Evidence for pre-Cenozoic extension in the eastern Main Ranges of the southern Canadian Rockies

The Cordillera of western North America comprises a collage of terranes that have accreted onto the North American tectonic plate (Fig. 1; Monger et al., 1982). Crustal features included in this collage of terranes are peri-Laurentian terranes that are interpreted to be slivers of North American crust that rifted off the continent during the opening of the Slide Mountain back-arc basin in Late Devonian to early Carboniferous time (Fig. 1; Colpron and Nelson, 2009). Radiometrically dated alkaline igneous bodies and bimodal volcanics that occur along the Cordillera mark the onset of rifting (Colpron and Nelson, 2009). By the end of the Triassic Period, the Intermontane superterrane had accreted back onto North America with the closure of the Slide Mountain Ocean (Fig. 1; Colpron and Nelson, 2009). An inferred suture that formed as a result of the closure of the ocean is marked by discontinuous slivers of ophiolitic rocks of the Slide Mountain terrane (Fig. 1; Nelson et al., 2013).

Cratonward of the terranes, there is a belt of contracted North American strata (Fig. 1): The southern Canadian Rocky Mountain thrust-and-fold belt is a tectonic retrowedge that was formed through convergence between the accreted terranes and the North American tectonic plate (McMechan et al., 2018, and references therein). The belt, located along the eastern margin of the North American Cordillera (Figs. 1 and 2), experienced protracted contraction from the Late Jurassic Period of the Mesozoic Era into the Eocene Epoch (Fig. 3; McMechan et al., 2018, and references therein). Within this collisional setting, a network of normal faults has been identified and mapped in the eastern Main Ranges of the southern Canadian Rockies (Fig. 2, indicated as blue comb lines; Price and Mountjoy, 1970; Cook, 1975; Price, 1981, 2000, and references therein). These normal faults have been postulated to flatten at depth and merge smoothly with older thrust faults (Price and Mountjoy, 1970; Cook, 1975). The development of the normal faults is inferred to have happened during the major Eocene to Oligocene transtensional episodes (Fig. 3) that occurred along and west of the Rocky Mountain Trench (Fig. 2; Price and Mountjoy, 1970; Fermor and Moffat, 1992; Price, 1994). The normal-sense dip-slip movement was proposed to have been made possible by the reactivation of thrust faults (Price and Mountjoy, 1970; Cook, 1975; Price, 2000). An earlier investigation in this study area (Fig. 2), however, found evidence indicating that these normal faults developed prior to a contractional episode (Cook, 1975). The evidence supporting early normal fault development published by Cook (1975) is reviewed in the “Regional Geology” section below. This study examined the nature and timing of several normal faults located within the eastern Main Ranges (Figs. 2 and 4). The characteristics of the faults indicate that the extensional episode during which the normal faults developed occurred prior to contractional episodes (Fig. 3).

The study area lies within the Prophet Trough, a linear basin paralleling the western edge of the craton that contains relatively thick Carboniferous and Permian strata (Richards et al., 1994; Henderson et al., 1994). The development of the basin was related to the rifting and subsequent opening of the Slide Mountain Ocean and the Antler foreland basin of the western United States (Fig. 1; Richards et al., 1994). Located only 10–20 km south of the study area (Fig. 2), the presence and radiometric age of the Ice River complex point to a period of rifting coincident with the opening of the Slide Mountain Ocean. The Ice River complex is an alkaline igneous body that has been dated to ca. 360 Ma (Parrish et al., 1987; Mumford, 2009). Root (2001) interpreted deformation at the Delphine Creek area (Fig. 2; ~100 km south of study area) as an early phase of the Antler orogeny. This study proposes that a common cause led to the development of the network of normal faults within the study area, the development of the Prophet Trough, and the formation of the Ice River complex; each is a product of and expression of the Slide Mountain back-arc basin. The subsequent contractional phases (Fig. 3; Jurassic–Paleogene Periods) then inverted the rifted margin and carried the network of normal faults northeast to its present location, over unattenuated crust. Inversion of rifted margins is an outcome that has been identified throughout many parts of the world (Lowell, 1995). The structural restoration of the faults to their location of origin yields insight into the tectonics that affected the extensional area.

Major changes in stratigraphy and structural style occur across the Main Ranges of the southern Canadian Rocky Mountains (Fig. 2). To the northeast in the eastern Main Ranges (EMR in Fig. 2), open folds of platform-margin carbonates of the Bow Platform (Norford, 2012) are carried on a few major thrust sheets (Price and Mountjoy, 1970; Cook, 1975; Price, 1981, 2000). The thrust sheets are dissected by a network of normal faults that trend north to north-northwest, oblique to the regional trend (Fig. 2). To the southwest in the western Main Ranges (WMR in Fig. 2), basinal and slope successions of argillaceous deposits of the White River Trough (Norford, 2012) exhibit ductile and penetrative deformation (Cook, 1975). The western Main Ranges and eastern Main Ranges are separated by the Kicking Horse Rim, a paleo–platform margin of Laurentia (Fig. 2; Aitken, 1971). The Kicking Horse Rim marks the break between the outer edge of a shallow-marine carbonate bank, which had subtle paleotopographic relief, and the deeper marine slope (Aitken, 1971). The carbonate bank margin lies along a 130 km, continuous, slightly arcuate belt that parallels the paleomargin of Laurentia (Fig. 2). The location of the carbonate bank margin shifted throughout the Cambrian Period but did not deviate by more than 16 km (Aitken, 1971). The Kicking Horse Rim is interpreted to have a tectonic origin, and biogenic processes may have played an important role in maintaining the structure (Aitken, 1971; Root, 1987; Collom et al., 2009; Johnston et al., 2009).

The Cathedral Crags anticline, a large-scale open anticline, persists along the length of the Kicking Horse Rim (Fig. 2; Cook, 1975). To the southwest of the Kicking Horse Rim, bedding attitude steepens sharply. Coincident with the steepening of bedding, there is a zone, <10 km wide, of complex folding and faulting of variable stratigraphy that parallels the Kicking Horse Rim over its entire length (Cook, 1975; Stewart, 1991; this study). The complex deformation of the southwest-dipping strata is interpreted to be the result of compression of weak argillaceous rocks against the stronger carbonate platform rocks (Cook, 1975). Mega-truncation surfaces that are inferred to represent the proximal portions of marine gravity-slide scars are observed along this zone and indicate that the southwest limb of the Cathedral Crags anticline was once a marine slope environment (Stewart, 1991; Stewart et al., 1993). Previously interpreted as tectonic normal faults, these slide surfaces have been distinguished from tectonic faults by detailed field mapping (Cook, 1975; McIlreath, 1977; Stewart, 1991). The slide surfaces tend to be sharp and listric in form; the upper body onlaps the truncation surface and often contains large olistoliths (Stewart, 1991). The Walcott Quarry, in which the Burgess Shale fossils are found, is located midway between Mount Field and Wapta Mountain (Fig. 4); the fossil site lies adjacent to the Cathedral Escarpment, the headwall scarp of a Cambrian-age marine slide (Stewart, 1991).

