sign in sign up
<<
Home , Geology of Chile

INTERNATIONAL SOCIETY FOR SOIL MECHANICS AND GEOTECHNICAL ENGINEERING

This paper was downloaded from the Online Library of the International Society for Soil Mechanics and Geotechnical Engineering (ISSMGE). The library is available here: https://www.issmge.org/publications/online-library

This is an open-access database that archives thousands of papers published under the Auspices of the ISSMGE and maintained by the Innovation and Development Committee of ISSMGE. Paper No. ANAHE

ANALYSIS OF THE 2007 -INDUCED PUNTA COLA ROCKSLIDE AND TSUNAMI, AYSÉN , PATAGONIA, (45.3º S, 73.0° W)

Tim F. REDFIELD1, Reginald L. HERMANNS1, Thierry OPPIKOFER1, Paul DUHART2, Mauricio MELLA2, Patricio DERCH2, Ignacio BASUÑAN2, Manuel ARENAS3, Javier FERNANDEZ3, Sergio SEPÚLVEDA4, Sofia REBOLLEDO4, Simon LOEW5, Freddy X. YUGSI MOLINA5, Anna ABÄCHERLI5, Iain H.C. HENDERSON1, Michel JABOYEDOFF6, Vidar KVELDSVIK7

ABSTRACT

The 2007 Aysén earthquake directly triggered many large-volume rockslide releases into the fjord, in turn causing tsunami-related destruction and death. One of the rockslide scars, a ~12 million cubic meter release at Punta Cola, was studied in initial detail with Terrestrial Laser Scanning (TLS) data and direct acquisition of field measurements. A Basal Failure Surface (BFS) and an ESE-trending Lateral Release Surface (LRS) were identified. A crush zone sub-parallel to a mapped , characterized by extremely high fracture density and trending approximately N-S, may have destabilized the uppermost portion of the compartment in a manner similar to a ‘back-crack.’ Release occurred towards a ‘free face’ formed by the deeply-incised pre-slide drainage system.

The crystalline bedrock into which Aysén fjord is incised hosts numerous joint swarms, some of which developed into brittle faults. Mylonite bands up to several meters wide also occur throughout the region, and may have helped localize subsequent jointing and brittle faulting. Twenty-four brittle faults were observed in the Punta Cola slide scar, and 53 elsewhere in Aysén fjord. Striated and mineralized fault surfaces carry dip-slip and oblique lineations, exhibiting both normal and reverse sense of shear. The zone affected by faulting and jointing is very broad, and is interpreted to be directly associated with the Liquiñe-Ofqui Fault Zone (LOFZ). The LOFZ is interpreted to be the principal controlling structure with respect to the imposition of destabilizing rock fabrics. Variation along its strike may play a significant role in defining the nature and style of geohazard in Patagonia.

Keywords: Aysén Earthquake Rockslide Patagonia Geohazard Liquiñe-Ofqui

1 International Centre of Geohazards, Geological Survey of Norway, 7491 Trondheim, Norway, e-mail: [email protected] 2 Servicio Nacional de Geología y Mineria, Puerto Varas, Chile, e-mail: [email protected] 3 Servicio Nacional de Geología y Mineria, , Chile, e-mail: [email protected] 4 Departamento de Geología, Universidad de Chile, Santiago, Chile, e-mail: [email protected]. 5 Engineering Geology Group, Federal Institute of Technology ETH Zürich, 8092 Zürich, Switzerland, e- mail: [email protected] 6 Institute of Geomatics and Analysis of Risk, University of Lausanne, 1015 Lausanne, Switzerland, e- mail: [email protected] 7 International Centre for Geohazards, Norwegian Geotechnical Institute, 0855 Oslo, Norway, e-mail: [email protected] 5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile INTRODUCTION

Puerto Aysén is located in southern Chilean Patagonia, just inboard of the innermost mapped trace of the Liquiñe-Ofqui Fault Zone (LOFZ; Figure 1a). On 21 April 2007 the Aysén fjord hosted a Mw = 6.2 shallow crustal earthquake. The surface rupture of the main event is believed to lie on the bottom of the Aysén Fjord some 20 km west of the city (Vargas et al., submitted). The event was the climax of a seismic swarm that started on 22 January 2007.

As many as 538 mass movements of widely-varying composition and dimension were triggered on steep slopes of the fjord and surrounding glacial valleys during the period of time covered by the seismic swarm (Naranjo et al., 2007; Arenas et al., 2008; Naranjo et al., 2009; Sepúlveda et al. 2010). Many failures occurred as catastrophic releases of large volume compartments of rock. Directly following the main earthquake, several large bodies of rock collapsed catastrophically into the fjord, spawning deadly tsunami waves (Naranjo et al., 2009; Sepúlveda & Serey, 2009). One of the largest failures occurred at Punta Cola, where collapse of a compartment containing an estimated 12 million cubic meters of rock occurred very close to the located epicentre of the main seismic event (Figure 1c).

