DO NOT CITE THIS WEB PAGE -- ALWAYS READ & CITE THE
PUBLISHED TEXT
This document is the manuscript that was submitted to the Journal
of Structural Geology.
It may have been altered slighly by the editors.
Annotations to the text will appear in the same color as this note
Department of Geology
Portland State University
Portland, OR 97307-0751
On the southwest limb of Salt Valley Anticline are three sets of joints in the Entrada Sandstone covering an area of about 6 km2. Within the 20 m thick Moab Member, a single joint set is found in three distinct areas, separated by a second set of joints, at a 35° angle to the first. Joint interaction features show that the second set is younger than the first. This illustrates that joints of a single set do not have to fill the entire area across which the stresses that formed the joints were acting. The underlying Slickrock Member contains the third set of joints, which is at an angle of 5°-35° to joints in the Moab Member. The Slickrock set nucleated from the lower edges of joints of all orientations in the overlying Moab Member. Thus, the fracture pattern evolved both horizontally, within the same unit, and vertically between units. The sequence of jointing is determined by establishing the relative ages of each joint set. Each joint orientation is best interpreted as representing a direction of maximum compression, ruling out the possibility that the joints are a conjugate set. The joints, and an earlier set of deformation bands, record a 95° counterclockwise rotation of the direction of maximum compression.
Joints are spectacularly developed and displayed on the flanks of Salt Valley Anticline in SE Utah (Figs. 1, 2a). Joints south of Klondike Bluffs on the SW flank of the anticline appear in at least three sets. Zones of joints at some places within the Entrada Sandstone have weathered as long, narrow grooves to produce elongate rock fins, while in other places there has been almost no erosion along joint traces. Areas with little erosion are ideal for studying joint patterns. Although the joint pattern is very simple in some locations (e.g.; Cruikshank et al. 1991), there are sites with multiple sets of joints; the Arches National Park and vicinity provides an excellent area for studying the geometry of either a single joint set, or multiple joint sets, at a kilometre scale. While the spectacular joint patterns around Salt Valley have attracted attention and are featured in several textbooks (e.g. ; Davis 1984), there have been few detailed studies of the joints (Dyer 1983; Cruikshank et al. 1991).
Fig. 1 (a) Simplified geologic map and stratigraphy for Arches National Park, located in SE Utah. The study area is at the NW corner of the park (map courtesy of Utah Geological Survey). (b) Cross-section through the Klondike Bluffs area. (Doelling 1988, section J p. 48). The Navajo and Wingate sandstones are shaded for reference. The study area is on the SW edge of the section. (c) Simplified stratigraphy for Arches National Park. The joints shown in Fig. 4 are within the Moab and Slickrock Members of the Jurassic Entrada Sandstone.
Fig. 2 (a) Oblique aerial photograph looking north across the study area. Photograph shows the grooves on a dip slope of the Moab Member of the Entrada Sandstone on the southwest limb of Salt Valley Anticline, in the vicinity of Klondike Bluffs. The grooves generally correspond to zones of proto-joints. An abrupt change in the orientation can be seen near the center of the photograph (below "Tower Arch" label), which corresponds to the change in orientation shown in Figs. 4 & 3, near coordinates 5500N. (b) Oblique aerial photograph of the Klondike bluffs area, looking south, showing fault scarp and relation of joints to scarp. Notice that not all joints cross the fault surface. In the area where the systematic joints do not cross the surface, there is a polygonal-like joint network. The photograph covers the area shown in Fig. 7.
According to studies of sheet structures by Holzhausen (1977; 1989), each joint set should represent a different direction of maximum compression. This would imply, though, that the various joint sets, which appear to be conjugate are, in fact, independent. One purpose of this investigation is to determine the age relations of joints in the various sets to establish whether their ages differ. Another is to determine whether there is a pattern to the change in direction of the maximum compression in space and time. This investigation also aims to determine if there are different joint sets in the Moab Member and the underlying Slickrock Member. A final objective is to ascertain whether joints in one set influenced those in another set. To answer these questions, joints were mapped at a scale of 1:4500 over an area about 6 km long, and 0.5 to 1.5 km wide (Figs. 3 & 4). Within the mapped, area joint orientations change abruptly by up to 35° laterally within the same unit, and vertically between units.