Farther west, within the western Main Ranges and parallel to the Kicking Horse Rim, there is the Porcupine Creek anticlinorium, a large-scale, 350-km-long, 35-km-wide, arcuate, asymmetric, fan-shaped anticlinorium (Fig. 2; Balkwill, 1972; Price, 1986). The formation of the fan has been proposed to have occurred as a consequence of contraction above a detachment zone that lies immediately above the thick Lower Cambrian Gog Group orthoquartzite (Balkwill, 1972; Price, 1986). For the purpose of this paper, a detachment surface or detachment zone is defined as the surface between two adjacent blocks of rock that have moved relative to each other along a bedding-parallel surface or zone. The northeastern margin of the anticlinorium passes through the southwest quadrant of the study area (Figs. 2 and 4), and, in this location, it contains two sets of cleavage, suggesting that the area experienced two distinct episodes of contraction (Balkwill, 1972). The earlier set dips moderately to the southwest (25°–40°) and is truncated and offset by the younger, subvertical cleavage set.

The network of normal faults that is the focus of this study (Fig. 2) consists of subparallel, en echelon, left-stepping faults with dip displacements rarely exceeding 1 km (Cook, 1975; Price, 2000). Summed across the study area, the network has a net horizontal extension of <2 km (estimated from Cook, 1975, his cross-sections A–H). In several locations, the sense of net displacement observed along the surface trace of a fault changes between reverse and normal (Price and Mountjoy, 1970), demonstrating the polyphase nature of the faults, but not the order of movement nor the number of distinct phases of tectonism. In other cases, the surface trace of a fault branches into a normal segment and a reverse segment (see, for example, Price, 2005; Price and Mountjoy, 1978), again demonstrating a history of multiple tectonic episodes but not necessarily the order. The network of normal faults in the northwest quadrant of the map (Fig. 2) is confined to the eastern Main Ranges and is bounded to the west by a facies change west of the Kicking Horse Rim. The north- to north-northwest–striking normal faults can be traced toward the south, where the direction of strike curves to the southeast as the fault surfaces merge with the Cathedral Crags anticline and dip-slip displacement along the faults decreases to zero (Cook, 1975).

Seventy kilometers northwest of the study area, the final stage of a latest Neoproterozoic to earliest Cambrian rifting episode is marked by outcropping syndepositional marker beds cut by normal faults (Fig. 2, NW corner; Lickorish and Simony, 1995). In the study area and in the eastern Main Ranges neighboring the study area, thicknesses of units younger than this rifting episode do not change across the normal faults (see maps and accompanying cross sections of: Price and Mountjoy, 1978; Price et al., 1980; Price, 2005; Cook, 1975; this study). Therefore, significant displacement along the normal faults occurred after the deposition that occurred in the Cambrian Period. This observation does not preclude minor or isolated development of faulting during the Cambrian Period.

The Emerald Pass fault (Fig. 4), a normal fault of particular interest for this study, shows an apparent, progressive decrease in stratigraphic separation along the fault from north to south. The fault crops out at Emerald Pass (Fig. 4), where the fault shows stratigraphic separation of ~600 m (Cook, 1975, his section C), while farther south at Burgess Pass, the separation is ~300 m (Fig. 4; Cook, 1975, his section D), and the stratigraphic separation diminishes until the fault tips out where it crosses Dennis Pass (Fig. 4).

In the eastern Main Ranges neighboring the study area, all strata younger than early Carboniferous have been eroded away (Richards et al., 1994), and evidence that establishes the timing and development of the normal faults is incomplete. Both the Amiskwi Pass–Emerald Pass fault system and the Stephen-Cathedral fault (Fig. 4) continue north of the study area, where outcropping stratigraphic separation across the faults juxtaposes Upper Devonian strata against Ordovician- and Cambrian-age strata (Price, 2005). Farther north, within the footwall of the Simpson Pass thrust, stratigraphic separation across the normal faults places lower Carboniferous strata against Cambrian strata (Fig. 2, north border; Pana and Elgr, 2013). Based on limited evidence, the network of normal faults was active at least as late as earliest Carboniferous time. Basinward of the Kicking Horse Rim, and across the 35 km width of the Porcupine Creek anticlinorium to the Rocky Mountain Trench, evidence of extension is absent at the latitude of the study area (Fig. 2). The Porcupine Creek anticlinorium separates the Rocky Mountain Trench from the network of normal faults located in the eastern Main Ranges (Fig. 2; Cook, 1975; Balkwill, 1972).

In regions of the southern Canadian Rockies neighboring the study area, fault properties indicate complicated tectonic histories. The Redwall fault is located ~120 km south of the study area and adjacent to the Rocky Mountain Trench (Fig. 2, lower center). The complex nature of the fault indicates that it has experienced multiple episodes of deformation (McMechan et al., 2018, and references therein). Farther south along trend, ~25 km north of the U.S.–Canada border, the Howell Creek structure (Fig. 2, lower right) has been interpreted to have gone through alternating episodes of contraction and extension prior to movement on the Oligocene-age extensional Flathead fault (Stockmal, 2018). The findings of these recent studies suggest that the tectonic history of the southern Canadian Rockies involved multiple episodes of contraction and may have included an earlier episode of extension. A straightforward tectonic history involving contractional episodes followed by a late extensional period may not be adequate to account for field observations.

Structural features mapped in the study area by Cook (1975) led him to interpret that a compressional event occurred after the normal faults had developed. The Ogre synclinorium, a large-scale (km+) structure, is found in the hanging wall of the Amiskwi Pass fault (Fig. 4, northwest quadrant; Cook, 1975, Geological Survey of Canada Map 1368, his Howse Pass fault). The magnitudes (amplitude and wavelength) of the folds in the adjacent footwall are much smaller, however (Fig. 4), and Cook (1975) inferred that the synclinorium and minor folds developed independently of each other, bounded by the preexisting normal fault.