Regional tectonic setting In a series of studies, Hervé, Cembrano, and coworkers defined the LOFZ as the principal structural feature of the Patagonia region (see Hervé & Thiele, 1987, Hervé 1994, Cembrano et al., 1996, 1999; 2000, Cembrano & Lara, 2009). Although enormous temporal gaps remain, fieldwork has documented a lengthy kinematic history dominated by strike-slip offset in both ductile and brittle regimes. Resolution of plate kinematic forces suggests the LOFZ developed as a dextral transpressional structure (Figure 1b; see Somoza, 1998; Cembrano et al., 2007). Cembrano & Lara (2009) concluded that the LOFZ has been active as a transpressional dextral strike–slip structure at least over the last 6 Ma, although circa 25 Ma it was more likely active as a transtensional 'leaky transform'. However, unequivocal piercing points along the LOFZ have not been identified, and paleomagnetic studies have not been able to demonstrate plate-scale lateral motion (García et al., 1988; Rojas et al., 1994).

Thompson (2002) presented Zircon and Apatite Fission Track (ZFT, AFT) data from Patagonia. To explain the relatively rapid exhumation demanded by the AFT age and track length data from the southern end of the LOFZ Thompson (2002) suggested that fault linkage at depth engendered a crustal-scale dextral transpressional flower structure or crustal pop-up. Thompson (2002) suggested that the onset of recent, rapid AFT-indicated denudation correlated with the collision of segments of the spreading center.

That the LOFZ exerts control over the localization of volcanic eruptive centers (Figure 1a) has been previously noted (e.g. Cembrano et al., 1996). Here, we extend the range of LOFZ-related geohazards to include the Aysén fjord rockslides (Figure 1c). We specifically consider the structure of one large volume rockslide at Punta Cola that occurred within the zone of deformation associated with one of the principal LOFZ strands. We obliquely suggest that the regional structuring imposed by LOFZ related deformation may be important over a broader region than simply Punta Cola, or even Aysén fjord. For these purposes, we consider the LOFZ to encompass a complex of fault structures that span much of Patagonia, whose kinematic nature and history may have been quite variable in both space and time.

Local Geology Andean between 40-48°S is dominated by calc-alkaline plutonic rocks of the - North Patagonian Batholith. These plutons intruded 1) a late to early Mesozoic metamorphic accretionary complex at the eastern margin of the Coastal Ranges (Hervé and Fanning, 2001; Hervé et al., 2003); 2) volcanic silicic rocks in the eastern flank of the Main Range (Pankhurst et al., 1999) and the overlying early sedimentary marine and volcanic successions 5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile of the Aysén Basin (Suarez and De La Cruz, 2000); and 3) the current volcanic arc and Pleistocene glacial deposits. The Puerto Aysén region is comprised entirely by NPB plutonic rocks. Although one emplacement age of 98 Ma near Aysén fjord was reported in Thompson (2002), the operating hypothesis is that many intrusive ages and the age of high pressure exhumation are Miocene (see Thompson, 2002). However, radiometric ages describing the rocks near Puerto Aysén are in general very sparse.

Figure 1. a) Map redrawn from Cembrano et al. (1996) and Thompson (2002) showing the Liquiñe- Ofqui Fault System (LOFZ). ANT = plate. NAZ = . SAM = plate. RMF = Rio Mañihuales Fault. ATF = Azul Tigre Fault. White arrows: convergence direction of NAZ (fixed SAM) using rotation poles derived by Somoza (1998). b) Plots of the relative NAZ/SAM convergence velocity and angle (Tangential/Normal) since ~33 Ma. A very high angle of present-day convergence is evident with respect to N-S trending LOFZ structures. Prior to ~25 Ma, a greater tangential component of motion is predicted by the generalized plate kinematic model. c) Location of mass movements induced by the earthquake in the Aysén Fjord (Sepúlveda et al., 2010).

The youngest plutons spatially associated with the central LOFZ are interpreted as syntectonic intrusions (Hervé et al., 1993), although such data from Aysén are lacking. Many exhibit centimeter to meter thick mylonitic bands, indicative of significant shearing during their ductile phase.

5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile

Figure 2. a) Photograph of the Punta Cola rockslide scar with the exposed Basal Failure Surface (BFS) and the Lateral Release Surface (LRS). b, c, d) examples of normal, strike-slip, and reverse brittle faults found in the LRS. e) Reverse fault in tilted glacial varves at the Punta Cola beach.

REGIONAL OBSERVATIONS IN AYSÉN FJORD

During the 2010 field campaign in Aysén, one rockslide, Punta Cola (Figures 1c and 2), was mapped and studied in some detail. Here we present preliminary results and interpretations from the Punta Cola event, supported by generalized observations obtained more regionally throughout Aysén fjord.

