Gold Open Access: This paper is published under the terms of the CC-BY license.


The Cenozoic Susitna basin lies within an enigmatic lowland surrounded by the Central Alaska Range, Western Alaska Range (including the Tordrillo Mountains), and Talkeetna Mountains in south-central Alaska. Some previous interpretations show normal faults as the defining structures of the basin (e.g., Kirschner, 1994). However, analysis of new and existing geophysical data shows predominantly (Late Oligocene to present) thrust and reverse fault geometries in the region, as previously proposed by Hackett (1978). A key example is the Beluga Mountain fault where a 50-mGal gravity gradient, caused by the density transition from the igneous bedrock of Beluga Mountain to the >4-km-thick Cenozoic sedimentary section of Susitna basin, spans a horizontal distance of ∼40 km and straddles the topographic front. The location and shape of the gravity gradient preclude a normal fault geometry; instead, it is best explained by a southwest-dipping thrust fault, with its leading edge located several kilometers to the northeast of the mountain front, concealed beneath the shallow glacial and fluvial cover deposits. Similar contractional fault relationships are observed for other basin-bounding and regional faults as well. Contractional structures are consistent with a regional shortening strain field inferred from differential offsets on the Denali and Castle Mountain right-lateral strike-slip fault systems.


The Susitna basin of south-central Alaska (Figs. 1 and 2) consists of ∼4–5 km of Cenozoic strata (Stanley et al., 2014). The basin is bounded on the southwest by the intrusive, metavolcanic, and sedimentary rocks of Mount Susitna and Beluga Mountain, on the northwest by the Kahiltna flysch sequence (Wilson et al., 2012) that outcrops in low hills, on the northeast by igneous and metamorphic rocks of the Talkeetna Mountains, and on the southeast by the Castle Mountain strike-slip fault.

Existing structural information for the basin is sparse; aside from the Castle Mountain fault, none of the bounding structures are exposed. Subsurface data consist of a modest number of vintage seismic lines (e.g., Lewis et al., 2015) and a few wells, none of which drilled deep enough to reach crystalline basement. Other data include 1970s and newer gravity data, mostly collected by the State of Alaska and the U.S. Geological Survey (USGS; see Supplemental File1) and aeromagnetic data from surveys collected in 2000 and 2012 by the USGS ( = 4247, = 10001).

Based on access to early exploration data and models, Kirschner (1988, 1994) depicted the Susitna basin as bounded entirely by normal faults. However, this depiction was inconsistent with an earlier gravity interpretation of Hackett (1977a, 1977b) that showed a reverse geometry for the Beluga Mountain fault that bounds Susitna basin to the southwest.

The dip of the Beluga Mountain fault is an ideal target for gravity anomaly investigation. The significant lateral density contrast between the igneous bedrock of Beluga Mountain and the Cenozoic sedimentary deposits of the Susitna basin creates an easily measurable gravity gradient. If the surface projection of the presumed fault structure can be identified, then the overall dip of the structure can be uniquely determined from gravity modeling (e.g., Saltus and Blakely, 2011, and references therein). Hackett (1977b) showed a reverse fault geometry model for the Beluga Mountain fault based on an assumption of the surface fault projection at the topographic mountain front. In this report we present an update of his model, using best available gravity and airborne magnetic data for the Beluga Mountain fault. Using this well-constrained model, we point out other likely thrust or reverse structures based on similar gravity anomaly-topographic front associations in the Susitna lowland region and discuss the relevance of these features in the context of southern Alaska margin geodynamics. An improved structural model for Susitna basin is important for better understanding of hydrocarbon resources and modern seismic hazards close to the population center of Anchorage, Alaska (e.g., Gillis et al., 2013).



The Susitna basin is separated from the adjacent Cook Inlet basin by the seismically active Castle Mountain fault (CMF, Fig. 1) with estimated right-lateral displacements of ∼26 km since 35 Ma (Haeussler and Saltus, 2005) and ∼110–130 km since the Late Jurassic (Trop et al., 2003). We suggest that the difference between these two estimates may indicate that tens of kilometers of right-lateral movement occurred along the Castle Mountain fault during the Cretaceous and/or Paleogene. If this interpretation is correct, then the Cook Inlet and Susitna basins did not form adjacent to each other and may have formed in different tectonic settings.