Cook (1975) observed and measured slickensides on fracture surfaces adjacent to the Cataract Brook and Stephen-Cathedral normal faults (Fig. 4, red ovals; Cook, 1975, his Fig. 16). The slickensides indicate horizontal right-lateral strike-slip movement, even though the displacement along the normal faults is apparently dip-slip (at all observed bedding dips; Cook, 1975). Cook (1975) noted the connection between the steep normal faults and adjacent steep slickenside surfaces at the Cataract Brook and Stephen-Cathedral faults and the development of distinct fold structures on opposite sides of the normal fault at the Ogre synclinorium to the northwest (Cook, 1975). Based on these observations, he proposed that a compressional episode had occurred, resulting in movement along and across the preexisting normal faults.

The study area lies within a single large thrust sheet that is carried by the Simpson Pass thrust (Figs. 2 and 4; Cook, 1975). The Simpson Pass thrust is a large-scale, 275-km-long thrust (Price, 2000) that emerges in outcrop within the Miette Group strata at the northeast corner of the study area (Figs. 2 and 4). Two minor thrust faults (Wapta Mountain and Martin Creek thrusts) occur within the Simpson Pass thrust sheet and emerge at the surface within the study area (Fig. 4; Cook, 1975). The Martin Creek thrust has been interpreted as the surface expression of a detachment zone that carries the Porcupine Creek anticlinorium (Kubli and Simony, 1994). Near the southern termination of the Porcupine Creek anticlinorium, the Lucier River Stock, dated to 109 Ma, crosscuts and overprints structures and faults of the Porcupine Creek anticlinorium (Fig. 2; Larson et al., 2006), indicating that the Porcupine Creek anticlinorium had developed by Albian time and was “passively” carried by underlying thrusts during the Late Cretaceous to Eocene phase of contractional episodes (Fig. 3; McMechan et al., 2018). The Wapta Mountain and Martin Creek thrusts have been interpreted to be truncated and offset by normal faults (Fig. 4; Cook, 1975). Major thrust faults, such as the Simpson Pass thrust, however, do not show evidence of truncation or offset by the network of normal faults mapped in the vicinity of the study area (Fig. 4; Cook, 1975). Thrust-fault gouge of the Simpson Pass thrust, sampled 10 km north of the study area, showed a radiometric age of 161.7 Ma (Pana and van der Pluijm, 2015, on ⁴⁰Ar/³⁹Ar illite). Although the estimate suggests a Late Jurassic age, it is unclear what stage in the fault's development this age represents.

Field mapping was conducted with particular attention to locations that present stratigraphic and structural views of importance to the understanding of the geologic history of the area. Mapped geologic features were traced onto photographs and projected onto cross sections using locally derived structural trends. Mapping was facilitated by thin marker beds and a dense network of measured outcrops. The mapping results revealed stratigraphic and structural relationships in considerable detail and extent and facilitated a revised interpretation and reconstruction of an updated tectonic history of the region.

A brief introduction to the stratigraphy of the area is presented below, followed by the methods employed in this study. The greater part of this paper focuses on observations and an interpretation of the nature of the Emerald Pass fault at the west end of Burgess Pass.

A schematic stratigraphic section that is representative of the geology along the upper Kicking Horse valley was constructed by incorporating mapped stratigraphy with measured sections completed by previous workers (Fig. 5). The stratigraphic section illustrates the transition from a marine carbonate platform (to the NE) to a slope and deep-water argillaceous sequence (to the SW). A detachment zone has been interpreted above the kilometer-thick, mechanically strong orthoquartzite of the Gog Group (Balkwill, 1972). The mechanical stratigraphy (and deformation styles) of the units above the Gog Group changes abruptly at the Kicking Horse Rim (Fig. 5, annotated at top) as the thickness ratio of carbonate to shale increases platformward (Cook, 1975).

Geologic and accompanying positional data were collected during the new study from more than 80 day-long traverses completed during the field seasons of 2013–2021. The data were integrated with and positioned onto the Canadian Digital Elevation Model (DEM; 0.75″ 25 m spatial resolution; Natural Resources Canada, 2010) using the Global Mapper® v. 13.1.2 software program. The work of previous investigators, including more than 60 measured sections completed in the study area, was incorporated into the new mapping. Traverses for each measured section were hand-digitized and input into Global Mapper®. The initial structural and stratigraphic framework of the mapping was guided by the “A” series maps produced by the Geological Survey of Canada (GSC; Fig. 4). These publications are based on the findings of the GSC's 1960s Operation Bow–Athabasca project, which produced a coherent stratigraphic and structural framework of the southern Canadian Rockies (Sears et al., 2007; Gadd, 2012). The “A” series maps of the GSC were imported into Global Mapper®, and formation contacts, faults, and bedding and cleavage orientation measurements were hand-digitized. Interpreted photographs and sketches from previous investigators complemented the initial map.

A Garmin GPSMAP® 62st instrument produced a digital track log containing location (latitude and longitude), time, and date of each point (variable spacing averaging <10 m) along the traverses. The digital map in Global Mapper® was updated with the new field observations, and the integrated and updated digital map was exported from Global Mapper® to Google Earth Pro® in order to associate mapped features with satellite imagery and photographs taken along the traverses. The orientations of bedding and cleavage were measured using a Brunton GEO Transit; the mapping coordinates of the outcrop were key-entered into a Notepad file, which was subsequently imported into Global Mapper®. The elevations of geologic features, including the bedding orientation, digitized contacts, and faults, were derived from the DEM.

A high-resolution version of the map (Fig. 4) is included in the Supplemental Material¹. Eight parallel structure cross sections (locations indicated on the geologic map), oriented perpendicular to the structural and stratigraphic fabric of the region, were constructed for this study and can be found in the Supplemental Material. Further details on methods used in the course of this study, and measured data are included in the Supplemental Material.

Digital images of outcrops and surrounding landscape were obtained using a Canon D5 Mark II camera. The camera location for each image was determined by matching the date-time stamp attached to the digital image with the date-time stamp recorded on the global positioning system (GPS) in the corresponding track log of the traverse. Digital images were imported into Adobe Illustrator®, and evident features, such as formation contacts and faults, were hand-drawn (digitized) by inspection in the program. Where coverage of a photograph overlapped with the measured section of a previous worker, the interpreted geology drawn on the photograph was aided by the earlier interpretation. In some cases, the interpretation of less distinct geologic features was achieved with the use of additional magnified photos taken with a zoom lens. Where discrimination was unclear in the photo, interpretation of geologic features was inferred by overlaying the geologic map over the satellite imagery of Google Earth Pro®; the interpreted geologic feature was then hand-drawn onto the image (in Adobe Illustrator) based on information from the map and the overlay of the satellite imagery.

