Typical landforms and their subsurface structures formed by mass rock creep. Roman numerals show the structure types defined by Chigira (1992). Interpretive cross-sections at lo wer right (Arabic numerals) are derived from: (1) Mahr and Nemcok, 1977; (2) Ando et al., 1970; (3) Tabor, 1971; (4) Radbruch-Hall, 1978 and Shimuzu et al.,1980; (5) Jahn, 1964. 

Contexts in source publication

Context 1

... term “sackungen” (from the German verb “to sag”) describes a family of landforms in mountainous areas that include crestal troughs, antislope scarps, and closed depressions ( Fig. 1). Although many authors have concluded that sackungen result from slow mass rock creep (e.g., Chigira, 1993), the more linear sackungen also resemble tectonic fault scarps. In addition, sackung- like landforms have formed or have been rejuvenated during historic earthquakes (Dramis and Sorriso-Valvo, 1983; Wallace, 1984; Cotton et al., 1990; Ponti and Wells, 1991; Nolan and Weber, 1992). Sackungen may therefore have formed by: 1) displacement on tectonic faults, 2) gravity failures caused by earthquake shaking, or 3) gravity failures unrelated to tectonics. Criteria for distinguishing among these three possible origins are geomorphic, structural, and stratigraphic. Geomorphic criteria are based on qualitative observations of scarp length, continuity, plan shape, and relation to topography. The ratio of scarp height: length may be a useful quantitative criterion. Fault scarps of tectonic origin typically exhibit small (<10) height: length ratios, whereas gravity scarps are commonly much shorter for a given height. In addition, gravity-related sackungen are typically short, discontinuous, arcuate, and occur in swarms of multiple parallel scarps, whereas fault scarps are longer, continuous, linear, and singular. Structural criteria address the morphology of the shear zone beneath sackungen landforms and contemporary movement. Subsurface shear zones created by gravity creep are consistently asymmetrical, with a sharp upper contact and a transitional lower contact (Fig. 2). Tectonic shear zones are commonly more symmetrical, especially if they formed at deep crustal levels and are now exposed after considerable erosion. Some very, linear antislope scarps, originally interpreted as coseismic fault scarps, possess geodetically documented aseismic slip rates of up to 10 mm/yr (Bovis and Evans, 1995); such contemporary movement strongly suggests a gravity origin. Gravity shear zones also have predictable locations and orientations with respect to the present mountain ridge topography, relations that would be fortuitous for tectonic faults. For example, the toppling test of Goodman and Bray (1976) can indicate whether preexisting discontinuities ought to be failing under gravity stresses. Finally, the stratigraphy and deformation of fine-grained Holocene sediments in sackung- related troughs and depressions can indicate whether the formation of these landforms was slow and gradual, or episodic. If formation was slow and gradual, then the features cannot be coseismic. Landforms created episodically could be of either tectonic or gravity origin, because even landslides may undergo episodic movement in response to climatic forcing, episodic basal erosion, or episodic loading. In summary, determining a gravity- versus tectonic origin for sackung- like landforms (and their underlying shear zones) is best assessed by the application of multiple geomorphic, structural, and stratigraphic criteria. Use of these criteria requires geodetic, geomorphologic and microstratigraphic investigations (such as from trenching) similar to those utilized for the paleoseismic study of tectonic ...