Previous authors have stated that the Susitna basin is a northern continuation of the Cook Inlet forearc basin (e.g., Rouse and Houseknecht, 2012, p. 2; Craddock et al., 2014, p. 48), but we disagree with this interpretation because the two basins differ from each other in significant ways. The Cook Inlet basin is a 200-million-year-old forearc basin with Late Paleocene to Quaternary nonmarine strata up to 8 km thick (LePain et al., 2013); these strata rest unconformably on a pile of Late Triassic to Late Cretaceous marine strata more than 10 km thick (included in the Peninsular terrane of Nokleberg et al., 1994). In contrast to the Cook Inlet basin, the Susitna basin is less than 60 million years old and consists of nonmarine strata of Late Paleocene to Quaternary age that have a maximum thickness of ∼4–5 km (Stanley et al., 2014). The nonmarine strata of the Susitna basin rest unconformably on Triassic and older igneous, metamorphic, and sedimentary rocks of the Wrangellia terrane (Nokleberg et al., 1994; Schmidt and Rogers, 2007) and Cretaceous to Paleogene plutonic intrusions that represent the roots of a subduction-related magmatic arc. We believe that these differences in age, thickness, and tectonic setting (forearc basin versus magmatic arc) indicate that the Cook Inlet and Susitna basins are distinct from one another, and that the Susitna basin is not a northward continuation of the Cook Inlet basin. Furthermore, we note that comparative stratigraphic studies (Trop et al., 2003; LePain et al., 2013) indicate that the likely northern continuation of the Cook Inlet forearc basin is located in the Matanuska Valley and southeastern Talkeetna Mountains on the opposite (northern) side of the Castle Mountain fault.

The Susitna basin can be divided into two depocenters, the Susitna depocenter and the Yentna depocenter (Figs. 1–3). In the Susitna depocenter, recent investigations using data from exploratory wells, seismic-reflection, gravity, and aeromagnetic data (Shah et al., 2014; Stanley et al., 2014; Lewis et al., 2015) indicate that a deeply buried Paleogene sequence, known primarily from two deep exploratory wells (the Pure Kahiltna Unit 1 and Trail Ridge Unit 1; see Fig. 2), is unconformably overlain by a Neogene and Quaternary sequence that is present in several exploratory wells and scattered surface outcrops (Gillis et al., 2013). The Paleogene basin fill includes a lower interval, more than 800 m thick, of interstratified nonmarine sedimentary and volcanic rocks, including basaltic and andesitic rocks that have yielded whole-rock 40Ar/39Ar ages of ca. 57.3 to ca. 54.3 Ma (Stanley et al., 2014). This volcanic-bearing interval is conformably overlain by a nonmarine sequence ∼1300 m thick that includes sandstone, conglomerate, siltstone, mudstone, and coal with Paleocene to Middle Eocene palynomorphs. The Paleogene volcanic and sedimentary sequence was interpreted by Stanley et al. (2013, 2014) to record volcanism, subsidence, and sedimentation that accompanied eastward passage of a slab window related to subduction of a spreading ridge, consistent with tectonic models proposed by Ridgway et al. (2012) and Benowitz et al. (2012a).

The Paleogene sequence in the Susitna depocenter is unconformably overlain by a package of Neogene and Quaternary nonmarine conglomerate, sandstone, siltstone, mudstone, and coal with Early Miocene to Quaternary palynomorphs (Stanley et al., 2013). The Miocene-on-Paleogene unconformity is not precisely dated but may record uplift and erosion that accompanied the early stages of Yakutat microplate subduction beneath south-central Alaska; speculatively, this uplift and erosion may have coincided with an episode of exhumation ca. 23 Ma inferred from thermochronologic data by Benowitz et al. (2012a, p. 13). The thickest Neogene deposits in the Susitna depocenter, ∼2500 m thick in the Trail Ridge Unit 1 well, occur in a broad synform that is bounded on its western and eastern margins by north-striking reverse faults (Stanley et al., 2014; Lewis et al., 2015). We hypothesize that Neogene subsidence and deposition in the synform, as well as movement along the reverse faults, resulted from contractional deformation associated with Yakutat microplate subduction. It is possible that the north-striking Neogene reverse faults seen on seismic-reflection profiles (Lewis et al., 2015) are reactivated Paleogene normal faults, but this hypothesis remains to be evaluated.