Fig. 3. Mosaic of aerial photographs of the area shown in Figs. 2a & 4. Salt Valley is in the NE corner of the photograph. Zones of proto-joints show up well as lineations on the photographs. Most joints are exposed on a dip-slope of the Moab Member of Entrada Sandstone. The Entrada Sandstone forms a NE facing cuesta overlooking Salt Valley. The Slickrock Member is more deeply dissected, and forms a series of tall fins, such as in the vicinity of Tower Arch. The coordinate system for the photograph is the same as in Fig. 4. The mosaic is composed of images AB4UPRM0001656A-77 and -78 available from U.S. Geological Survey EROS Data Center.
Fig. 4. Map showing traces of joints and faulted-joint in Entrada Sandstone on the NW limb of Salt Valley Anticline. All the units in the area dip gently to the southwest. The coordinate system is aligned along true North, and is in metres from an arbitrary datum used in Cruikshank et al. (1991). Mapped area covers parts of Sections 22, 26, 27, 34, & 35 T23S R20E, U.S. Geological Survey Klondike Bluffs 7.5' quadrangle. The blank area in the vicinity of (7500N 27500W) represents an area where zones of joints are wide, sediment filled grooves. This area could not be mapped with the same detail and quality as the rest of the study area.
In the Garden area - 3 to 4 km south of the current map area - Cruikshank, Zhao and Johnson (Cruikshank et al. 1991; 1991; Zhao & Johnson 1991; 1992) studied joints, faulted-joints, and deformation bands (Aydin 1978; 1978) within the Moab Member of the Entrada Sandstone over an area of about 1 km2. They showed that, in order to understand the deformation represented by fractures, it is important to distinguish between deformation bands and faulted-joints. Deformation bands form in shear and have a few mm to cm of shear offset, while faulted-joints form as extension fractures (joints) but later slip a few mm to cm. Thus, a fracture which originally formed as a tension crack may now show shear displacements. We use the term proto-joint to describe a fracture that formed as an opening-mode (mode I) fracture, but may have slipped since formation.
The sequence of deformations established for the Garden area were (Zhao and Johnson 1992): (1) Formation of a set of conjugate, strike-shift deformation bands with maximum compression oriented about N45°W. (2) Local formation of joints along the traces of a few deformation bands. (3) Formation of a systematic joint set trending N15°W to N5°W, parallel to the maximum compression direction. The joint segments that formed along the deformation bands slipped at this time, with a sense of shift opposite to that of the deformation bands. (4) The final deformation recorded in the Garden area was shear offset of a few mm to cm along the systematic joint set, forming faulted-joints. Some traces of zones of joints had right-lateral offsets in the southern part of their traces, no offset in the central area, and left-lateral offsets in the northern part of the Garden area. The sequence of deformations in the Garden area can be explained as a counter-clockwise rotation of the direction of principal compression, from about N45°E to N5°W. The sequence of structures in the current study area records a larger counter-clockwise rotation in the direction of maximum compression, from about N65°E to at about N30°W.
Dyer (1983; 1988) studied proto-joints in the Moab Member of the Entrada Sandstone at three locations. Dyer's domain B (1983, ch. 2; 1988, Fig. 3) is on the south-west limb of Salt Valley Anticline and corresponds to the southern part of the Anniversary Arch domain in this paper (south of 5000N, Fig. 4). Dyer noted that, in this area, proto-joints in the Slickrock Member have a very different orientation from those in the overlying Moab Member (e.g., 4700N 1200W, Fig. 4). He also studied the Garden area (1983, p. 83) and near the Arches National Park campground (Dyer 1988, Fig. 2), on the east flank of Salt Valley (Fig. 1a). Dyer's study (1988) focused on the interpretation of cross-joints found between an older systematic joint set.
Salt Valley Anticline is a breached, asymmetric, salt-cored anticline at the NE edge of the Paradox basin (Fig. 1a). Jurassic and Cretaceous sedimentary rocks exposed on the flanks of Salt Valley Anticline are inclined up to about 15°; within the core of the valley, however, bedding may be vertical or overturned (Shoemaker et al. 1958; Elston et al. 1962; Doelling 1985). The salt is part of the late Paleozoic Paradox Formation, and started to move soon after the Paradox evaporites were covered with other sediments. Unconformities in sediments overlying the Paradox Formation suggest that there were several periods of rapid salt movement throughout the Mesozoic and Cenozoic, and deformed Quaternary deposits in Salt Valley and neighboring salt-anticline valleys indicate that the salt is still moving (Colman et al. 1988; Oviatt 1988).