Lines representing the map location of each vertical cross section produced in this study were marked on the map in Global Mapper®, and the corresponding cross sections were produced in Adobe Illustrator®. Along each cross section, geologic information was extracted from Global Mapper®, with corresponding x, y, and z coordinates. The exported data were imported into Microsoft Excel®. The data points were projected onto the plane of the cross section using the local structural trend and appropriate trigonometric formulae. The projected points were then imported into Adobe Illustrator™ using the Scatter Graph Tool and plotted onto the artboard at an appropriate scale. Contacts and faults were hand-drawn in Adobe Illustrator® by connecting the projected points. Data points were collected in the field at a considerable range of elevations. Due to the local topographic relief (up to 1500 m), the data points were projected onto cross sections at a corresponding vertical range on the cross sections, and the relief was sufficient to resolve key geologic features with no vertical exaggeration. Apparent dips, along the line of section, were computed from field-measured orientations of bedding and cleavage. Dip bars were oriented and positioned onto projected points on the artboard. These dip bars were also posted onto photographs of outcrops that coincided with locations along the line of section.

Outcrops in the upper Kicking Horse region in Yoho National Park are predominantly of Cambrian sedimentary deposits that have experienced multiple phases of deformation (Figs. 3, 4, and 5). The study area spans the transition from the eastern Main Ranges (open folds of platform-margin carbonates) to the western Main Ranges (basinal and slope successions) of the southern Canadian Rockies. Along the eastern Main Ranges of the study area, the folds plunge very gently to the north-northwest, exposing the oldest strata to the south-southeast and younger strata to the north-northwest (Fig. 4, northeast half). The northwest-trending “stripes” of outcropping strata that extend from the southeast corner of the map to the upper-middle western border of the map trace the southwest-dipping limb of the Cathedral Crags anticline (Fig. 4, red ribbon), which lies on or near the Kicking Horse Rim (Cook, 1975). The striped band of dipping strata marks the northeast margin of the western Main Ranges (and White River Trough). The southwest boundary of the eastern Main Ranges (and Bow Platform) lies immediately northeast of the band, near the crest of the anticline. The normal faults in the vicinity of the study area are observed to cut down at least to the Miette Group (Figs. 4 and 5).

Targeted fine-scale field mapping (as fine as ~30 m spatial sampling) was conducted on the southeast slopes of Mount Burgess, the southwest slopes of Mount Field, and along Burgess Pass where the Emerald Pass fault outcrops (Fig. 6). A map, cross sections, interpreted photographs, and stereograms illustrating the outcropping geology are presented below. In the discussion that follows, the interpreted structural relationships and history of the study area emerge from the evidence presented in these illustrations. Observed structural and stratigraphic features are described starting at the outcropping Emerald Pass fault at the west end of Burgess Pass and systematically moving away from the fault to the northeast (footwall) and then returning to the fault and systematically moving downdip toward the southwest (hanging wall; Fig. 6).

The surface trace of the Emerald Pass fault traverses the southwest end of Burgess Pass and the village of Field (Fig. 6, center), following a generally north to south direction. Northwest of the pass, the surface trace of the fault follows a northwest trend. Good-quality exposure of the fault occurs on the northwest slopes of Burgess Pass (Fig. 6, north oval; Figs. 7A and 7B, fault marked by red-and-white striped line). Here, the fault surface is vertical to steeply dipping to the southwest (Fig. 6; Fig. 7A, marked with red-and-white striped line).

Within ~100 m northeast of the fault, the “basinal” Stephen Formation of the footwall exhibits intense penetrative deformation that has obliterated most traces of bedding (Fig. 7C). The term “basinal” Stephen Formation designates the thicker Stephen Formation unit found basinward of the Cathedral Escarpment (Aitken, 1997). The cleavage dips steeply to the southwest and appears to parallel the nearby fault surface. From ~100 to 200 m northeast of the fault, cleavage is prominent in outcrops, whereas bedding is rarely found. The observed cleavage is consistently vertical and strikes southeast-northwest. Some of the cleavage is deformed into anastomosing kink bands (Fig. 7D). Cook conducted an investigation of outcrops and thin sections from the study area in order to determine the nature of cleavage development (Cook, 1975). It was determined that development of penetrative cleavage is an indicator of significant shortening in the direction perpendicular to the cleavage planes (Cook, 1975). The intense subvertical cleavage adjacent to the Emerald Pass fault indicates that the shales of the “basinal” Stephen Formation have shortened and thickened significantly through penetrative deformation. The presence of chlorite specks in the “basinal” Stephen Formation (and other argillaceous units in the study area) indicates that the cleavage developed through pressure solution at the subgreenschist metamorphic grade (McIlreath, 1977; Aitken, 1997; Read et al., 1991).

The occurrence of cleavage within the “basinal” Stephen Formation at Burgess Pass progressively decreases beyond 200 m to the northeast of the fault. Bedding tends to be diffuse where it occurs, and the orientation of cleavage varies from vertical to dipping moderately to the southwest. Beyond ~1500 m northeast of the fault and along the southern slopes of Mount Field (Fig. 6), bedding is much more pronounced, and cleavage surfaces are minor elements compared to bedding.

Along the slopes of Burgess Pass, the more calcareous-rich beds of the “basinal” Stephen Formation form steep cliff bands that can be traced for more than 1 km (Fig. 6; footwall of Emerald Pass fault; e.g., Fig. 7B, left side). These cliff bands dip moderately to the southwest, and are subplanar and subparallel to each other, indicating that shortening and thickening of the unit were penetrative and homogeneous. Folding is minor, and no through-going thrust fault was found in the “basinal” Stephen Formation at Burgess Pass.

The intense shortening and thickening that occurred in the footwall of the Emerald Pass fault contrasts with the undeformed carbonates of the Tokumm Formation in the hanging wall, adjacent to the fault (Figs. 6 and 7E). Within ~500 m southwest of the fault, the carbonates in the hanging wall show very little deformation except for a minor fold near the fault and scattered occurrences of irregular-shaped, subhorizontal white veins. Here, the carbonates are strong and can sustain a subvertical cliff wall more than 500 m in height (Figs. 6 and 7E). Throughout the study area, argillaceous beds are observed to be strongly susceptible to cleavage development, whereas carbonate units exhibit only minor cleavage features (Cook, 1975; this study). The width of the fault zone is 1–10 m, and it has weathered significantly in many places, leaving a narrow depression that retains snow and ice, even into late summer (Fig. 7F).