Seismic-reflection and aeromagnetic data show that Paleogene strata in the eastern part of the Susitna depocenter are folded and cut by reverse and thrust faults (Shah et al., 2014; Lewis et al., 2015); the timing of the folding is unclear, but at least some of the northeast-striking faults appear to have surface expression and therefore may be geologically young. Geophysical evidence, discussed in detail in this report, indicates that the southwestern margin of the Susitna depocenter is the Beluga Mountain fault, a northwest-striking, southwest-dipping thrust fault.

In the Yentna depocenter (Peters Hills basin of Haeussler, 2008), surface geologic investigations show that Neogene nonmarine strata consist of conglomerate, sandstone, siltstone, mudstone, and coal, which have yielded palynomorphs of Middle to Late Miocene and Pliocene age (Wolfe et al., 1966; Haeussler, 2008; Gillis et al., 2013; LePain et al., 2015). The thickness of the Neogene strata is ∼3 km (Stanley et al., 2014; based on gravity modeling by R.W. Saltus). The Neogene strata rest in angular unconformity on strongly deformed and locally metamorphosed marine strata of the Kahiltna flysch sequence of Cretaceous and Jurassic age (Wilson et al., 2012; Hults et al., 2013), which is interpreted to record collision of the Wrangellia terrane with North America (Trop and Ridgway, 2007). No seismic-reflection profiling or exploratory drilling has been done in the Yentna depocenter. The Neogene basin in this area is interpreted as a piggyback or wedgetop basin (Ingersoll, 2012, p. 22) that developed above a hypothetical, southeast-directed thrust fault, the Broad Pass fault of Haeussler (2008). In this interpretation, movement along the Broad Pass fault is strongly linked to right-lateral displacement along the Denali fault system (Haeussler, 2008).

Possible sources of sediment for Tertiary strata in the Susitna basin include the Talkeetna Mountains and the Alaska Range. Thermochronologic studies document exhumation episodes in the western Alaska Range ca. 56–35 Ma, 23 Ma, and 6 Ma (Benowitz et al., 2012a; Gillis et al., 2014), prolonged exhumation of the central Alaska Range near the Denali fault starting ca. 28 Ma (Benowitz et al., 2012b), and the eastern Alaska Range since ca. 24 Ma (Benowitz et al., 2011, 2014).

The Susitna lowland was heavily glaciated during Pleistocene time (e.g., Karlstrom, 1964; Schmoll and Yehle, 1986) and is largely mantled by glacial and fluvial deposits in the shallow subsurface. Wahrhaftig (1951) notes that the northern part of the lowland has a “great system of parallel north-south ridges, 10 to 50 feet high, and looks as if it had been furrowed by a great plow.” He also notes a broad medial moraine that extends southwest from the Talkeetna Mountains and forms a hilly country with numerous lakes. The significant glacial morphology in the lowland region obscures direct observation of many structural features including the fault trace of the Beluga Mountain thrust fault.

The northeastern slope of Beluga Mountain follows a generally linear trend striking about N50W, roughly paralleling the trend of the Yentna River, which lies ∼15–20 km to the northeast. The Beluga-Susitna mountain front is the northeastern edge of a triangle-shaped region of elevated topography (Figs. 1–3). The southwestern edge of the elevated triangle trends about N20°W. The shorter southeastern edge of the triangle parallels the Castle Mountain fault. Beluga Mountain, with a maximum elevation of 3670 ft, occupies the northwestern point of the elevated triangle and consists of a broad complex of northeast- and northwest-trending ridges. Northeast-trending Bear Creek follows the southeastern flank of the Beluga Mountain complex and drains to the northeast into Alexander Lake. Mount Susitna occupies the southeastern corner of the elevated triangle and attains a maximum elevation of 4396 ft. The southern flank of Little Mount Susitna (maximum elevation 3035 ft) occupies the southern corner of the elevated triangle. Between Beluga Mountain and Little Mount Susitna, the intervening Wolf Lakes elevated region has more subdued topography, culminating in a broad north-trending ridge with a peak elevation of 2209 ft.

The elevated triangular region encompassing Beluga Mountain, Mount Susitna, and Little Mount Susitna consists of Cretaceous to Paleocene magmatic rocks underlain by rocks of the late Mesozoic and Cenozoic magmatic arc (Wilson et al., 2012). Beluga Mountain, the central Wolf Lakes region, and the valley between Mount Susitna and Little Mount Susitna consist of metamorphosed intermediate volcanic and sedimentary rocks of Cretaceous age (map unit “Kivs” of Wilson et al., 2012). Mount Susitna is mapped as Late Cretaceous granodiorite, tonalite, and quartz monzodiorite. Little Mount Susitna is made up of granitic plutons of Paleocene age. A small region at the top of Beluga Mountain is mapped as intermediate to mafic volcanic rocks of Paleocene age.

