fig. 3 Prospective view with 5 m contour lines of a completely subaqueous debris flow at Lake
Albano (Mazzanti et al. 2007). Red dashed line bounds the landslide scar; yellow dashed lines identify the landslide erosive channel; green dashed line bounds the main landslide deposit; blue full ellipse encloses an oriented block. Distance between the front and the background is nearly 650 m u
0
+ (
¶ u /
¶ y)y u
0 a
L
J
Fig. 4 Scheme for the calculation of the torque acting on the block axis nearly parallel to the flow direction. Let us assume a very simple model of a block as a cylindrical ellipsoid of semi-axes a, L, and thickness W. The block is immersed in a fluid with a velocity gradient as shown in Fig. 4. Because of the velocity gradient, the block is subjected to a torque that we approximate as:
M ≈
1
2 r
/ 2
W
[
( ) cos
−
L
L / 2
ϑ ]
2
(5)

148 P. Mazzanti and F.V. De Blasio where J is the angle between the block axis and the velocity current and W is the block thickness (perpendicular to the plane of Fig. 4). Assuming a constant velocity gradient u(y) = u
0
+ ( ∂ u/ ∂ y)y, it is readily found that:
M ≈ −
1
12
W u
0 r ( ∂ ∂ y L 3 ϑ
(6)
In the absence of a velocity gradient, a torque tends to line up the block with axis perpendicular to flow (Batchelor 1967):
T ≈
1
4
W r u L 2
⎝
⎜
⎛
1 − a 2
L 2
⎟
⎞
⎠ sin cos ϑ (7)
We assume that the two contributions can be linearly summed:
N = M T
1
12
W u
0 r ( u / y L 3 ϑ +
1
4 r 2
⎜
⎛
⎝
1 − a
L 2
2
⎟
⎞
⎠ sin cos ϑ (8)
Thus, in this approximation the block has a tendency to line up with the axis parallel to the flow if: u
0
∂
∂ y
>
3 cos sin 2 ϑ
ϑ
⎝
⎜
⎛
1 − a
L 2
2
⎟
⎞
⎠
(9) where K ª 1 in the present approximations. As a result, parallel orientation is possible if the velocity gradient is sufficiently large in relation to the block length and the average landslide velocity. Note also that the factor sin 2 J in the denominator of
Eq. 9 implies that the block must have a minimum obliquity with respect to the velocity vector to start rotation. Assuming a length of 20 m, a velocity of 20 m/s, J
= 45°, and a = 5 m, we find that a gradient of about 4 s −1 is necessary for the block to line parallel to the flow, but with J = 80° only 0.5 s −1 is necessary. The highest gradients are obtained at the sides of the mass flow, where the velocity abruptly goes to zero. We thus predict that parallel blocks could be more common at the sides of the mass flow, whereas in the central part they will mostly flow with the main axis perpendicular to the flow. To illustrate this phenomenon, we performed some simple preliminary experiments with a viscous fluid simulating an incoming mass flow and some pieces of rubber standing for blocks (Fig. 5). We observed that irrespective of the original orientation of the rubber blocks, they tend to become parallel to flow at the sides of the fluid, and maintain a perpendicular orientation in the middle of the flow (characterized by small velocity gradient). We noticed however that as predicted by Eq. 9, blocks acquired a rotation with difficulty when they are perfectly parallel to the mass flow velocity.
Figure 6 reports some calculations of a theoretical subaqueous landslide performed by DAN3D model (Mc Dougall and Hungr 2004). The lines of constant velocity illustrate the presence of a strong gradient at the sides of the flow. According to our model, a block at the sides of the mass flow will rotate as shown in the figure.
Note that in this preliminary work the movement of the block is not calculated, but only superimposed on the calculated curves for illustrative purposes. It would be interesting to include the presence of blocks in a three-dimensional model.