From the location where the Emerald Pass fault crops out at the west end of Burgess Pass, the fault surface cuts down steeply though the Tokumm Formation (hanging wall) and “basinal” Stephen Formation (footwall) as a steep ramp (Fig. 6, center; Fig. 7A; Fig. 8B, center, solid white line; Fig. 8C, distance 9900 m). The fault surface continues as a steep ramp (hanging wall and footwall) for ~500 m (Fig. 8C) and then curves to become a moderately southwest-dipping flat (hanging wall and footwall) at the contact between the “basinal” Stephen and Takakkaw Tongue Formations. The fault continues to the southwesternmost Takakkaw Tongue Formation outcrop (Fig. 8C, distance 9100 m, elevation 1600 m).

Mapping of the footwall strata on Mount Field and Burgess Pass revealed minor apparent tectonic deformation in units from the Gog Group to the Takakkaw Tongue Formation (Fig. 8B, photo, right half; Fig. 8C in cross section, distance 9000–12,000 m). This minor deformation contrasts with the overlying “basinal” Stephen Formation argillites, which have shortened and thickened significantly through cleavage development (Fig. 7C, photo). It is proposed that a detachment surface exists at the base of the shortened and thickened “basinal” Stephen Formation, separating it from the underlying less deformed Takakkaw Tongue Formation.

Mapping in the vicinity of the Emerald Pass fault at Mount Burgess was aided by the Field Member (thin argillaceous marker bed), formational contacts, and a dense network of outcrop observations (Figs. 8B and 8C, center-left). The hanging-wall block of the Emerald Pass fault (Figs. 8B and 8C, center-left) was found to contain far more structurally complex features than the simple shortened and thickened adjacent footwall block (Figs. 8B and 8C, center-right). Moreover, the hanging-wall block of the Emerald Pass fault contains structures more complex than the hanging-wall block of the neighboring foreland-directed Martin Creek thrust (Figs. 8B and 8C, left). While the hanging wall of the Martin Creek thrust contains a planar, southwest-dipping panel of rocks with minor internal deformation, the hanging-wall block of the Emerald Pass fault contains a number of (contrarily) hinterland-facing minor folds and back thrusts, indicating a more complex stress pattern at this location (Figs. 8B and 8C, left).

The hanging-wall block of the Emerald Pass fault includes a kilometer-scale hinterland-facing anticline and syncline (Fig. 9, red and blue ribbons, respectively, distance 8300–9300 m, elevation 1500–2500 m). Back thrusts pervade the Tokumm carbonates of the anticline. The Tokumm-Duchesnay contact in the southwest-dipping limb of the syncline (Fig. 9, distance 9000–9300 m; Fig. 10A, upper-right; Fig. 10B close-up with stereogram) was previously mapped as a steep, southwest-dipping normal fault (Cook, 1975, GSC Map 1368A; Balkwill et al., 1980, GSC Map 1496A; Price et al., 1980, GSC Map 1483). The fault was continued down the “East gully” (Fig. 10A) to the talus slope below where back thrust 3 has been mapped. The structural geology of the hanging-wall block of the Emerald Pass fault at Mount Burgess was characterized as extensional to the southwest (Cook, 1975), whereas this study describes the geology of this zone as contractional, as implied by the hinterland-directed thrusts and hinterland-facing folds.

The Field Member, where it appears in outcrop at the hinterland-directed anticline, is offset by minor steep-to-vertical to ostensibly “overturned,” irregular, hinterland-directed thrusts (Fig. 9, cross section, distance 8400–8900 m, thrusts numbered 1–4; Fig. 10A, photo, thrusts numbered 1–4). These hinterland-directed thrusts and their flanking structures are illustrated and discussed below starting with back thrust 3 at the northeast.

Back thrust 3 dips steeply to the northeast (Fig. 9, cross section, distance 8800 m; Fig. 10A, photo, lower right), with a flanking hanging-wall anticline and smaller footwall syncline at the Field Member level. Stratigraphic separation along the back thrust at the Field Member cutoff is ~100 m (Fig. 9, white ovals). The back thrust branches upward into an inaccessible cliff wall (Fig. 10A).

The next back thrust, back thrust 2 (Fig. 9, cross section, thrust 2, distance 8600 m; Fig. 10A, photo, thrust 2), is subvertical to locally back-rotated to a normal fault orientation. Above the upper hanging-wall cutoff of the Field Member (Fig. 11A, photo, lower), the thrust branches above, and the main thrust curves toward the southwest (Fig. 11A, upper). A close-up view of the hanging wall (Fig. 11A, inset polygon; Fig. 11B) shows an anticline that contains a minor isoclinal fold. Higher up the cliff wall, back thrust 2 tips out in a hinterland-facing syncline (Fig. 9, distance 8600 m, elevation 1800 m; Fig. 10A, center-left, blue line; Fig. 11C). The syncline is connected to an overlying anticline by a subvertical to overturned limb (Figs. 10A and 11C). Stratigraphic separation along back thrust 2 at the Field Member is ~70 m (Fig. 9, thrust 2, upper pair of white ovals). This fold-fault geometry indicates that the structure was formed by a hinterland-directed fault-propagation folding mechanism.

Back thrust 4 offers a clear illustration of fault-propagation folding related to a hinterland-directed thrust (back thrust). The thrust (Fig. 10A, inset box thrust 4), shown in close-up in Figure 11D (thrust 4), terminates in a fault-propagation fold in the Tokumm carbonates.

Thrusts 2, 3, and 4 have contributed to the development of the kilometer-scale hinterland-facing anticline (Fig. 9, center). The impact of this folding on the overlying and adjacent strata is illustrated and discussed below. The argillite of the Duchesnay unit adjacent to the southwest-protruding “nose” of the Tokumm-Duchesnay contact (convex hinterlandward) exhibits intense penetrative deformation (Fig. 9, red closed curve); vertical to steeply southwest-dipping cleavage is prominent in outcrops, whereas bedding is rarely found (Fig. 12A). Above and below, but adjacent to the same contact, bedding of the Duchesnay unit remains unaltered, and cleavage is largely absent (Fig. 8, green closed curves; Fig. 12B). The relatively unaltered zones above and below the “nose” (Fig. 9, green closed curves) are recessed forelandward of the anticline. Bedding of the Duchesnay unit can be traced from the lower unaltered zone (Fig. 9, lower green closed curve) to the northeast, where the argillite is fractured. Below the protruding “nose” of the anticline, moderately to steeply southwest-dipping argillite beds of the Duchesnay unit contain moderately northeast-dipping fracture sets (millimeter to decimeter spacing) with hinterland-directed millimeter- to centimeter-scale displacements along the fractures (Fig. 9, larger yellow oval; Fig. 13A). About 100 m higher up, bedding is steeply overturned, and back thrust 1 steeply dips to the northeast (Fig. 9, smaller yellow oval; Fig. 13B). The locations of the “nose” of the anticline relative to the back thrusts within the Tokumm Formation and of the hinterland-directed fracture sets within the Duchesnay unit indicate that the hinterland-facing anticline developed through hinterland-directed shear. The shape of the southwest-protruding “nose” of the Tokumm Formation and the penetrative deformation in the adjacent Duchesnay unit also indicate that this area was subjected to intense horizontal compressive stress related to the localized hinterland-directed shear; the Duchesnay unit has shortened and thickened, accommodating the displacement of the “nose” toward the hinterland.