Haeussler (2008) cites three seismically active structures at the margins of the Susitna basin: (1) an actively uplifting structure, with an inferred N-S trend and east-directed reverse motion, crossing a broad bend in the Skwentna River (Willis and Bruhn, 2006; “SS” in Fig. 2); (2) the Pass Creek fault (PCF in Fig. 2), which is roughly 17 km long, strikes northeasterly, and is identified as a north-side-down normal fault; and (3) N-S–striking normal faults to the west of the Skwentna structure. Flores and Doser (2005) describe a north-south zone (the Talachulitna zone) of active seismicity striking north-northwest at the eastern edge of the Beluga basin and dipping to the northeast beneath the Beluga triangular highland (TZ in Fig. 2). Flores and Doser (2005) cite one reliable focal mechanism within this zone to indicate reverse-oblique motion. A fault is mapped (Wilson et al., 2012) in the north-south Wolverine–Lewis Creek drainage between Mount Susitna and Little Mount Susitna. Wilson et al. (2012) do not draw a fault along the Beluga-Susitna mountain front. Gillis et al. (2013) report that they did not find any surface evidence of a fault contact during a reconnaissance traverse through Quaternary cover and into volcaniclastic and metavolcanic rocks within a deeply incised drainage along the Beluga Mountain topographic front. However, a normal fault was drawn along or parallel to this mountain front by Kirschner (1994) and Trop and Ridgway (2007).


A low gravity anomaly has long been identified in the Susitna basin (Hackett, 1977a, 1977b, 1978). Barnes et al. (1994) showed a roughly oval gravity low of ∼50 mGal. Hackett (1977b, 1978), Meyer and Boggess (2003), and Meyer (2005) collected and published gravity data for the Susitna basin. For this study, we acquired additional gravity data along several transects across the Beluga Mountain to Susitna basin gravity gradient to support more detailed profile interpretation (Fig. 3). We compiled historical USGS and State of Alaska data and data from our new transect onto the IGSN71 datum and calculated complete Bouguer anomalies using the standard Bouguer reduction density of 2670 kg/m3 (gravity data are available in the Supplemental File [see footnote 1]).

New gravity data better constrain the southwestern portion of the gravity gradient from the elevated Beluga-Susitna region to the gravity low centered on the Susitna depocenter (Fig. 3). This allows for more robust modeling of the gradient sources. The deepest portion of the Susitna gravity low is located ∼15 km northeast of Beluga Mountain. A steep gravity gradient (4.3 mGal/km) parallels the Beluga-Susitna mountain front. The highest gravity anomaly values occur in the elevated region, generally ∼10 km southwest of the mountain front, so that the total width of the gradient zone is ∼15 km.

Two USGS public-domain, moderately low level (flown with a nominal draped flight height of 1000 ft above ground) aeromagnetic surveys encompass the Beluga-Susitna mountain front and the transition from the elevated region into the Susitna basin ( = 4247). These surveys were collected and released as part of the Anchorage Urban Region Aeromagnetics project of the USGS, which focused primarily on interpretation within the Cook Inlet basin (Saltus et al., 2001).

Complex high-amplitude magnetic anomalies are observed within the study area (Fig. 4), reflecting highly magnetic igneous rocks within and beneath the elevated Beluga-Susitna region and within the deep basement of the Susitna basin. In particular, the basin-facing edge of the Wolf Lakes (Fig. 2) central section of the Beluga-Susitna highland, as well as Little Mount Susitna and the western part of Mount Susitna, correlate with very high magnetic anomaly values. The Beluga Mountain region correlates with moderately high magnetic anomaly values. The central section of the Beluga-Susitna triangle, southwest of the basin-facing high, is a region of low magnetic anomaly values with abrupt linear boundaries in an orthogonal pattern parallel and perpendicular to the trend of the Beluga-Susitna mountain front. The moderate and high magnetic anomalies along the Beluga-Susitna mountain front extend basinward ∼5 km into the region of mapped cover deposits of the Susitna basin (magenta line on Fig. 4).