Context 2

... near old faults were probably seismogenic, but that most of the shorter, more sinuous scarps at the lips of steep slopes were of gravitational origin. A second series of studies in California also resulted in controversy. The extensive ground cracking caused by the 1989 Loma Prieta, California earthquake was coincident with antislope scarps, benches, and ridge top grabens at Summit Ridge, Santa Cruz County. Trenches excavated by Cotton et al. (1990) and Nolan and Weber (1992) across the 1989 ground cracks clearly showed that most of them had experienced prehistoric dip-slip displacement, which in turn had created the sackung-like landforms at the ridge crest. However, workers disagreed about the mechanism that created the ground fissures in 1989, and by analogy, the underlying prehistoric displacements. ‘The debate over the origin of the 1989... ground ruptures centers around whether the ridge-top fissures formed primarily as a result of tectonic deformation produced by the earthquake or whether they mostly reflect gravity-driven processes triggered by strong ground motions” (Ponti and Wells, 1991, p. 1495). Cotton et al. (1990) argue that fissures and faults resulted from normal bending- moment slip on bedding plane faults, caused by coseismic folding adjacent to the San Andreas fault. Because they considered the landforms to be the result of repeated secondary faulting caused by Loma Prieta-type earthquakes, they suggested that trenching studies in grabens could reconstruct a proxy paleoseismic history of the San Andreas fault. In contrast, Ponti and Wells (1991) concluded that the cumulative extensional and vertical displacements on cracks in 1989 were 35 times larger than could be explained by a bending- moment fault model. They concluded that approximately 90% of the ground displacement was due to downslope movement (landsliding) caused by prolonged seismic shaking. Similar controversies have arisen in Canada, where occasional long, linear antislope scarps occur in areas with more abundant short, arcuate bedrock scarps. For example, Eisbacher (1983) identified the prominent linear antislope scarp on Mt. Currie (British Columbia) as a young fault scarp. In contrast, Evans (1987) and Bovis and Evans (1995) documented that: 1) the joints beneath the scarp were predicted to fail by the toppling test of Goodman and Bray (1976), and 2) between 1987 and 1991 up to 80 mm of horizontal displacement and 40 mm of vertical displacement occurred across the scarp, in the absence of any earthquakes. They thus concluded the scarp, despite its linearity, was a gravitational failure, The Hell Creek scarp (British Columbia) has been interpreted as a Holocene tectonic reactivation of an older fault (Psutka, 1995), based mainly on evidence for episodic displacement (observed in trenches) with a lateral component (inferred from weak geomorphic evidence). In contrast, Clague and Evans (1994) consider the scarps to reflect nonseismic gravity failure, similar to many other scarps in the area. The purpose of this paper is to: 1) summarize the identifying characteristics of sackungen, and 2) to provide criteria that enable the differentiation between sackungen- like landforms that are tectonic versus nontectonic in origin. This summary paper is based on a compilation of literature on sackungen and other scarplike landforms throughout the world, and on a synthesis of field reconnaissance and detailed field studies in western North America with the worldwide data base. Zischinsky (1966, 1969) first proposed the term ‘sackung” for the surface manifestations of deep-seated rock creep in foliated bedrock of the Alps. In this paper the term “sackung” refers to the process of deep-seated sagging, whereas ‘sackungen” is used as a generic term to describe any landforms such as antislope scarps or ridge-crest depressions, regardless of origin. Other workers have referred to the same slow rock deformation process as mass rock creep (MRC; Radbruch-Hall, 1978), depth creep (Ter-Stepanian, 1966), deep-seated creep (Nemcok, 1972), deep-seated continuous creep (Hutchinson, 1988), bedrock flow (Varnes, 1978), or gravitational spreading (Radbruch-Hall et al., 1977; Varnes et al., 1989). Unfortunately, these processes have usually been inferred to explain landforms whose exact mode and rate of formation is unknown, and which could be (in part) created by rapid displacements accompanying surface fault rupture or earthquake shaking. Sackungen are landforms, so inferences concerning their origin have traditionally been based purely on geomorphic evidence. This evidence emphasizes that sackungen must have a deep-seated rock- failure origin because they occur in topographic positions (high on ridge flanks, at ridge crests) where an erosional origin is extremely unlikely. The most common geomorphic features can be divided into four categories: downhill- facing scarps, double-crested ridges, uphill- facing scarps, and erosional notches on