Mylonite zones in Aysén fjord vary in width from centimetres to several meters. Very few firm indications of their sense of shear could be documented. In some cases, mylonitic zones were observed to coincide with elevated fracture densities and dikes, and in at least one case, on Mentirosa Island (Figure 1c) a ca. 1 meter wide mylonite zone appears to strike into a large release plane of one of the principle earthquake- triggered and tsunami-genic rockslides. At other outcrops, mylonite fabrics were observed in close association with brittle fracture lenses, also on strike with and directly below large rockslide compartments that released during 2007. Lineations and kinematic indicators for normal, reverse, and oblique to strike slip deformation under brittle conditions were observed. Mineralization that is possibly syn-tectonic includes chlorite, epidote, and calcite. Deformation products include cataclasite, (rare) fault breccias, and many proto-breccia fracture systems. Fault kinematics include normal (11 examples) and reverse (9 examples). Of these 20 fault planes, 8 fault planes with lineations shallower than 45 degrees plunge were observed. Both dextral and sinistral senses of shear were recorded.

5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile

Figure 3. Hillshade of the HR-DEM created from the TLS point clouds of the Punta Cola rockslide scar and the rock avalanche deposits (coordinates in UTM 18S). The inset shows an orthophoto of the epicentral area of the 21 April 2007 earthquake and the major earthquake-triggered rockslides of Punta Cola and Mentirosa Island.

A thrust fault plane cutting unconsolidated clay varves oriented 261°/50° was observed near Mentirosa Island (Figure 2e). beach deposits at Punta Cola dip consistently 12 degrees to the NE. Similar beach deposits elsewhere are flat-lying, indicating both that the supposition of original horizontality is reasonable and that Holocene to Recent strain is differentially partitioned between localities.

THE PUNTA COLA ROCKSLIDE

One of the largest failures occurred at Punta Cola (Figures 1c, 2, 3). A rock mass estimated to comprise 12 million cubic meters failed in a rock avalanche that ran completely down the 1.5 km long valley, impacted the fjord, and contributed to the Aysén tsunami waves. The rock avalanche caused significant erosion in the valley and had a maximum run-up height of 150 m on the opposite valley flank. Five major secondary landslides occurred in the valley of the Punta Cola rock avalanche (Figure 3). Lobate landslide deposits with high soil content overlying fresh blocky rock avalanche deposits suggest that these secondary landslides occurred after the main Punta Cola rock avalanche. They may have been triggered by earthquake aftershocks or by scouring of valley flanks by the main rock avalanche. 5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile The main rock body appears to have failed along a WNW-dipping Basal Failure Surface (BFS) that is exposed in the north-eastern part of the rockslide scar (Figures 2, 3). The scar is nearly 1000 m long, up to 760 m wide, and has a height difference of more than 530 m. The rockslide scar is delimited to the south by an up to 115 m high WNW-ESE-trending sub-vertical Lateral Release Surface (LRS; Figures 2, 3).

Detailed terrestrial laser scanning (TLS) and ground-based photogrammetry data were acquired in Aysén during January 2010 in order to characterize the principal structures involved in earthquake-triggered rockslides. Both techniques permit the creation of high-resolution digital elevation models (HR-DEM) that may be used in detailed structural analyses of the cliffs (Lim et al., 2005; Jaboyedoff et al., 2007; Oppikofer et al., 2009; Sturzenegger and Stead, 2009). Here we present a preliminary map and structural analysis based on the TLS dataset and existing field investigations. Our analysis will be completed by upcoming ground-based photogrammetry results and additional field investigations programmed for 2011.

Structural analysis based on TLS Our TLS analysis of the Punta Cola rockslide scar used Coltop3D software (Jaboyedoff et al., 2007; www.coltop3d.ch). This software computes the spatial orientation (slope direction and slope angle) of each point of a TLS point cloud with respect to its neighborhood and displays the orientation using a unique color code. If the assumption that rock slopes are shaped by major discontinuities is valid, the slope orientations computed in Coltop3D will in large part reflect the orientation of discontinuities.

Based on field observation, we divided the Punta Cola slide into mappable units including what we interpreted to be the BFS and LRS (see Figure 3). Within Coltop3D, selections of points and surfaces with similar orientation were then made to extract a statistical representation of the orientation of the main discontinuity sets present on the BFS (Figure 4a) and the LRS (Figure 4b).

Figure 4. Stereonets of the major structures in the Punta Cola rockslide scar obtained by Coltop3D analysis of the TLS dataset for a) the basal failure surface (BFS); b) the lateral release surface (LRS). Field measurements made on the Punta Cola rockslide scar are shown as poles. Note: density stereonets are based on selections of 1000 points per set. Density stereonets for each discontinuity set show a good match between the mean orientation and the highest density of poles.

5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile Although there are differences between the two domains, the principal sets (D1, D2, D3 and D3') are detected in both zones. Some discontinuity sets can be divided into several subsets in one zone (e.g. D3, D3' and D6 on the BFS; D1, D1' and D1" on the LRS), while they appear to form a single set in the other zone. Note that some discontinuity sets cannot be detected due to occlusions in the TLS point cloud and orientation biases; see Sturzenegger et al. (2007). This situation may occur when the dip direction of a set is more or less perpendicular to the TLS view direction (e.g. D4 and D8 cannot be detected on the BFS), or when they do not actively shape the rock surface (e.g. set D5 is not detectable on the LRS as it dips into the cliff).