Proprietary industry seismic-reflection data were acquired in Susitna basin from the 1960s to the 1980s (Lewis et al., 2015). The data have relatively low signal-to-noise ratios but clearly show layered reflectors throughout the basin consistent with the significant sedimentary section known from exploratory wells including the Pure Kahiltna Unit 1 and Trail Ridge Unit 1. Six seismic lines are close enough to the Beluga-Susitna mountain front to potentially yield useful structural information about the basin margin. Of these, the one that contains the most interpretable reflections at the Beluga Mountain margin (Lewis et al., 2015, their fig. 30) shows no evidence for a normal fault offsetting a basement reflector. Although this line is heavily disrupted in the region of the magnetic edge (magenta line on Fig. 4), it shows evidence for both the northeast-vergent Skwentna reverse fault and Beluga Mountain thrust fault. This seismic line and several other lines that approach the basin margin show strong northeasterly apparent dips on Tertiary reflectors on the upthrown (southwest) sides of the Beluga Mountain and Skwentna faults, a structural pattern that is indicative of contractional fault-propagation folding and reverse or thrust faulting along the southwest margin of the Susitna basin, rather than normal faulting. While sparse and somewhat indistinct, this seismic-reflection evidence for the Beluga Mountain and Skwentna faults can be interpreted as steeper-dipping, near-surface splays that have propagated from the deeper, regional thrust fault modeled by the potential field data.


Profile Interpretation—Beluga Mountain Fault

Three key observations relate to the structure of the Beluga Mountain fault along the southwest margin of Susitna basin: (1) a mapped geologic and/or physiographic boundary that juxtaposes high-density and variably magnetic igneous rocks on the southwest against surficial glacial and/or fluvial deposits on the northeast; (2) a broad gravity gradient transition from low-gravity anomaly values over the Susitna depocenter to high-gravity values over the igneous rocks of the Beluga highland; the gradient is roughly centered on the physiographic boundary; (3) moderate to high magnetic anomaly values associated with the exposed igneous rocks along the Beluga-Susitna physiographic boundary; these moderate to high magnetic values extend ∼5 km basinward from the physiographic front (magenta line on Fig. 4).

To explore the implications of these key observations, we model a southwest to northeast cross-section profile spanning the Beluga mountain front (location shown on Figs. 3 and 4). The gravity gradient shows a similar profile if drawn across the mountain front anywhere along the front from north of Beluga Mountain southeast to Mount Susitna; so gravity models crossing the front will yield similar results along this part of the margin. The magnetic signal, on the other hand, is more complex and shows considerable lateral variation along the mountain front. On the chosen profile, parallel to Bear Creek, the magnetic anomaly pattern is relatively simple and appears to reflect primarily the geometry of the basin-bounding fault. The magnetic signature along this profile is a single magnetic high that terminates at the toe of the inferred thrust front.

For discussion and demonstration of modeling uncertainties, we show three models for this profile (Fig. 5): (A) a thrust fault geometry using observed geologic constraints; (B) a normal fault geometry using observed geologic constraints; and (C) a more complicated model constructed specifically to force a normal fault geometry. Model (A) is our preferred model. It fits the data and requires the fewest ad hoc assumptions. The fundamental assumptions underpinning model (A) are: (1) the simplification of the geology into two bodies, a dense (igneous and metamorphic) upper plate and a lower density “Susitna basin” lower plate allows for reasonable estimation of fault geometry; (2) the shallow subsurface magnetic rocks modeled are part of the hanging wall (i.e., their location can be used to map the hanging wall where it is concealed by the shallow basin sediments); (3) there is no hidden density contrast lurking beneath the basin-bounding structure (as in model C below). The main reason that our preferred model has a shallower dip angle for the Beluga Mountain fault compared to the steeper dips of Hackett (1977a, 1977b) results from our inferred position for the toe of the thrust fault. Hackett (1977a, 1977b) placed the toe close to the mountain front. Also we have more complete gravity data coverage that shows the gravity gradient to be somewhat broader than shown in Hackett’s original data (1978).

Model (B) shows that a normal fault geometry does not fit the data. The fundamental problems with this or any related model are the broad and smooth nature of the gradients and the fact that the gravity gradient extends well to the southwest of even the most mountain-ward possible location (i.e., the edge of the surficial basin sediments) for the surface fault trace. Any normal (or vertical) fault geometry for a two-body solution will not match the observed gravity gradient.