The fractures we mapped are in the Moab and Slickrock members of the Jurassic Entrada Sandstone (Fig. 1). The Entrada Sandstone rests unconformably on the Jurassic Navajo Sandstone. The Navajo Sandstone thins considerably over the anticline, and may not have been deposited over the entire crest. Although the normal thickness of the Navajo Sandstone in this area is on the order of 150 m, a well near the Rock Corral area (~6 km N of study area) went through 55 m of Navajo Sandstone directly into the Paradox Formation. (Doelling 1988, p. 27). The Entrada Sandstone was deposited over the entire anticline, although it may be slightly thinner over the crest of the anticline (Dyer 1983; Doelling 1988).
The Moab Member is a light-colored, clean, fine- to medium-grained sandstone, about 20 to 30 m thick. Underlying the Moab Member is the Slickrock Member, 60 to 160 m of dark-red, massive, fine-grained sandstone. The lowest member of the Entrada Sandstone is the Dewey Bridge Member. The Dewey Bridge is a series of interbedded siltstones and shales, about 10 to 20 m thick. The Entrada Sandstone is overlain by interbedded sandstones, cherts, and shales of the Cretaceous Morrison Formation.
Joints in the Entrada Sandstone appear to be related to the salt-cored Salt- and Cache-Valley structures, and do not reflect a regional pattern (Kelley & Clinton 1960; Doelling 1988). These joints are approximately parallel to the axis of Salt Valley; near Fiery Furnace, on the northeast side of Salt Valley, they change orientation and become parallel to Cache Valley. On the northeast flank the joint spacing increases away from Salt Valley. Throughout most of the Salt Valley area, joints are almost vertical, while bedding may be inclined up to about 15°. Although the pattern is related to the local structure, the joints are not surficial features, since they intersect the current topography (Dyer 1983). It is important to realize that Salt Valley Anticline was well established at the time the Entrada Sandstone was deposited, and that fractures in the Entrada Sandstone can represent only part of the history of the anticline.
The study area is located in the central portion of the southwest flank of Salt Valley Anticline, at the NW corner of Arches National Park (Fig. 1a). The area is marked by a NW-striking dip-slope of the Entrada Sandstone, the gentle dip being about 7°-15° to the SW. The ground surface is essentially the stripped upper surface of the Moab Member, which forms a NE-facing cuesta overlooking Salt Valley (Fig. 1c). The axis of Salt Valley Anticline in this area strikes about N45°W.
Within the mapped area, all proto-joints are within 10° of vertical, so mapped traces appear as relatively straight lines. Each line in Fig. 4, however, represents the trace of a zone of joints (Hodgson 1961; Dyer 1988; Cruikshank et al. 1991). In their simplest forms, the traces are composed of en échelon proto-joint segments, generally stepping a few cm to m, that are misaligned up to 10° from the trace of the proto-joint zone. Individual segments range from a few metres to tens of metres in length. The zones are segmented in both a lateral and vertical direction.
A joint is a fracture that is developed in opening-mode (mode I) (Pollard & Aydin 1988). Thus, although a joint forms in response to pressure in a crack or to far-field tension, it propagates as a result of high tensile stresses developed at its tip.
One of the most important principles in joint formation is that joints propagate in the direction of maximum compression. The importance of high compression was documented in detail with numerous field examples for sheet structures, a type of horizontal joint, generally in granitic rocks, by Holzhausen (1977; 1989). In his words, "many sheeted rock masses are under high in situ compressive stress parallel to the ground surface, as evidenced by rockbursts, expansion of blocks upon quarrying, and in situ stress-relief measurements. No other stress states have been detected for sheeted rock masses. Such a state of stress is apparently responsible for the formation of sheet structure" (1977, p. 120). His conclusion was verified theoretically by Cotterell & Rice (1980), who show that, even if perturbed, a mode I fracture has a strong tendency to reorient to become parallel to maximum far-field compression.
A corollary of this principle is the idea that joints with different orientations formed under different stress states. Further, the notion of "conjugate joints" is unfounded. Accordingly, since we recognize three distinct sets of joints in the SW flank of Salt Valley Anticline, the joints must represent three different stress states, presumably of different ages.
There are several features that can be used to discern the relative ages of joints, namely abutting relationships, joint interactions, and secondary joints due to shear displacement on existing joints (e.g. ; Twiss & Moores 1992, Fig 3-10). These patterns are summarized in Fig. 5.