At the hinge of the kilometer-scale anticline described above, the top of the Tokumm Formation is positioned in close proximity to the considerably younger Oke unit (Fig. 9, distance 8500 m, elevation 2200 m). The Oke unit has been displaced down to the southwest by at least 400 m relative to the Tokumm Formation. A fault zone (southwest side down) is found in outcrop above the hinge of the anticline (Fig. 9, dark blue oval) and can be traced along the base of the Duchesnay unit down to the southwesternmost back thrust (Fig. 9, thrust 1). The geometry and sense of displacement along the fault seem to indicate that it is a normal fault that may have developed during an extensional episode (Fig. 9, dark blue oval). However, the fault appears to have contributed to the thickening of the Tokumm Formation. The fault's apparent linkage with the back thrusts within the Tokumm Formation is more likely an indication that the “normal” fault developed in a contractional regime concurrent with the development of the kilometer-scale anticline, and that it is a back thrust. It remains uncertain whether portions of back thrusts 1 and 2 (Fig. 9) have been rotated into an “overturned” attitude or if variations in the principal stress orientations over a narrow zone caused the “overturned” orientation.

In order to resolve the pattern of observed deformation, the nature of the hanging-wall structures must be integrated with the fault geometry and footwall structures. The results of this integration are discussed in the interpretation section below.

The prevalent hinterland-directed shearing and thrusting and the hinterland-facing folding on Mount Burgess escaped mention by previous field mapping projects (Cook, 1975; Balkwill et al., 1980; Price et al., 1980). In the strata of Mount Burgess, the hinterland-directed shearing and associated structures indicate that a barrier or steep ramp impeded foreland progression of deformation. The Emerald Pass fault is located immediately forelandward of this deformation, and it is plausible that the fault served as the barrier. As tectonic wedging may result in very complex deformation (Price, 1986), it is necessary to study and analyze the subtleties of each element (back thrust and associated folding, and penetrative deformation) in detail.

Integration of stratigraphic and structural relationships of the hanging wall and footwall provides the basis for an interpreted tectonic history of the Emerald Pass fault. The “deconstruction” of stratigraphic and structural features into related elements allows the parsing of deformed features into their respective structural domains. Ramps, flats, and flanking structures play a critical role in determining the strain path. The inferred barrier includes the mechanically weaker “basinal” Stephen Formation, which is intensely shortened and thickened adjacent to the Emerald Pass fault (Figs. 7B, 7C, and 7D). The structural restoration of the deformed “basinal” Stephen Formation allows the determination of a predeformational geometry of the Emerald Pass fault surface. Similar analyses of the Martin Creek and Wapta Mountain thrusts add further insight into the nature and relative timing of the development of each of the thrusts and the Emerald Pass fault.

Mapped outcrops at Mount Burgess and Burgess Pass have shown that the Emerald Pass fault has a flat-ramp geometry (see discussion above; Fig. 14, distance 9100–10,000 m; Fig. 15A). Southwest of the flat-ramp, the units of the hanging-wall block dip more steeply to the southwest (Fig. 14, Tokumm-Duchesnay contact, distance 8100–8500 m). A corresponding steeper dip is inferred to occur in the Emerald Pass fault, with the downturn interpreted to occur at the southwesternmost Takakkaw Tongue Formation outcrop of the footwall (Fig. 14, distance 9100 m). Though not constrained by surface geology, the fault is interpreted to cut down to the southwest through the Gog Group (Figs. 14 and 15A), in a manner analogous to the Cataract Brook and Stephen-Cathedral normal faults to the east (Fig. 4). The Emerald Pass fault at Mount Burgess is interpreted to have a lower-ramp–middle-flat–upper-ramp geometry (Figs. 14 and 15A).

Two plausible end-member kinematic models of the development of the Emerald Pass fault at Mount Burgess are discussed below. The conventional tectonic model, in which the strain path consists of contraction followed by extension, is presented first. An alternate model, which follows an inversion strain path (extension followed by contraction), is presented second.

The Tokumm Formation lies directly on the Takakkaw Tongue Formation because of the removal of the “basinal” Stephen Formation (Fig. 14, distance 9000–9800 m, elevation 1500–1700 m). This removal could have been accomplished through a single dip-slip extensional movement of the hanging wall over a series of ramp-flat steps (Fig. 15B). The hanging-wall flat would have slid down past the footwall ramp to lie above the footwall flat. The hanging-wall ramp then would have slid down past the hanging-wall flat to lie adjacent to the footwall ramp (Fig. 15B). The correspondence of hanging-wall and footwall ramp geometries across the upper ramp (Fig. 15A), and the correspondence of hanging-wall and footwall flat geometries across the flat, requires matching lengths of hanging-wall and footwall flats and a displacement along the fault that is commensurate with the dip-directed lengths of the flat and ramp (Fig. 15B), which presents a remarkable coincidence. The dip-directed displacement measured along the inferred fault trajectory at the cutoffs of the Tokumm-Stephen contact is ~1260 m (Fig. 14). This displacement is more than double the displacement noted at Emerald Pass to the north (Fig. 4; as previously discussed in “Fault Characteristics Occurring in the Network of Normal Faults” section). In light of the fact that the Emerald Pass fault tips out at Dennis Pass to the south (Fig. 4), the inferred displacement at Burgess Pass is anomalously large and requires justification under this model.