Basal failure surface (BFS) As mapped within Coltop3D, the BFS is formed by 7 distinct discontinuity sets. Five sets (D3 to D7) are dipping in W to NW direction and shape the exposed WNW-dipping failure plane (Figure 5). Two sets (D1 and D2) dip to the S to SE. The major part of the BFS appears to be formed by discontinuity sets D3 and D3', shown in green in Figure 5. When present, the shallowly-dipping D7 set (white in Figure 5) is the limiting factor governing slope stability since it is the discontinuity with the smallest dip angle.

Only two field measurements on a slide plane were obtained in 2010. Their orientations correspond to discontinuity set D7. The relatively low dip angle of D7 (30° ± 5°) and the lack of field evidences for low friction angles along the basal sliding surface suggests that a seismic acceleration was necessary to trigger the Punta Cola rockslide. This assumption needs further testing and validation by slope stability models, supported by much more on-site data.

Figure 5. 3D view of the TLS point cloud of the Punta Cola basal failure surface. Points with slope orientation corresponding to one of the discontinuity sets are displayed with the specific colour of the discontinuity (see stereonet). All other points are shown in grey. The analysis is limited to the exposed basal failure surface (excluding surrounding scree deposits and vegetation).

5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile Lateral release surface LRS) The south-bounding cliff of the Punta Cola rockslide comprises the LRS. The LRS is shaped by several steeply dipping to sub-vertical discontinuity sets (Figure 5b). North-dipping and south-dipping discontinuities (D1, D1’, D1”, D4, D6 and D8) define the LRS map shape. Discontinuity sets (D2, D3 and D3’) constitute subvertical NW-dipping extensional structures that may have linked successive lateral release surfaces.

Potential failure mechanism (PFM) Based on the observed structures and assuming a NW-dipping slope face of the pre-rockslide topography, it appears that sets D7 and D8 are the limiting structures for sliding, acting respectively as basal sliding surfaces and lateral sliding surfaces. This simple mechanical analysis also suggests that the main LRS also accommodated some shear stresses to form a wedge sliding mechanism with a wedge intersection line that is slightly oblique (302°N) to the dip direction of D7 (283°).

Field Observations The Punta Cola rockslide was the only location where both lateral release surface and basal failure surface were physically accessed and field measurements were acquired directly. Faults, fractures, dykes, and foliation constituted the principle planar structures. Most structures measured were exposed in the LRS, but measurements were made also in the Punta Cola valley and along the beach. Figure 6 provides a set of summary plots of all data acquired at Punta Cola and more regionally in the Aysén fjord for comparison with the TLS plots of Figure 4.

Joints Much of the Punta Cola failure occurred within gneissic to tonalitic bedrock. High angle fracturing occurs in swarms separated by much less brittly-deformed bedrock. In its unfaulted state, the bedrock tends to be massive: the background fracture density was non-quantitatively characterized as very low. Figure 6 shows joint populations that, given the low number of observations, compare favourably with the TLS- postulated shallow-dipping D7 and steeply-dipping D8 structures.

Faults Eleven normal and 9 reverse brittle faults were observed in the LRS during field inspection. Fault planes are in general only lightly mineralized and kinematic indicators in superlatively short supply. Where possible, kinematic interpretations were made from striated fault planes. However, in many cases, secondary indicators were needed. For example, a very low angle fracture system appeared normally displaced in 2D plane view; elsewhere, thin (ca. 1 mm) bands of calcite displayed crystallization directions perpendicular to the vein fracture walls, suggesting open-space deposition under a tensile regime. Nearby, a set of open fractures document a pattern consistent with normal movement (Figure 2b). Interpretations of reverse movement relied dominantly on fracture patterns, very poor quality fault plane ‘steps’, and in one instance, the displacement of a vein in 2D section (Figure 2d). Fault strikes vary, but the majority lies in the SSW quadrant (dip follows right hand rule). Of the above 20 examples, 8 planes with lineations shallower than 45 degrees plunge were observed. Both dextral and sinistral senses of shear were noted. Poorly-defined slickenlines indicate both dip-slip and oblique components of motion, but no preferred sense of throw is resolvable within our currently very limited data set.

At the BFS, an andesitic dike intruded parallel to a prominent fracture set outcropped near the basal slide plane. Its orientation (dip direction/dip angle: 253°/22°) probably represents an important discontinuity set such as D7 (Figure 4). BFS failure may have occurred along a linked set of similar planes, possibly guided by weakened zones such as those discussed below.

5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile

Figure 6. Stereonets showing poles to joints (purple colors), lineated fault planes (pink colors) and the striations themselves (blue colors) at Punta Cola and in Aysén in general. Stars represent mean striation directions. Steeply-plunging lineations seem to be more common than shallow lineations.

Near faults, fracture density increased significantly. On one notable fault, shear occurred entirely within a basaltic dike carrying olivine phenocrysts (Figure 2c). This fault absorbed virtually all local brittle deformation: the tonalite on either side is completely unfractured.