(Full GIF image: 24 Kb)
Fig. 5. Relationships used to determine relative ages of joints. In (a) the younger joint turns towards the older joint and becomes either parallel- or perpendicular-to the older joint. In (b) the joints are of a similar age, and produce a characteristic hook-shaped geometry. In (c) the younger joint is propagating normal to the plane of view. In this case the younger joint breaks down into a series of en échelon segments that turn towards the older joint. In (d) the younger joints are a result of left-lateral shear displacement on the older joint. These tail, or horsetail joints initiate from the tips of older joints, or from roughness elements along the length of the older joint (see Fig. 8). Similar patterns and relationships are discussed by Dyer (1988), and Twiss & Moores (1992, Fig 3-10 p. 42).
A propagating joint will re-orient itself to stay in the plane of minimum compression. A joint propagating as an opening-mode fracture will continue to propagate in its own plane unless the tip is subjected to some shearing. Thus, if a joint curves or changes direction, the joint tip was subjected to shearing. Joints will deviate from straightness either during formation of the joint (e.g., Figs. 5a-c), or in response to some changed stress state at some time after the joint formed (e.g., Fig. 5d).
When a joint's orientation is changed by its propagating into a steadily-changing stress field, the joint will gradually change orientation; the fracture is said to veer. Veering may result from stress perturbations produced either by other near-by fractures, or by bedding interfaces. Thus, if one fracture curves towards a sensibly straight fracture, then locally the curved fracture is younger than the straight fracture (Fig. 5a). This interaction may be the curving parallel or curving perpendicular type (e.g. ; Dyer 1988), depending on the angle that the fractures approach each other (Thomas & Pollard 1993). If both joints veer, producing a double hook-shaped geometry (Fig. 5b), they were mutually interacting and are of similar age (Kulander et al. 1979; Pollard and Aydin 1988; Olson & Pollard 1989).
In cases where a fracture is propagating normal to the plane of view, fracture terminations with series of en échelon fractures near the tips probably represent a breakdown fringe (or gradual twist hackle (Kulander et al. 1979)) on the parent fracture, the tip of which is above or below our observation plane (Fig. 5c) (Pollard et al. 1982; Cruikshank et al. 1991, fig. 24). Where such a fringe exists next to a sensibly straight fracture, and joint segments that make up the fringe turn towards the straight fracture, the fracture associated with the fringe is probably locally younger (Cruikshank et al. 1991; Twiss and Moores 1992, Fig 3-10).
A proto-joint trace that abruptly changes orientation (i.e., a kink or wing fracture, Fig. 5d) forms in two-stages: first, a parent joint forms and, as the parent joint is later sheared, the tip may propagate at an abrupt angle to the parent fracture. Kinks are a diagnostic feature of faulted-joints. The kink angle gives the sense of shear on the older, parent joint (Cotterell and Rice 1980; Cruikshank et al. 1991). A variation on the kink is a kink-like fracture that initiates from irregularities along the length of the fracture, called horsetail fractures (Fig. 5d) or pinnate joints (Price 1966; Engelder 1989).
The configurations described above provide a basis for establishing the relative age of most interacting or intersecting fractures. Using these rules we find a consistent relationship between proto-joints of different orientations in the Entrada Sandstone. Thus we can determine the relative age relationships between proto-joint sets, and understand complex-looking fracture patterns, such as that shown in Fig. 4, in terms of the sequential development of joint sets.
In the mapped area (Fig. 4) there are three sets of proto-joints. Each set has a definite spatial distribution and orientation. The limited spatial extent of some joint orientations allows us to define four proto-joint domains, a domain being an area of outcrop with a well-defined proto-joint pattern, usually dominated by single proto-joint set. The three proto-joints domains within the Moab Member are: Klondike Bluffs domain (North of 8600N, Fig. 4), Anniversary Arch domain, (West of 2000W, and between 5000N and 8600N, Fig. 4), and the Tower Arch domain. (East of 2000W and south of 6000N, Fig. 4). The transition between the Tower Arch and Anniversary Arch domains (5000N 1800W to 6000N 2200W, Fig. 4, Fig 9) is a 400 m wide zone containing both NE-striking and NW-striking proto-joints. The transition between the Klondike Bluffs and Tower Arch domains (8400N to 8800N, Fig. 4, Fig 7) is defined by the southern termination of a set of North-striking proto-joints. The upper and lower boundaries of domain in the Moab Member are, respectively, the contacts with the Morrison Formation and the Slickrock Member. There is one joint domain in the Slickrock Member, and it is found throughout the mapped area. The upper-boundary of this domain is the Moab-Slickrock contact.