Model (C) fits the position of the gravity gradient but is not preferred because it relies on insertion of an ad hoc structural element at depth. We construct this model as an illustration of uncertainties in potential field modeling. It is generally possible to concoct a model that incorporates additional unseen (and therefore unconstrained) geologic elements to force a match to a given potential field profile. For this model, we add an ad hoc deeper body with density intermediate to the two surface blocks. The ad hoc body is centered in the middle of the gradient with flanks that dip (in a “normal” geometry) in both directions. On average, this body allows for a smoothing of the gravity gradient by smearing the overall lateral density contrast in both directions from the center. We can think of no reasonable geologic rationale for such a body, and furthermore, there is no seismic indication of a normal fault interface on the basin side (as discussed above).

Other Features in the Gravity and Magnetic Anomaly Data

As previously noted by Hackett (1977b, 1978), other major faults in the region around Susitna basin have a gravity gradient signature consistent with contractional fault geometries (Fig. 6). The gravity gradient associated with the Castle Mountain fault undercuts the southeast flanks of Mount Susitna and Little Mount Susitna. Along with the magnetic expression, in which an intense magnetic high is truncated along a line parallel to the Castle Mountain fault but located ∼5–7 km to the northwest (coincident with the high edge of the gravity gradient), this supports a contractional geometry with a dip to the northwest. Similarly, the gravity gradient on the northern margin of the Matanuska lowlands undercuts the southern flank of the Talkeetna Mountains as expected for a reverse geometry. On the northwest flank of the greater Susitna basin, the gravity gradient bounding the Yentna (Peters Hills) basin undercuts the southeast flanks of the adjacent ridges of the Alaska Range, again suggesting a reverse geometry. Gravity gradients undercut the flanks of the Yenlo Hills, but not as steeply as along the Beluga-Susitna front, suggesting a steeper angle to the reverse structures facing both the Susitna basin to the southeast and the Yentna (Peters Hills) basin to the northwest. In contrast, the gravity gradient on the eastern side of the Susitna basin, adjacent to the Talkeetna Mountains front, tapers into the basin suggesting an onlapping or normal fault geometry (as shown in Stanley et al., 2014).

The USGS Anchorage Urban Region Aeromagnetics (AURA) magnetic survey data (Fig. 4; Saltus et al., 2001) reveal complex magnetic anomalies in the Beluga-Susitna region. Shah et al. (2014) report on preliminary interpretation of magnetic anomalies and possible relationships to basement lithology and structure beneath the Susitna basin. A detailed analysis of the magnetic anomalies of the Beluga-Susitna triangular highland and surrounding region is beyond the scope of this paper.


The thrust-fault geometry model for the Beluga-Susitna front is similar in some ways to the COCORP deep seismic-reflection and/or gravity model for the Wind River thrust fault in Wyoming (Smithson et al., 1978). Like our preferred model, the Smithson et al. (1978) model uses a thrust fault to juxtapose high-density basement of the Wind River uplift (2670 kg/m3) in the hanging wall against the low-density (2370–2600 kg/m3) sedimentary section of the Green River basin in the footwall block. The amount of lateral overthrust (heave) is ∼15 km (Smithson et al., 1978), again similar to our model. In the Wind River case, the total gravity gradient spans nearly 80 mGal, almost twice our gravity range, but the thickness of the overthrust Green River basin section is nearly 10 km compared with our modeled 4 km. The excellent COCORP seismic data have much better signal penetration and lateral coverage across the thrust compared to the shallow exploration seismic data available for Susitna basin. The COCORP data were collected continuously from the basin, through the Wind River uplift, and into the Wind River basin on the other side of the range, imaging the entire crust down to the Moho; whereas the Susitna basin industry seismic data, at best, only image to pre-Tertiary basement on one side of the fault.

We reiterate that we are not the first to interpret the gravity gradient along the Beluga-Susitna mountain front as a contractional fault. Hackett (1977b) states: “This gravity feature is therefore interpreted to be the expression of a high-angle reverse fault dipping 60-75 degrees, up-thrown on the southwest along the Susitna lowland-Beluga Mountain boundary.” His reverse fault interpretation (Hackett, 1977b, 1978) has not been widely appreciated in our opinion. For example, both Kirschner (1994) and Trop and Ridgway (2007) cite Hackett (1977b) but draw the Beluga-Susitna edge of Susitna basin as a normal fault. Flores and Doser (2005) cite both Hackett’s 1977 and 1978 reports and mention his reverse fault interpretation.