An alternate kinematic model of extension followed by contraction can account for all of the observations and explain the remarkable correspondence of the ramps and flats. The deformation at Mount Burgess and Burgess Pass can be explained through the restoration of the distorted Emerald Pass fault. The restoration of the distorted Emerald Pass fault at Mount Burgess addresses the shortened and thickened “basinal” Stephen Formation in the footwall block of the fault. The first step is accomplished by sliding the Tokumm Formation in the hanging-wall block to the southwest along the flat at the base of the “basinal” Stephen Formation (Figs. 15A and 15C). A detachment at the base of the “basinal” Stephen Formation was proposed previously in the “Emerald Pass Fault at Burgess Pass: Footwall” section. The “basinal” Stephen Formation of the footwall block that was shortened and thickened through intense cleavage development is restored in the same step. After the first step, the flat is eliminated, and the Emerald Pass fault resumes its former steeply southwest-dipping planar surface. In the second step, the hanging-wall strata, from the “basinal” Stephen Formation and younger units, are restored by reversing displacement on the back thrusts, opening the folds, and reversing penetrative strain (Figs. 15C–15D). The restoration of the shortened and thickened “basinal” Stephen Formation in the hanging-wall block of the Emerald Pass fault effectively lowers the overlying Tokumm Formation, positioning it against the Takakkaw Tongue Formation of the footwall block (Fig. 15D). The two steps reverse the effects of a contractional episode that the Emerald Pass fault is interpreted to have experienced after it was initially formed.

Hinterland-directed thrusts along with associated hinterland-facing folds and intense localized penetrative deformation indicate that a barrier or steep ramp impeded foreland progression of deformation. This study proposes that the Emerald Pass fault provided the basis for tectonic wedging (see other examples in the area in Price, 1986; Colpron et al., 1998). The wedge is situated in the hanging-wall block of the fault, and it is bounded below by foreland displacement along the detachment zone within the “basinal” Stephen Formation and above by the hinterland-directed thrusts within the Tokumm Formation (Fig. 15C). Earlier normal displacement along the Emerald Pass fault placed the lowermost Tokumm Formation (hanging wall) adjacent to the relatively competent Takakkaw Tongue Formation (footwall) (Fig. 15D). The subsequent contraction produced hinterland-directed thrusts and associated folds within the Tokumm Formation of the hanging-wall block of the Emerald Pass fault. Continued contraction shortened and thickened the “basinal” Stephen Formation of the hanging-wall block, elevating the overlying Tokumm Formation (Fig. 15C). Once the Tokumm Formation was positioned adjacent to the “basinal” Stephen Formation of the footwall block, the Tokumm Formation began to slide forelandward over the Takakkaw Tongue Formation of the footwall over a detachment surface that developed at the base of the “basinal” Stephen Formation. Movement was accommodated through penetrative shortening and thickening of the footwall's “basinal” Stephen Formation (Fig. 15A).

The structural patterns associated with the Emerald Pass fault that are observed at Mount Burgess (Fig. 14) can be attributed to an earlier extensional episode that created the normal fault, followed by a contractional episode that distorted the fault surface and deformed the flanking strata. The slickenside data obtained from fractures associated with normal faults in the Gog Group orthoquartzites that were measured by Cook (1975) (Fig. 4) support an episode of contraction after the normal faults had developed. This study contends that the strain patterns (which developed along preexisting normal faults during contraction) were controlled by the rheology of the rock units and the mechanical stratigraphy of the adjacent strata. At any location along the normal faults, the contractional strain pattern was a product of the contrasting rock properties of the neighboring units across the fault. When movement occurred along the normal fault surface during contraction, the positioning of neighboring units may have changed, and the strain patterns changed as the fault evolved. Deformation of hanging-wall and footwall block strata matches the distortion of the normal fault surface. In order to detect and resolve these patterns with a high degree of certainty, mapping at a fine scale (~30 m) was required.

A kinematic model that agrees with the convention of contraction followed by extension (first model) requires an unspecified feature or property to account for the prevalence of hinterland-directed thrusts and hinterland-facing folds (Fig. 9). The first model does not explain the intense shortening of the “basinal” Stephen Formation adjacent to the fault (Figs. 7C and 7D). As illustrated in Figure 15B, the first model would require the hanging-wall Tokumm Formation carbonates to have folded twice as the hanging-wall block slid over the fault bends, yet the carbonate strata show only minimal deformation (Fig. 7E).

The geometry of the Martin Creek thrust provides clues about how the “basinal” Stephen Formation argillites and overlying Tokumm Formation carbonates behaved under compressive stress (Fig. 6, center-southwest; Fig. 14, distance 7000–8000 m). The Tokumm Formation, where it occurs in the hanging-wall block of the thrust, is a flat superimposed against a ramp in the corresponding footwall (Fig. 14, distance 7500–7800 m, elevation 1600–2000 m). The absence of “basinal” Stephen Formation in the hanging wall of the Martin Creek thrust implies that the fault surface follows a detachment surface at the top of the Stephen Formation (Figs. 14 and 15E). The detachment surface is interpreted to separate the less competent argillaceous “basinal” Stephen Formation from the more competent carbonates of the Tokumm Formation. When the thrust is restored by sliding the hanging-wall block down along the thrust surface, it is apparent that the fault continues to the southwest along the top of the “basinal” Stephen Formation (Figs. 15E–15F). The thrust fault does not continue cutting down through the argillaceous “basinal” Stephen Formation as a discrete planar surface. Rather, the contractional deformation of the argillaceous “basinal” Stephen Formation is interpreted to have been expressed as ductile penetrative deformation with detachment surfaces at its top and base (Figs. 15E and 15F). Similar ductile penetrative deformation above a detachment surface was inferred in the “basinal” Stephen Formation in the footwall of the Emerald Pass fault at Burgess Pass.

The very steep southwest dip of the Martin Creek thrust can be attributed to the effects of hinterland-directed shear and rotation that occurred on the foreland side of the thrust (Fig. 14). The Martin Creek thrust is interpreted to have developed as a low-angle, foreland-directed thrust and was later rotated into its steeper inclination by hinterland-directed shear. The Martin Creek thrust is interpreted to have initiated before the hinterland-directed shear occurred; however, the age of the thrust relative to the Emerald Pass fault cannot be determined with confidence at this location.