The proto-joint network shown in Fig. 4 was preceded by the development of a network of deformation bands. Jointing started with the development of a NE-striking proto-joint set at the northern (Klondike Bluffs domain) and southern (Anniversary Arch domain) ends of the mapped area within the Moab Member. The area between these two domains (Tower Arch domain) was then filled with a NW-striking set, again confined to the Moab Member. NW-striking joints then formed in the Slickrock Member underling the entire area. NW striking joints in the Slickrock have propagated upwards into the Klondike Bluffs and Anniversary Arch domains. These stages in the evolution of the fracture pattern are discussed in more detail below.
The oldest fractures in the area are deformation bands, which are thin, segmented tabular zones of deformed sandstone across which there is a small amount of shear offset [Aydin, 1978 #30; Aydin, 1978 #31; Aydin, 1983 #32; Johnson, in prep. #3615; Antonellini, in press #3797]. Deformation bands are present throughout the study area. Traces of individual deformation bands are irregular, as it the spacing between bands. There are at least two orientations of bands with a similar strike-about N55°-75°E-but an opposite dip direction. Offsets across bands is oblique- or normal-shift, as opposed to strike-shift displacements observed in the Garden area 3 km to the south (Zhao and Johnson 1991; 1992). Thus, the earliest recorded deformation is that of maximum compression in a plane striking about N65°E.
The presence of deformation bands has not affected the development of systematic joints except where a slip surface has formed along a zone of deformation bands. As documented in the Garden area, joints formed locally along zones of deformation bands (Zhao and Johnson 1992). In the present study area jointing along zones of deformation bands followed formation of the systematic joint set; they terminate against members of the systematic sets shown in Fig. 4. Most E-W grooves seen in Fig. 3 represent weathering along joints formed within zones of deformation bands. Thus, these E-W grooves give a rough indication of the spacing and distribution of deformation bands in this area.
The northern boundary of the mapped area is at the Klondike Bluffs fault, a normal fault with down-throw to the north. The maximum fault offset, up to 20 m, is approximately equal to the thickness of the Moab Member in this area. Displacement on the Klondike Bluffs fault decreases from east to west. The displacement is about 20 m at about 8500N 2300W (Figs. 2 & 7), while 500 m further to the west, the displacement is only a few cm.
The oldest proto-joint set in the mapped area has a strike of about N5°-15°E (Fig. 6a), and is confined to the Moab Member of the Entrada Sandstone. The northern area of the NE-striking joints is the Klondike Bluffs domain, and the southern area is the Anniversary Arch domain. There are a few isolated segments of NE-striking proto-joints about 1 km south of the Klondike Bluffs fault (Fig. 6a). Proto-joints in each of these areas probably formed around the same time and under the same stress state, since they have similar orientation in each area and are, throughout, older than NW-striking joints.
(Full GIF image: 69 Kb)
Fig. 6. Three jointing stages identified in the study area. A period of deformation recorded in deformation bands and movement on the Klondike Bluffs fault preceded the sequence of jointing shown here. (a) Joints in the Moab Member of the Klondike Bluffs and Anniversary Arch domains are the oldest set. (b) Joints of the Tower Arch domain form in the Moab Member. (c) NW striking joints form in the Slickrock Member. These joints formed in response to a slightly different stress field from those in the overlying Moab Member, and nucleated from the lower edge of joints in the Moab Member. After formation of joints in the Slickrock domain, some blades grew upwards into the Moab Member, producing the NW-striking cross-joints. Figure 4 shows the result of the superposition of all these events.
Proto-joints in the vicinity of the Klondike Bluffs fault (Fig. 7) strike about N7°E, almost normal to the Klondike Bluffs fault. In the vicinity of 8750N 3400W, the joints intersect the fault slip surface; about 250 m to the east, however, the joints end within the zone of deformation bands near the fault, 10 to 20 m before the fault surface (Fig. 2). Near the fault, where the NE set does not cross it, there is a series of joints with a polygonal-like pattern. Individual polygons are about 5 m across. Proto-joints on the northern side of the fault in the Moab Member have a pattern similar to that just south of the fault (Fig. 3).