In the uppermost, easternmost segment of the Punta Cola LRS a very intense zone of brittle deformation was discovered; there, fracture density was elevated to an essentially un-measureable degree. This zone appears to trend N to NNE (Figures 2a, 3), and may have contributed to the pre-slide instability (see below).

DISCUSSION

Rock slope failures triggered by the 2007 Aysén earthquake Field and TLS data suggest that prior to release, compartmentalization structures were plentiful in the Punta Cola rockslide body. Failure is hypothesized to have occurred along low-angle, west-dipping planes, guided laterally by steeper, more northerly dipping planes of the LRS (Figures 4, 5). The intensely-fractured zone exposed in the upper part of the LRS can be followed on the morphology by a flatter area above the exposed BFS (Figure 3). Within this zone, rock strength is probably very much reduced. A direct relationship to failure cannot be drawn with the data at hand, but it is plausible that this zone was an important component of block compartmentalization, perhaps functioning as back-bounding structure for the release of the main part of the Punta Cola rockslide. This hypothesis is supported by the discontinuity set D3' that is roughly parallel to the fractured zone and that locally creates step fractures on the complex BFS (Figure 5).

Since the slide and tsunami occurred almost immediately following the 21 April 2007 earthquake, there seems little doubt the trigger was seismic. An outstanding question deserving further research is whether or not structural compartments that likely exist elsewhere in the Aysén fjord region also require a seismic trigger, or if gravitational processes are sufficient to engender failure. Similarly, future research should also address whether or not the slopes that failed during and after the 2007 earthquake exhibited signs of pre-failure deformation.

While it is not yet possible to claim certainty, we firmly hypothesize that compartmentalization and catastrophic release of the 2007 Punta Cola rockslide was directly related to a network of faults and fractures permeating the underlying bedrock. These brittle structures, we opine, symbolize the widely- distributed zone of deformation that is the LOFZ itself. This portion of the LOFZ is subjected to ongoing 5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile compressive stress (Figure 2e); new structures may well be forming throughout the region. Consequently, we are intrigued by the idea that the broad zone of mylonites, brittle faults, and open fractures in the Aysén region might represent a relatively recent stage of fault strand evolution in the LOFZ, perhaps spawned by upper/lower crust decoupling in response to changes in relative plate motion, buttressing, and block rotation.

A possible role of regional tectonics in the large rock slope failures of southern Chile Citing Pankhurst et al. (1999), Thompson (2002) presented a map showing intrusion ages for the North Patagonia Batholith. The four intrusive ages nearest to the Aysén region are 10 Ma, 20 Ma, 39 Ma, and 98 Ma. Representative ZFT/AFT pairs near the Punta Cola slide are (10 +/-1; 5 +/-1) Ma, (11 +/-1; 5 +/-1) Ma, and (12 +/-1; 4 +/-1) Ma (Thompson, 2002).

Although a proper interpretation of AFT thermal models demands geological sanity as opposed to irrational exuberance (see snippets of Redfield, 2010), the young ZFT and AFT ages from near Punta Cola do impose a requirement of relatively rapid cooling since the middle Miocene. ZFT ‘ages’ correspond to temperatures ranging from 235°C to 375°C (~7 to ~14km depth) whilst AFT ‘ages’ imply temperatures between 60°C to 110°C (~2.5 to ~5km depth). The long AFT mean track lengths and narrow, unimodal histograms reported by Thompson (2002) suggest rapid transit through the Apatite Partial Annealing Zone (APAZ) ~ in accord with the preponderance of high-angle structures with steep lineations found throughout the Aysén region during our 2010 field work (Figure 6).

The mechanism of exhumation of the ductile, basement-crosscutting mylonite shear zones to the surface is uncertain, but may be relatively important to our study. The depth of mylonite formation was probably on the order of 35-55km (Pankhurst et al., 1999; Chlieh et al., 2004). Strike-slip faulting with an upward obliquity of 10 degrees requires some 200 km of particle motion for 35 km exhumation, illustrated by the geometric relation (Equation 1):

l = /d sin θ (1) where l is the path length, d is the emplacement depth, and θ is the angle from the horizontal. Much shorter path lengths (and thereby less horizontal translation) are required if exhumation pathways are steeper. Table 1 shows such a simple rate calculation for various strike-slip exhumation angles for each of the four intrusive ages suggested by Thompson (2002) from an assumed emplacement depth of 35 km.

Table 1. Calculated exhumation rates along the oblique path defined by the angle away from the horizontal for a depth of emplacement at 35 km. Angle θ Path length l Exhumation rate along path [km/Ma = mm/yr] [°] [km] 10 Ma 20 Ma 39 Ma 98 Ma 0 Infinity Infinity Infinity Infinity Infinity 10 202 20.16 10.08 5.17 2.06 20 102 10.23 5.12 2.62 1.04 30 70 7.00 3.50 1.79 0.71 40 54 5.45 2.72 1.40 0.56 50 46 4.57 2.28 1.17 0.47 60 40 4.04 2.02 1.04 0.41 70 37 3.72 1.86 0.96 0.38 80 36 3.55 1.78 0.91 0.36 90 35 3.50 1.75 0.90 0.36

5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile Under oblique , detached slivers of continental crust can be translated parallel to a convergent margin (Beck, 1983; 1989; Jarrard, 1986). However, if actual transport is less than the maximum margin- parallel component of relative motion, the sliver will be subject to shear and may fracture into rotating blocks (Beck, 1989). Paleomagnetic data indicate that plate scale northwards translation is absent in the Chilean (e.g. Cembrano et al., 1996). Rather, statistically significant block rotations have been paleomagnetically documented in southern Patagonia (García et al., 1988; Beck et al., 1993; Rojas et al., 1994). To a certain degree, then, the brittle upper crust is probably structurally decoupled from the ductile lower crust.