ridge axes (Fig. 1). Double-crested ridges (the “Doppelgrat” of Zischinsky, 1969) are a classic sackung landform. The axial depression typical of Doppelgrat is difficult to explain by an erosional process, because streams are unlikely to flow down the crest of a ridge and the closed depressions could not have been excavated by running water. Uphill- facing scarps and sidehill benches (degraded scarps?) are probably the most common sackung landform (e.g., Varnes et al., 1989). Along strike, sackungen may grade from antislope scarps to benches, and grabens may grade into irregular closed depressions; along-strike changes in height and morphology are common. An inventory of published sackung scarp dimensions (McCleary et al., 1978, Table 2) yields these typical ranges: scarp length, 15-300 m; scarp height, 1-9 m; slope height, 400-1200 m; slope gradient, 25-50 degrees. However, Salvi and Nardi (1995) interpret a trough 100 m deep, 700 m wide, 9 km long in the Apennines (Italy) as an earthquake- induced sackung. Almost all of these landforms are found at or near the crests of slopes, in the zone of tensional failure. In contrast, very few distinctive landforms have been observed on the lower slopes, except where authors have postulated that lower slopes have been “oversteepened” or bulged outward by compressive forces at the “toe” of a creeping rock mass. Two opposing hypotheses have been proposed for the subsurface geometry of sackungen in massive competent rocks. One, held by Zischinsky (1969) and other European and American workers, proposes that “a well-defined slide plane near the headscarp passes downward into a broader zone of rock creep. Consequently the lower portion of this type of failure simply bulges out into the valley” (Morton and Sadler, 1989, P. 302). The slide plane may dip either into or out of the slope. Such slow, deep-seated failure results in “half- a-landslide” morphology (Morton and Sadler, 1989), with well-developed tensional features near the head, but often with no recognizable evidence of medial landslide features or compressional morphology downslope from the scarp. Radbruch-Hall (1978) claims that rock creep can extend to depths of several hundred meters. However, there are few locations where the depth or shape of the failure plane can be measured with certainty, making the “half-a- landslide” hypothesis difficult to directly test. The second hypothesis is that sackungen are shallow surface manifestations of toppling and flexural slip along discontinuities that dip steeply into a mountain mass, but which do not penetrate to any great depth (Jahn, 1964, his Fig. 9; Beck, 1968). Bovis (1982) termed this process “flexural toppling” and cit ed model studies (Barton, 1971) and studies in quarries (Goodman and Bray, 1976) as support for this non-penetrative mode of extensional deformation. During flexural toppling outward rotation of blocks and dilation of sackung cracks lead to attenuation of movement with time, which Bovis (1982) compared to strain- hardening in granular materials. Given the steep dip inferred in this model, the sackungen could possibly connect to seismogenic faults at depth, although the vertical separation of the ground surface along sackung scarps would not necessarily bear any relation to fault movement at depth. The most detailed study of subsurface deformation features (folds, faults) associated with mass rock creep is that of Chigira (1992). Chigira was mainly concerned with characterizing the micro- and meso-scale characteristics of the causative shear zones that underlay sackungen at depths of 10s to 100s of meters. Chigira (1992, p. 174-175) describes the discrete “fault” zones that underlie areas of mass rock creep as “a pulverized zone with fault gouge and a phyllitic or brecciated zone (Fig. 2). In a densely foliated rock (Fig. 2A1, 2A2), the phyllitic zone is formed by microscopic slip along foliations. In a sparsely foliated rock (Fig. 2B1, 2B2), the brecciated zone is formed by random crushing. In massive rocks (Fig. 2C1, 2C2), a brecciated zone is formed through networks of tension fractures, but not if it is formed through connection of shear fractures. The pulverized zone, which usually forms only in the upper part of the shear zone, is formed through grinding by a downsliding block on the fault.” He assumed that the features he described were of gravitational rather than tectonic origin, and did not address the issue of possible shaking- induced gravitational movement. The details of Chigira’s study are presented in Sec. 4.2. Varnes et al. (1989) distinguish three types of sackung: 1) spreading of rigid rocks overlying soft rocks (Radbruch-Hall, 1978; Radbruch-Hall Ct al., 1976), 2) sagging and bending of foliated phyllites, schists, and gneisses (“true Sackung” of Zischinsky, 1969), and 3) differential displacements in hard but fractured crystalline igneous rocks. Sackungen have been observed in almost all rock types, including phyllite and schist ...