(Full GIF image: 110 Kb)
Fig. 7. Map showing joint traces in Entrada Sandstone in the Klondike Bluffs domain. All the joints shown cut pre-existing deformation bands. Some deformation bands are offset where joints have become faulted. Many joints striking NW are confined between NE-striking joints, and show curving-parallel or curving-perpendicular relationship to NE-striking joints. At the southern termination of some NE-striking joints there are a few NW oriented tail fractures. These relationships indicate that the NW-striking joints are younger than the NE-striking joints in the Klondike Bluffs region. Both orientations of joints show left-lateral offsets. The NW-striking joints are of the correct orientation to be related to shear on the NE-striking joints. Stippled areas are sand covered.
The southern terminations of NE-striking joints in the vicinity of the Klondike Bluffs fault are abrupt. There are a few that show en échelon steps or kinks in the tip region. Several NW-striking kink-like fractures can be seen at the southern termination of a NE-striking joint at the west side of the blank area in Fig. 7. Some NE-striking joints appear to change orientation gradually (e.g. 8370N 2950W), and become parallel joints in the Tower Arch domain. This change is actually accomplished by a series of en échelon stepping segments.
Northwest-striking proto-joints form two domains, the Tower Arch domain in the Moab Member, and the Slickrock domain. Joints in the Slickrock are consistently younger than through-going proto-joints in the overlying Moab Member. Within the Moab Member, the NW-striking proto-joints are younger throughout than the NE-striking proto-joints.
The pattern of jointing in around the isolated NE-striking proto-joints 1 km south of Klondike bluffs confirms that the NW-striking joints are younger. This pattern is diagnostic. Multiple nucleation points along their length suggest that the fracture walls were being held closed at the time the fringe fractures formed. This is consistent with the notion that zones of joints formed with high compression parallel to the joints, minimizing interaction between joint segments (Holzhausen 1977; Cotterell and Rice 1980; 1989; Olson and Pollard 1989; Cruikshank et al. 1991).
(Full GIF image: 77 Kb)
Fig. 8. Map of a portion of Fig. 4 showing details of tail and horsetail fractures formed along NE trending joints. The NE-striking joints were sheared in a left-lateral sense. This indicates that the NE-striking joints are older than the NW-striking joints.
Joint patterns at the boundary between the Tower and Anniversary Arch domains (Fig. 9) again show that NW striking joints nucleated from the older NE-striking set. NW-striking joints of the Tower Arch domain show left-lateral offsets, so the NE-striking, kink-like joints at the southern tips of these joints cannot be related to this shear offset. The numerous NW-striking tail and horsetail fractures on the NE joints are consistent with the left-lateral slip observed in the NE joints.
(Full GIF image: 167 Kb)
Fig. 9. Map showing details of the transition between the Tower Arch and Anniversary Arch domains. Proto-joints of the Tower Arch domain nucleate from the older proto-joints in Anniversary Arch area. At coordinate 5200N on the eastern margin of the map, the Slickrock Member of the Entrada Sandstone is exposed. Joints in the Slickrock are oriented about N20°W, while joints in the Moab Member strike about N10°E.
The NW joint set shows either no offset, or left-lateral offset of pre-existing deformation bands. In one location (6750N 3400W), on a joint oriented N15°W 80°E, the displacement vector was 5.5 cm in a direction 36°/N8°W. Thus the joint was faulted with a combination of 4.8 cm left-lateral and 2.5 cm normal offset (down towards the east). Offsets along joints in the Slickrock were not observed.
Because there is only one joint domain in the Slickrock Member, spectacular transitions-such as between the Tower Arch and Anniversary Arch domains (Fig. 9)-are seen only in the Moab Member. Joints in the Slickrock Member do not show an abrupt change in orientation in this area. Whereas joints in the Slickrock domain are rotated only a few degrees counter-clockwise with respect to joints in the Tower Arch domain, they are at a 35° angle to joints in the Anniversary Arch domain.
Throughout the mapped area, the orientation of joints in the Slickrock Member differ from that of joints in the overlying Moab Member. Strikes of joints in the underlying Slickrock Member are rotated up to 35° counter-clockwise with respect to joints in the Moab Member in the Anniversary Arch domain, and 10° counter-clockwise with respect to joints in the Tower Arch domain (a distance of about 2.5 km). Joints in the Slickrock Member are younger than those in the overlying Moab Member. In the northern part of the area (7000N, Fig. 4) the Slickrock joints strike about N30°W; however, in the vicinity of Anniversary Arch, the set strikes about N15°W. Thus there is a 15° counter-clockwise rotation of the strike of joints moving from south to north. Over the same distance, proto-joints in the overlying Moab Member (Tower Arch domain) show a 7° clockwise change in orientation.