Decoupling may perhaps be important from the perspective of fault geometry. Figure 1 shows NAZ-SAM relative plate motion convergence vectors resolved against a N-S trending backstop at 45°S. The magnitude of the tangential (strike-slip) component dropped dramatically circa 25 Ma (Somoza, 1998) (Figure 1b). Near Aysén, from 25 Ma on, strike-slip plate tectonic components have averaged less than 30 mm/yr, whilst averaged orthogonal components slightly exceed 100 mm/yr. Table 1 reveals exhumation rates that approach plate tectonic rates. Assuming NAZ-SAM coupling was/is 100% efficient, all particle path rates shown in Table 1 are permissible. However, the paleomagnetically-derived evidence cited above for upper/lower crustal decoupling suggests high efficiency is not likely. A more reasonable (albeit arbitrary) value of 10% that reflects significant crustal decoupling implies that any Table 1 particle path greater than ~3 mm/yr is probably unacceptable. Under this hypothetical, tectonically exhuberent model, exhumation of mylonites by strike-slip faulting is likely only if the intrusion ages of the Aysén country rock are greater than ~40 Ma. If a ~20 Ma intrusion age is adopted, particle paths greater than 40° from the horizontal are likely necessary. Studies of country rock and mylonite ages, pluton emplacement mechanisms, and pluton exhumation kinematics may have a certain bearing on geomorphic stability in the Aysén region.

Figure 8. Schematic model linking regional tectonics to the Punta Cola rockslide. Left: overview diagram after Thompson (2002) showing an hypothesized LOFZ flower structure geometry. Steeply-dipping mylonite fabric (blue lines) is assumed to be a product of LOFZ transpressive deformation. Passage of mylonites through the ductile-brittle transition imposes a brittle fabric parallel or sub-parallel to ductile shear planes. Exhumation along 'pop-up' or flower structure significantly widens the zone of brittle deformation, offering new passageways for melts such as the olivine-bearing basaltic dike observed at Punta Cola (Figure 2c). Right: The strikes of olivine- bearing basaltic dike/fault, stockwork of very dense fractures, and the pre-failure drainage are sub- parallel to the main trend of the LOFZ (see Figure 1c). Mylonitic fabrics observed at Isla Mentirosa appear to strike into release planes of rockslide scar on the opposite hillside. Thrust fault and tilted beds indicate compressive deformation continues today. 5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile

We raise also the question of how rockslide geohazard might differ along the length of the LOFZ. Figure 1c illustrates both the diffuse nature of the southern LOFZ and its extremely high level of susceptibility to catastrophic failure, perhaps due to ‘stockworking’ of deformation in the uppermost, most brittle crust. At its northern end, the LOFZ becomes again a diffuse zone of deformation. Large volume rockslides also occur and can be linked to neotectonic structures with evidence of Quaternary activity (Folguera et al., 2004). In between, current mapping suggests the fault strands comprising the LOFZ become less diffuse and more well-defined at the regional scale. There, fewer large rockslides have been reported. A regional mapping programme documenting LOFZ structures and LOFZ-related rock instabilities is important for the future, and should consider a wide range of disciplines as opposed to solely rock mechanics.

Tilted Quaternary sedimentary layers and a small but well-defined thrust fault in Quaternary clays (Figure 2e) indicates that compressive deformation in Aysén continues today. However, many other studies (Cembrano et al., 1996; 1999; 2000) suggest strike-slip structures dominated the central LOFZ. This ‘discrepancy’ underlines the need for a judicious subdivision and categorization of the LOFZ into structural domains based on their structural style, the types and quantities of geohazards, and how deformation is partitioned with respect to the overall plate-scale convergent stress field. A pertinent question, hinted towards in Figure 8, is whether buttressing along the LOFZ (e.g. Beck et al., 1993) helps define the rockslide hazard potential of the LOFZ. For example, 1) is the hypothesized LOFZ flower structure limited to where crustal decoupling is sufficiently great to permit or engender small block rotation, 2) does the widening of the LOFZ brittle regime relate directly to that crustal pop-up, and 3) might a change in geometry to a less diffuse collection of better-defined fault strands impose a natural limit to large-volume rockslides in the southern end of the LOFZ?

As a plate scale feature, the LOFZ has much to teach us about how large faults develop through time and space, and how their impact upon the landscape can affect our daily lives.