A thrust-fault geometry for the southwest side of the Susitna basin implies that a significant thickness of Tertiary nonmarine strata, including coal and other possible hydrocarbon source rocks, may have been overthrust and therefore experienced greater pressure and temperature histories relative to the rest of the basin. We speculate that thrusting may have promoted thermal maturation of petroleum source rocks, including coal and organic-rich shale, and migration of petroleum-bearing fluids from beneath the hanging wall to the foreland. These aspects of a thrust interpretation have implications for oil and gas potential, including a speculative overthrust play adjacent to the southwestern margin of Susitna basin.

The Beluga Mountain fault does not appear to be seismically active today. Instead, Flores and Doser (2005) note a linear band of seismicity (“TZ” on Fig. 2) that forms an apparent northeast-dipping band, with epicenter depths from the surface to 16 km, that they call the Talachulitna zone. This band of seismicity has a surface intercept in the Beluga basin on the western flank of the Beluga-Susitna triangular highland region. Flores and Doser (2005) note Hackett’s (1978) interpretation of the southwest-dipping structure associated with the Beluga-Susitna mountain front to the northeast but point out that, in general, earthquake solutions for the Talachulitna seismicity have very large inversion uncertainties, making it difficult to determine first-motion directions. Conversely, they cite a single Alaska Earthquake Information Center catalog solution for a 4.8-km-deep event with a northeast-dipping reverse-oblique solution.

The existence of east-west shortening for the Beluga and Susitna basin regions is consistent with current models for the overall crustal response to the collision of the Yakutat microplate with the subduction margin (e.g., Haeussler, 2008). Interpretation of modern plate motions from GPS studies (Freymueller et al., 2008; Haeussler, 2008) with constraints from seismic tomography (Eberhart-Phillips et al., 2006) suggests a model in which a rigid crustal plate is rotating counterclockwise at the northwest edge of the Yakutat microplate collision. The combination of shortening and rotation is thought to lead to a transition from margin-normal (margin-perpendicular) to margin-parallel shortening in the approximate location of the Beluga and Susitna basins (Fig. 7 based on fig. 9 from Haeussler, 2008). The Beluga-Susitna thrust fault provides strong evidence that the western edge of the Susitna basin is near the western edge of the counterclockwise-rotating southern Alaska block.


We concur with the earlier gravity interpretations of Hackett (1977a, 1977b, 1978) that the boundary between the high-density igneous rocks of the Beluga Mountain–Mount Susitna region has a contractional fault geometry against the low-density Tertiary sedimentary rocks of the adjacent Susitna basin. We jointly modeled gravity and magnetic anomalies across this margin with constraints from legacy industry seismic data to show an overthrust structure that resembles the Wind River thrust fault in Wyoming (Smithson et al., 1978). A thrust fault structure on the southwest margin of the Susitna basin fits well with current geodynamic models (e.g., Haeussler, 2008) and is consistent with the variation in observed slip between the Denali fault to the north and the Castle Mountain fault system to the south (Benowitz et al., 2012c; Riccio et al., 2014). A significantly buried and heated nonmarine, coal-bearing sedimentary section has implications for a possible overthrust hydrocarbon play and fluid migration in the Susitna basin. Although this particular thrust fault structure does not appear to be seismically active, the presence, history, and nature of thrust (and reverse-oblique) faults in this region have implications for seismic hazard evaluation.

We are grateful for careful journal reviews by Leland O’Driscoll and Bob Gillis and for excellent editorial suggestions from Jeff Benowitz. Our work has benefitted from discussions with Marwan Wartes and Bob Gillis. Thanks also to the pilots of Pollux Aviation and the proprietors of the Skwentna Roadhouse. We acknowledge our appreciation for David Barnes (USGS, deceased), Bob Morin (USGS, retired), and John Meyer, Jr. (Alaska Division of Oil and Gas, retired) for collecting and publishing gravity data in Susitna basin and the surrounding region.

We wish to recognize Steve W. Hackett (deceased) for the original observation of the association between gravity gradients and topographic features in and around the Susitna basin. Steve recognized the significance of the gravity features and made the initial models to illustrate reverse and thrust structures related to the Susitna and Beluga basins. His “Yentna-Beluga” lineament is essentially the same as our Beluga Mountain fault.

1Supplemental File. Gravity data in and around Susitna basin, Alaska. Please visit or the full-text article on to view the Supplemental File