Joints in the Slickrock domain have nucleated from the lower edge of joints in all domains in the overlying Moab Member. In places where the contact is gradational, joint surfaces in the Moab Member can be followed continuously into joint surfaces of the Slickrock domain. Surfaces in the Slickrock can also be followed, without a change in orientation, back into the Moab Member as NW-striking cross-joints (Cruikshank in review). Where the contact between members is abrupt-usually where there is a thin shale layer-joints in the Moab Member do not penetrate into the Slickrock Member; rather, they end abruptly.
In the Klondike Bluffs domain (Figs. 2 & 7), northwest-striking proto-joints within the Moab Member are confined between the older northeast-striking joints. They usually curve to become normal with the NW-striking joints, or break down into a series of echelon cracks. They do not intersect the northeast-striking proto-joints. This pattern continues well to the north of the Klondike Bluffs fault, and is probably not related to the transition between two domains. This persistent set of NW-striking cross-joints is explained by the upward-propagation of NW-striking joint segments from the underlying Slickrock Member.
The fracture pattern on the southwest limb of Salt Valley Anticline evolved in a patchwork fashion, both laterally within the same unit and vertically between units. There are three proto-joints in the Moab Member, with the fourth domain in the Slickrock Member. Domain boundaries could not be related to structural boundaries. Nonetheless, domains in the Moab are separated from that in the Slickrock by lithologic variations, such as a shale interbed or a gradational change in sandstone composition.
The formation of one proto-joint set influenced the development of subsequent sets, either by serving as a nucleus for new fractures, or by limiting the spatial extent of subsequent fractures. Interactions between proto-joints in each domain allowed the determination of the following sequence of deformations:
The fractures within the Entrada Sandstone on the southwest limb of Salt Valley Anticline record a counter-clockwise rotation in direction of principal compression, from about N65°E to N30°W - a 95° counterclockwise rotation. Previous studies in the Garden area, 3 to 4 km south of the present study area, recorded a counter-clockwise rotation of about 50° (Zhao and Johnson 1992).
In summary, a network of joints on the southwest limb of Salt Valley Anticline is composed of several overlapping domains of differently-oriented joints. The network evolved both laterally and vertically in a patchwork manner through a series of stages identified by establishing the relative ages of joint domains. The development of one joint domain affected the development of a later one, either by providing a nucleus for new fractures, or by limiting the lateral extent of new fractures. Within the Moab Member of the Entrada Sandstone are three major overlapping joint domains. The strike of joints may change between domains by as much as 35°. Joints in northern and southern domains strike about N5°-10°E, while joints in the central domain strike about N10°-20°W. Where joint domains overlap, interaction between members of each domain indicates that joints in the northern and southern domains are older than those in the central domain. Joints within all three domains have become faulted, and offset pre-existing markers in a left-lateral sense. There is one joint domain in the underlying Slickrock Member, which extends under all three domains in the Moab Member. Joints in the Slickrock domain nucleated from the lower edges of joints in all three domains of the Moab Member.
This work was supported by U.S. Department of Energy, National Science Foundation, and industrial affiliates of the Rock Fracture Project. Thanks to Marco Antonellini for assistance in the field. Marco and Emanuel J.M. Willemse provided reviews of an early version of the manuscript. Discussions with Arvid Johnson were many and fruitful. James Gardner edited the manuscript. Jim Evans, George Davis and Byron Kulander provided constructive reviews for the Journal of Structural Geology.
Aydin, A. 1978. Small faults formed as deformation bands in sandstone. Pure and Applied Geophysics 116, 913-930.
Aydin, A. & Johnson, A. M. 1978. Development of faults as zones of deformation bands and as slip surfaces in sandstones. Pure and Applied Geophysics 116, 931-942.
Colman, S. M., Choquette, A. F. & Hawkins, F. F. 1988. Physical, soil, and paleomagnetic stratigraphy of the Upper Cenozoic sediments in Fisher Valley, southeastern Utah. Professional Paper United States Geological Survey 1686.
Cotterell, B. & Rice, J. R. 1980. Slightly curved or kinked cracks. International Journal of Fracture 16, 155-169.