AKNOWLEDGEMENTS

Support for this work was provided in part by the Research Council of Norway through the International Centre for Geohazards (ICG). Their support is gratefully acknowledged. S. Loew and colleagues gratefully acknowledge funding by ETH CCES Project COGEAR. S. Sepúlveda and S. Rebolledo gratefully acknowledge funding from the Montessus de Ballore International Earthquake Research Center and Fondecyt project 11070107. This is ICG contribution no. 305. All authors thank the gallant crew of the BARCO for logistical assistance far beyond the call of duty, bottomless maté, and el cordero suprema.

REFERENCES

Arenas, M., Naranjo, J., Clavero, J. and Lara, L. (2008). “Earthquake-induced landslides: susceptibility mapping for crisis management”. Congreso Geológico Argentino, No. 17, Actas 1, p. 255. Beck, M. E. (1983). “On the mechanism of tectonic transport in zones of oblique subduction”. Tectonophysics, Vol. 93, pp. 1-11. Beck, M. E. (1989). “Block rotations in continental crust: examples from western North America”. In: C. Kiessel and C. Laj (eds), “Paleomagnetic Rotations and Continental Deformation”, Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 327-342. Beck, M.E., Rojas, C. and Cembrano, J. (1993). “On the nature of buttressing in margin-parallel strike- fault systems”. Geology, Vol. 21, pp. 755-758. 5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile Cembrano, J., Herve, F. and Lavenu, A. (1996). “The Liquiñe Ofqui fault zone: a long-lived intra-arc fault system in southern Chile”. Tectonophysics, Vol. 259, pp. 55-66. Cembrano, J., Lavenu, A., Arancibia, G., López, G. and Sanhueza, A. (1999). “Crustal-scale pop-up structure at the southern Andes plate boundary zone: A kinematic response to Pliocene transpression”. Fourth International Symposium on Andean Geodynamics, Göttingen, Germany, pp. 151-154. Cembrano, J., Schermer, E., Lavenu, A. and Sanhueza, A. (2000). “Contrasting nature of deformation along an intraarc shear zone, the Liquiñe-Ofqui fault zone, southern Chilean Andes”. Tectonophysics, Vol. 319, pp. 129-149. Cembrano, J., Lavenu, A., Yañez, G., Riquelme, R., Garcia, M., Gonzalez, G. and Hérail, G. (2007). “Neotectonics”. In: T. Moreno and W. Gibbons (eds), “The Geology of Chile”, The Geological Society, London, 414 pp. Cembrano, J. and Lara, L. (2009). “The link between volcanism and tectonics in the southern volcanic zone of the Chilean Andes: A review”. Tectonophysics, Vol. 471, pp. 96-113. Chlieh, M., de Chabalier, J.B., Ruegg, J.C., Armijo, R., Dmowska, R., Campos, J. and Feigl, K.L. (2004). “Crustal deformation and fault slip during the seismic cycle in the North Chile subduction zone, from GPS and InSAR observations”. Geophys. J. Int., Vol. 158, pp. 695-711. Folguera, A., Ramos, V.A., Hermanns, R.L. and Naranjo, J. (2004). “Neotectonics in the foothills of the southernmost central Andes (37°–38°S): Evidence of strike-slip displacement along the Antiñir- Copahue fault zone”. Tectonics, Vol. 23, TC5008, doi:10.1029/2003TC001533. García, A., Beck, M.E., Burmester, R.F., Hervé, F. and Munizaga, F. (1988). “Paleomagnetic reconnaissance of the region de los Lagos, southern Chile, and its tectonic implications”. Rev. Geol. Chile, Vol. 15, pp. 13-30. Hervé, F. and Thiele, R. (1987). “Estado de conocimiento de las megafallas en Chile y su significado tectonico”. Comun. Univ. Chile, Vol. 38, pp. 67-91. Hervé, F., Pankhurst, R.J., Drake, R., Beck, M.E. and Mpodozis, C. (1993). “Granite generation and rapid unroofing related to strike-slip faulting, Aysén, Southern Chile”. Earth and Planetary Science Letters, Vol. 120, pp. 375-386. Hervé, F. (1994). “The southern Andes between 39° and 44°S latitude: the geological signature of a transpressive tectonic regime related to a magmatic arc”. In: K.-J. Reutter, E. Scheuber and P.J. Wigger (eds), “Tectonics of the Southern Central Andes”, Springer, Berlin, pp. 243-248. Hervé, F. and Fanning, M. (2001). “Late detrital zircons in metaturbidites of the Chonos Metamorphic Complex, southern Chile”. Revista Geológica de Chile, Vol. 28, No. 1, pp. 91-104. Hervé, F., Fanning, M. and Pankhurst, R.J. (2003). “Detrital zircons age patterns and provenance of the metamorphic complexes of southern Chile”. Journal of South American Earth Sciences, Vol. 16, pp. 107-123. Jaboyedoff, M., Metzger, R., Oppikofer, T., Couture, R., Derron, M.