Peculiar Morphologies of Subaqueous Landslide Deposits 149
Fig. 5 Sketch obtained from a series of preliminary experiments where a viscous fluid flowing down a flat inclined plane (from left to right in the figure) strikes against some pieces of rubber and plastic material of different aspect ratios. The initial configuration is chosen with blocks perpendicular, parallel to the fluid velocity, or with random orientation. Irrespective of the initial orientation, after the passage of the fluid the blocks tend to lineate with flow only at the sides. The length of the experimental board is approximately 1 m
Fig. 6 Sketch showing the orientation mechanism of a block parallel to the landslide flow direction

150
P. Mazzanti and F.V. De Blasio
To gain insights into landslide flow dynamics in the subaqueous environment, we have analysed some peculiar morphologies of subaqueous landslide deposits. We focused on two morphologies which can be recognized in open, closed or semi-closed basins: (1) horseshoe-shaped deposits and (2) oriented blocks. These features are typical of certain types of subaqueous landslides. Hence, we suggest that their origin is closely related to the interaction between the water and the moving mass during the landslide propagation and emplacement. It has been shown that horseshoe-shaped deposits may be the results of a generalized flipping back of the landslide front during propagation and that the rising of the frontal part of the landslide can explain also the short runout of such landslides. Once validated and calibrated, the proposed model may allow us to infer the landslide velocity based on the morphology of its deposit. This is a relevant issue, because it represents a key requirement in the assessment of tsunami potential of coastal landslides. Furthermore, we suggest that the occurrence of the flip-back could significantly increase the tsunami potential since the sensitivity of tsunami-wave generation to the landslide front thickness has been widely demonstrated (Harbitz et al. 2006). We have also considered the problem of the enigmatic orientation of rafted blocks parallel to the axis of a subaqueous landslide. We show that the orientation parallel to flow may occur in zones of high velocity gradient, namely at the sides of the mass flow deposit.
This paper represents a preliminary step in the understanding of subaqueous landslides dynamics based on peculiar morphological features of their deposits. However, further investigations must be carried out in order to better understand and constrain such phenomena. Once constrained, these hypotheses could allow for an estimation of past landslides’ velocity of emplacement. Then, statistically based methods (based on the analysis of past events) could be designed that would allow us to perform forecasting analysis of future landslide velocity of propagation and emplacement.
Acknowledgments The authors wish to thank Prof. F. Bozzano and Prof. F.L. Chiocci for enjoyable discussions about the main topics of this paper and for financial support. The manuscript benefited from revisions by Dr. A. Cattaneo and Dr. M. Owen.
Batchelor G K (1967) An Introduction to Fluid Dynamics. Cambridge University Press, New York.
Blikra L H, Longva O, Braathen A, Anda E, Dehls K, Stalsberg K (2006) Rock slope failures in
Norwegian fjord areas: examples, spatial distribution and temporal pattern. In: Evans S G,
Scarascia Mugnozza G, Strom A, Hermanns R (eds) Massive Rock Slope Failure, NATO
Science Series, Kluwer Academic, Springer Netherlands, pp. 475–496.
Breien H, Pagliardi M, De Blasio F V, Issler D, Elverhøi A (2007) Experimental studies of subaqueous vs. subaerial debris flows – velocity characteristics as a function of the ambient fluid. In: Lykousis V, Sakellariou D, Locat J (eds) Submarine Mass Movement and Their
Consequence, Springer Heidelberg, pp. 101–110.
De Blasio F V, Elverhøi A (2009) Properties of mass-transport deposits as inferred from dynamical modelling of sub-aqueous mass wasting: a short review. In Shipp C, Weimer P, Posamentier H
(eds) The Importance of Mass-Transport Deposits in Deepwater Settings. Soc Sediment Geol
Spec Pub (In press).

Peculiar Morphologies of Subaqueous Landslide Deposits 151
De Blasio F V, Elverhøi A, Engvik L E, Issler D, Gauer P, Harbitz C (2006b) Understanding the high mobility of subaqueous debris flows. Nor J Geol 86: 275–284.
De Blasio F V, Engvik L, Harbitz C B, Elverhøi A (2004) Hydroplaning and submarine debris flows. J Geophys Res 109: C01002, doi:10.1029/2002JC001714.
De Blasio F V, Engvik L, Elverhøi A (2006a) Sliding of outrunner blocks from submarine landslides.
Geophys Res Lett 33: L06614, doi:10.1029/2005GL025165.
De Blasio F V, Elverhøi A, Issler D, Harbitz C B, Bryn P, Lien R (2005) On the dynamics of subaqueous clay rich gravity mass flow – the giant Storegga slide, Norway. Mar Pet Geol 22: 179–186.