Cruikshank, K. M. in review. Möbius joints. Journal of Structural Geology
Cruikshank, K. M., Zhao, G. & Johnson, A. M. 1991. Analysis of minor fractures associated with joints and faulted-joints. Journal of Structural Geology 13, 865-886.
Cruikshank, K. M., Zhao, G. & Johnson, A. M. 1991. Duplex structures connecting fault segments in Entrada Sandstone. Journal of Structural Geology 13, 1185-1196.
Davis, G. H. 1984. Structural geology of rocks and regions. New York, John Wiley & Sons.
Doelling, H. H. 1985. Geology of Arches National Park, with accompanying text. Map Utah Geological and Mineral Survey 74.
Doelling, H. H. 1988. Geology of Salt Valley anticline and Arches national park, Grand County, Utah. In: Salt deformation in the Paradox region (edited by Doelling, H. H., Oviatt, C. G. & Huntoon, P. W.). Utah Geological and Mineral Survey, Salt Lake City, Utah. 1-58.
Dyer, J. R. 1983 Jointing in sandstones, Arches National Park, Utah. Ph.D. thesis, Department of Geology, Stanford University, Stanford, California.
Dyer, J. R. 1988. Using joint interactions to estimate paleostress ratios. Journal of Structural Geology 10, 685-699.
Elston, D. P., Shoemaker, E. M. & Landis, E. R. 1962. Uncompahgre front and salt anticline region of Paradox basin, Colorado and Utah. Bulletin of the American Association of Petroleum Geologists 10, 1857-1878.
Engelder, T. 1989. Analysis of pinnate joints in the Mount Desert Island granite: Implications for postintrusion kinematics in the coastal volcanic belt, Maine. Geology 17, 564-567.
Hodgson, R. A. 1961. Regional study of jointing in Comb Ridge-Navajo Mountain area, Arizona and Utah. Bulletin of the American Association of Petroleum Geologists 45, 1-38.
Holzhausen, G. R. 1977. Axial and subaxial fracturing of Chelmsford Granite in uniaxial compression tests. Proceedings of the 18th U.S. Symposium on Rock Mechanics
Holzhausen, G. R. 1989. Origin of sheet structure, 1. Morphology and boundary conditions. Engineering Geology 27, 225-278.
Kelley, V. C. & Clinton, N. J. 1960. Fracture systems and tectonic elements of the Colorado Plateau. Publications in Geology University of New Mexico 6.
Kulander, B. R., Barton, C. C. & Dean, S. L. 1979. The application of fractography to core and outcrop fracture investigations. United Stated Department of Energy, Morgantown Energy Technology Center.
Olson, J. & Pollard, D. D. 1989. Inferring paleostress from natural fracture pattern: A new method. Geology 17, 345-348.
Oviatt, C. G. 1988. Evidence for Quaternary deformation in the Salt Valley anticline, southeastern Utah. In: Salt deformation in the Paradox region (edited by Doelling, H. H., Oviatt, C. G. & Huntoon, P. W.). Utah Geological and Mineral Survry, Salt Lake City, Utah. 61-78.
Pollard, D. D. & Aydin, A. 1988. Progress in understanding jointing over the past century. Geological Society of America Bulletin 100, 1181-1204.
Pollard, D. D., Segall, P. & Delaney, P. T. 1982. Formation and interpretation of dilatant echelon cracks. Geological Society of America Bulletin 93, 1291-1303.
Price, N. 1966. Fault and joint development in brittle and semi-brittle rock. Oxford, Pergamon Press.
Shoemaker, E. M., Case, J. E. & Elston, D. P. 1958. Salt anticlines of the Paradox basin. In: Guidebook to the geology of the Paradox basin (edited by Sandborn, A. F.). 39-59.
Thomas, A. L. & Pollard, D. D. 1993. The geometry of échelon fractures in rock: Implications from laboratory and numerical experiments. Journal of Structural Geology 15, 323-334.
Twiss, R. J. & Moores, E. M. 1992. Structural geology. New York, W.H. Freeman and Company.
Zhao, G. & Johnson, A. M. 1991. Sequential and incremental formation of conjugate sets of faults. Journal of Structural Geology 13, 887-895.
Zhao, G. & Johnson, A. M. 1992. Sequence of deformations recorded in joints and faults, Arches National Park, Utah. Journal of Structural Geology 14, 225-236.