Cook (1975) inferred that the Wapta Mountain thrust (Fig. 4, center, see Wapta Mountain, 2 km north of Burgess Pass) continued through the “basinal” Stephen Formation at Burgess Pass (Fig. 6). Maintaining the hypothesis that extension occurred after contraction, the interpretation further proposed that the thrust had been cut and offset by the Emerald Pass fault (Cook, 1975). Detailed field mapping in this study has not borne out the same conclusion, as no thrust fault was found to be cutting through the “basinal” Stephen Formation at Burgess Pass. Along the northwest slopes of Wapta Mountain and just above the Stephen Formation, the Wapta Mountain thrust carries Eldon Formation strata (platform equivalent to the Tokumm Formation) as a hanging-wall flat against a footwall ramp of the Eldon Formation (Fig. 4, center, northwest of Wapta Mountain). The absence of the Stephen Formation in the hanging wall is an indication that the thrust follows a detachment surface at the top of the Stephen Formation (Fig. 15G, upper right) in a manner similar to the Martin Creek thrust. The development of the Wapta Mountain thrust can be attributed to the effects of penetrative shortening of the underlying Stephen Formation, which is located within the footwall block of the Emerald Pass fault. A detachment surface, located at the base of the “basinal” Stephen Formation, was previously proposed by McIlreath (1977) as an element of the Wapta Mountain thrust at Wapta Mountain. Contractional deformation of the “basinal” Stephen Formation in the footwall block would have distorted the Emerald Pass fault into its ramp-flat-ramp configuration at Burgess Pass and Mount Burgess, and the same contractional strain would have been transferred to the overlying Eldon Formation as a thrust fault. This interpretation proposes that the origin of the Wapta Mountain thrust was a consequence of the contractional episode that distorted the Emerald Pass fault and that the Wapta Mountain thrust therefore postdates the development of the Emerald Pass fault.

The occurrences and characteristics of the complex contractional deformation that are observed in the slope facies in the vicinity of Mount Burgess indicate that the structural style is strongly dependent on the mechanical stratigraphy of the sedimentary rock sequence (the arrangement of alternating lithologic units). Carbonate units are observed to have contracted principally through the development of thrusts and folds. Shale and argillite units reveal significant shortening and thickening through ductile, homogeneous, penetrative deformation. This study proposes that a detachment surface between lithologic boundaries separates the contrasting styles of deformation.

No through-going thrust has been found in the shales of the “basinal” Stephen Formation at Burgess Pass. Given the deformation styles observed at the pass, it is more likely that detachment surfaces lie at the top and base of the “basinal” Stephen Formation, separating it from the less deformed underlying Takakkaw Tongue Formation and the overlying carbonates of the Tokumm Formation. The presumed structural relationship of the Emerald Pass fault cutting the Wapta Mountain thrust at Burgess Pass (Cook, 1975, contraction predating extension) could not be verified. An alternate tectonic model of extension predating contraction, discussed above, can account for the observations made in this study.

Field work for this study included up-close examination of the contractional character of the folds and thrusts on the south slopes of Mount Burgess. Accurate mapping of these hinterland-directed thrusts and hinterland-facing folds led to a new tectonic interpretation. The Emerald Pass fault is interpreted to have acted as a barrier to foreland-directed deformation, resulting in tectonic wedging and the development of hinterland-directed structures.

McMechan et al. (2018, and references therein) interpreted that the last phase of contraction in the southern Canadian Rockies occurred during the Late Cretaceous to Eocene time interval (Fig. 3). During this time period, the preexisting (Late Jurassic to Early Cretaceous) orogenic belts were carried northeastward to their present location. Evidence derived from geologic mapping of the upper Kicking Horse region of the southern Canadian Rockies (Fig. 4) indicates that the extensional episode that generated the network of normal faults predated the last contractional episode (Fig. 3, phase 3 contraction). Therefore, the normal faults are interpreted to have developed before the Late Cretaceous.

The absence of significant growth strata associated with the normal faults dates the normal fault development to a time after deposition occurred during the Cambrian Period. Normal faults are observed to cut latest Devonian and earliest Carboniferous strata (Price, 2005; Pana and Elgr, 2013); therefore, the normal faults are interpreted to have developed after the Devonian Period. This study proposes that the network of normal faults developed prior to the last major episode of contraction and that a restoration of the faults to their location of origin repositions the normal faults over thinned crust, west of the Rocky Mountain Trench (Fig. 2).

The reinterpretation presented in this paper concludes that the study area was once a rifted margin that developed at the time of the creation of the Slide Mountain Ocean during the latest Devonian to Permian and was subsequently inverted during the contractional phases of the Jurassic to Cretaceous Periods. The rifting of pericratonic terranes away from North America produced extensional features, including the network of normal faults that occur in the study area, as well as the alkaline intrusive units of the Ice River complex. The normal faults were later carried onto unattenuated crust during the last contractional episode and thus escaped overprinting by igneous activity and extensional episodes that occurred through the Paleogene to Neogene time periods. The evidence upon which this interpretation is based does not indicate the number of distinct episodes of extension that occurred in the study area, nor does it pinpoint the ages of the extensional periods. The area may have experienced multiple episodes of alternating extension and contraction.

The results of targeted, fine-scale mapping have provided insight into the evolution of the southern Canadian Rocky Mountain thrust-and-fold belt. A period of extension is interpreted to have been followed by a period of contraction. The later episode of contraction is substantiated by the deformation of normal fault surfaces and strata adjacent to the preexisting normal faults. The contraction along each of these normal faults is expressed in one of two ways, corresponding to the lithology found adjacent to the fault: Penetrative deformation is observed in argillaceous units, whereas thrust faulting and folding developed in carbonate units. A bedding-parallel detachment surface that separates penetrative deformation of argillaceous units from thrusting carbonates is proposed to have facilitated this difference in deformation styles. The updated tectonic model proposes that the network of normal faults developed over attenuated crust west of the Rocky Mountain Trench during rifting and development of the Slide Mountain back-arc basin and ocean. Significant extension is proposed to have commenced in the earliest Carboniferous and ceased during the Permian Period. The later contractional episodes of the Late Cretaceous to Eocene time period displaced the normal faults northeast of the location in which they were formed.

I would like to acknowledge the support from Parks Canada Staff. The field study occurred within Yoho National Park, which includes traditional territory of the Ktunaxa, Secwepemc, and Stoney Nations. Ian McIlreath suggested the study in the first place and provided valuable insight during the early stages of the project. Margot McMechan, Glen Stockmal, and Barry Richards (all from the Geological Survey of Canada) provided welcoming access to their expertise. Margot and Glen deserve special credit for providing critical reviews of an earlier version of the manuscript. Associate Editor Terry Pavlis provided stalwart support and Editor David Fastovsky demonstrated great patience while this manuscript went through several iterations. The feedback from these individuals, in addition to three anonymous reviewers and Geosphere staff, greatly improved the manuscript. Special acknowledgment goes to my wife Kathy, who spent a great amount of time helping me with this project. Any errors, omissions, and weak interpretations are mine.