-H., Locat, J. and Turmel, D. (2007). “New insight techniques to analyze rock-slope relief using DEM and 3D-imaging cloud points: COLTOP-3D software”. In: E. Eberhardt, D. Stead and T. Morrison (eds), “Rock mechanics: Meeting Society’s challenges and demands. Proceedings of the 1st Canada – U.S. Rock Mechanics Symposium, Vancouver, Canada, 27–31 May 2007”, Taylor & Francis, London, Vol. 1, pp. 61-68. Jarrard, R. D. (1986). “Relations among subduction parameters”. Reviews of Geophysics, Vol. 24, pp. 217-284. Lim, M., Petley, D. N., Rosser, N. J., Allison, R. J., Long, A. J. and Pybus, D. (2005). “Combined Digital Photogrammetry and Time-of-Flight Laser Scanning for Monitoring Cliff Evolution”. The Photogrammetric Record, Vol. 20, No. 110, pp.109-129. Naranjo, J.A., Arenas, M., Clavero, J. and Lara, L. (2007). “Estudio preliminar de peligros por remociones en masa por el sismo -Mw 6,2-del 21 de abril de 2007, Región de Aisén”. Servicio Nacional de Geología y Minería (SERNAGEOMIN), Santiago (unpublished). Naranjo, J.A., Arenas, M., Clavero, J. and Munoz, O. (2009). “Mass movement-induced tsunamis: main effects during the Patagonian Fjordland seismic crisis in Aisén (45°25’S), Chile”. Andean Geology, Vol. 36, pp. 137-145. 5th International Conference on Earthquake Geotechnical Engineering January 2011, 10-13 Santiago, Chile Oppikofer, T., Jaboyedoff, M., Blikra, L.H. and Derron, M.-H. (2009). “Characterization and monitoring of the Åknes rockslide using terrestrial laser scanning”. Natural Hazards and Earth System Sciences, Vol. 9, No. 3, pp. 1003-1019. Pankhurst, R.J., Weaver, S.D., Hervé, F. and Larrondo, P. (1999). “Mesozoic-Cenozoic evolution of the North Patagonian Batholith in Aysén, southern Chile”. Journal of the Geological Society, Vol. 156, pp. 673-694. Redfield, T. F. (2010). “On apatite fission track dating and the Tertiary evolution of West Greenland topography”. Journal of the Geological Society, London, Vol. 167, No. 2, pp. 261-271, doi: 10.1144/0016-76492009-036. Rojas, C., Beck, M.E., Burmester, R.F., Cembrano, J. and Hervé, F. (1994). “Paleomagnetism of the Mid- Tertiary Ayacara Formation, southern Chile: counterclockwise rotation in a dextralshear zone”. Journal of South American Earth Sciences, Vol. 7, No. 1, pp. 45-56. Sepúlveda, S.A., Serey, A., Lara, M., Pavez, A. and Rebolledo, S. (2010). “Landslides induced by the April 2007 Aysén Fjord earthquake, Chilean Patagonia”. Landslides, in press, doi: 10.1007/s10346- 010-0203-2. Sepúlveda, S.A. and Serey, A. (2009). “Tsunamigenic, earthquake-triggered rock-slope failures during the April 21, 2007 Aisén earthquake, southern Chile (45.5°S)”. Andean Geology, Vol. 36, pp. 131-136. Somoza, R. (1998). “Updated Nazca (Farallon)–South America relative motions during the last 40 My: Implications for mountain building in the central Andean region”. Journal of South American Earth Sciences, Vol. 11, pp. 211–215. Suarez, M. and De La Cruz, R. (2000). “Tectonics in the eastern central Patagonian Cordillera (45°30’- 47°30’ S)”. Journal of the Geological Society, Vol. 157, pp. 995-1001. Sturzenegger, M. and Stead, D. (2009). “Close-range terrestrial digital photogrammetry and terrestrial laser scanning for discontinuity characterization on rock cuts”. Engineering Geology, Vol. 106, No. 3- 4, pp. 163-182. Sturzenegger, M., Yan, M., Stead, D. and Elmo, D. (2007). “Application and limitations of ground-based laser scanning in rock slope characterization”. In: E. Eberhardt, D. Stead and T. Morrison (eds), “Rock mechanics: Meeting Society’s challenges and demands. Proceedings of the 1st Canada – U.S. Rock Mechanics Symposium, Vancouver, Canada, 27–31 May 2007”, Taylor & Francis, London, Vol. 1, pp. 29-36. Thompson, S. (2002). “Late Cenozoic geomorphic and tectonic evolution of the Patagonian Andes between latitudes 42°S and 46°S: An appraisal based on fission-track results from the transpressional intra-arc Liquiñe-Ofqui fault zone”. Geological Society of America Bulletin, Vol. 114, pp. 1159-1173. Vargas, G., Rebolledo, S., Sepúlveda, S.A., Thiele, R., Townley, B., Padilla, C., Rauld, R. and Herrera, M. (2010). “Submarine seismic rupture and active tectonics along the transpressional intra-arc Liquiñe- Ofqui Fault Zone in the Patagonian Andes (45.5°S)”. Submitted.