Gardner J V, Mayer L A, Hughs Clarke J E (2000) Morphology and processes in Lake Tahoe
(California-Nevada). Geol Soc Am Bull 112 (5): 736–746.
Gee M J R, Masson D G, Watts A B, Allen P A (1999) The Saharan flow: an insight into the mechanics of long runout submarine debris flows. Sedimentol 46: 317–335.
Hampton M A, Lee H J, Locat J (1996) Submarine landslides. Rev Geophys 34: 33–59.
Harbitz C B, Løvholt F, Pedersen G, Masson D G (2006) Mechanisms of tsunami generation by submarine landslides: a short review. Nor J Geol 86: 255–264.
Ilstad T, De Blasio F V, Elverhoi A, Harbitz C B, Engvik L, Longva O, Marr J G (2004a) On the frontal dynamics and morphology of submarine debris flows. Mar Geol 213: 481–497.
Ilstad T, Elverhøi A, Issler D, Marr J (2004b) Subaqueous debris flow behaviour and its dependence on the sand/clay ratio: a laboratory study using particle tracking. Mar Geol 213: 415–438.
Kuijpers A, Nielsen T, Akhmetzhanov A, De Haas H, Kenyon N H, van Weering T C E. (2001)
Late Quaternary slope instability on the Faeroe margin: mass flow features and timing of events. Geo-Mar Lett 20: 149–159.
Lastras G, De Blasio F V, Canals M, Elverhøi A (2005) Conceptual and numerical modeling of the Big’95 debris flow, western Mediterranean Sea. J Sed Res 75: 784–797.
Legros F (2002) The mobility of long-runout landslides. Eng Geol 63: 301–331.
Longva O, Janbu N, Blikra L H, Bøe R (2003) The 1996 Finneidfjord slide; seafloor failure and slide dynamics. In: Locat J, Mienert J, Boisvert L (eds) Submarine Mass Movements and Their
Consequences, 1st International Symposium, Kluwer Academic Publishers, Dordrecht, pp. 531–538.
Mazzanti P (2008) Analysis and modelling of coastal landslides and induced tsunamis. PhD
Thesis Università di Roma “Sapienza”, Dipartimento di Scienze della Terra.
Mazzanti P, Bozzano F, Esposito C (2007) Submerged Landslides Morphologies in the Albano
Lake (Rome, Italy). In: Proceedings of 3rd International Symposium “Submarine Mass
Movements and Their Consequences”, Series: Advances in Natural and Technological Hazards
Research, 27, Springer, Heidelberg, pp. 243–250.
McDougall S and Hungr O (2004) A model for the analysis of rapid landslide motion across threedimensional terrain. Can Geotech J 41: 1084–1097.
Mohrig D K, Whipple C E, Parker G (1998) Hydroplaning of subaqueous debris flows. Geol Soc
Am Bull. 110: 387–394.
Nielsen T, Kuijpers A (2004) Geohazard studies offshore the Faroe Islands: Slope instability, bottom currents and sub-seabed sediment mobilisation. Geol Soc Den Greenland Bull 4: 57–60.
Prior D B, Borhold B D, Coleman J M, Bryant W R (1982b) Morphology of a submarine slide,
Kitimat Arm, British Columbia. Geology 10: 588–592.
Prior D B, Borhold B D, Johns M W (1984) Depositional characteristics of a submarine debris flow. J Geol 92: 707–727.
Prior D B, Coleman J M, Borhold B D (1982a) Results of a known sea floor instability event. Geo
Mar Lett 2: 117–122.
Rhodes M (19989 Introduction to Particle Technology. Wiley, Chichester, 320 pp.
Schnellmann M, Anselmetti F S, Giardini D, McKenzie J A (2005) Mass movement-induced fold-andthrust belt structures in unconsolidated sediments in Lake Lucerne (Switzerland). Sedimentol
52: 271–289.
Vanneste M, Harbitz C B, De Blasio F V, Glimsdal S, Mienert J, Elverhøi A (2009) The Hinlopen-
Yermak landslide, Arctic Ocean. Geomorphology, landslide dynamics and tsunami simulations. In Shipp C, Weimer P, Posamentier H (eds) The Importance of Mass-Transport Deposits in Deepwater Settings. Soc Sediment Geol Spec Pub (In press).