Abstract

The Upper Jurassic and Lower Cretaceous part of the Brookian sequence of northern Alaska consists of syntectonic deposits shed from the north-directed, early Brookian orogenic belt. We employ sandstone petrography, detrital zircon U-Pb age analysis, and zircon fission-track double-dating methods to investigate these deposits in a succession of thin regional thrust sheets in the western Brooks Range and in the adjacent Colville foreland basin to determine sediment provenance, sedimentary dispersal patterns, and to reconstruct the evolution of the Brookian orogen. The oldest and structurally highest deposits are allochthonous Upper Jurassic volcanic arc–derived sandstones that rest on accreted ophiolitic and/or subduction assemblage mafic igneous rocks. These strata contain a nearly unimodal Late Jurassic zircon population and are interpreted to be a fragment of a forearc basin that was emplaced onto the Brooks Range during arc-continent collision. Synorogenic deposits found at structurally lower levels contain decreasing amounts of ophiolite and arc debris, Jurassic zircons, and increasing amounts of continentally derived sedimentary detritus accompanied by broadly distributed late Paleozoic and Triassic (359–200 Ma), early Paleozoic (542–359 Ma), and Paleoproterozoic (2000–1750 Ma) zircon populations. The zircon populations display fission-track evidence of cooling during the Brookian event and evidence of an earlier episode of cooling in the late Paleozoic and Triassic. Surprisingly, there is little evidence for erosion of the continental basement of Arctic Alaska, its Paleozoic sedimentary cover, or its hinterland metamorphic rocks in early foreland basin strata at any structural and/or stratigraphic level in the western Brooks Range. Detritus from exhumation of these sources did not arrive in the foreland basin until the middle or late Albian in the central part of the Colville Basin.

These observations indicate that two primary provenance areas provided detritus to the early Brookian foreland basin of the western Brooks Range: (1) local sources in the oceanic Angayucham terrane, which forms the upper plate of the orogen, and (2) a sedimentary source region outside of northern Alaska. Pre-Jurassic zircons and continental grain types suggest the latter detritus was derived from a thick succession of Triassic turbidites in the Russian Far East that were originally shed from source areas in the Uralian-Taimyr orogen and deposited in the South Anyui Ocean, interpreted here as an early Mesozoic remnant basin. Structural thickening and northward emplacement onto the continental margin of Chukotka during the Brookian structural event are proposed to have led to development of a highland source area located in eastern Chukotka, Wrangel Island, and Herald Arch region. The abundance of detritus from this source area in most of the samples argues that the Colville Basin and ancestral foreland basins were supplied by longitudinal sediment dispersal systems that extended eastward along the Brooks Range orogen and were tectonically recycled into the active foredeep as the thrust front propagated toward the foreland. Movement of clastic sedimentary material from eastern Chukotka, Wrangel Island, and Herald Arch into Brookian foreland basins in northern Alaska confirms the interpretations of previous workers that the Brookian deformational belt extends into the Russian Far East and demonstrates that the Arctic Alaska–Chukotka microplate was a unified geologic entity by the Early Cretaceous.

INTRODUCTION

Foreland basins are the receptacles for detritus shed from adjacent thrust belts and provide a stratigraphic record of their evolution (e.g., Graham et al., 1976). In addition to the well-documented linkages between foreland basins and their adjacent thrust belts, recent work has shown that the detrital record of some basins also contains sediment derived from distant parts of the same orogenic belt and from other geologic provinces in the autochthon or distant hinterland (e.g., Lawton et al., 2003, 2009; Dickinson and Gehrels, 2008; Saylor et al., 2011; Dickinson et al., 2012). Such sediments commonly are transported long distances down the axis of foreland basins by longitudinal dispersal systems and mix with detritus derived from local source areas by short-headed and transverse dispersal systems. A full understanding of the contributions from local and distant source areas can provide a regional context of the orogenic activity as well as provide a record of deformation in the adjacent thrust belt.

Investigations of the interplay between transverse and longitudinal dispersal systems in foreland basins can be approached through the study of the compositional changes in a foreland basin through time. The early history of foreland basins is commonly not well preserved, however, because sedimentary successions deposited near the thrust front get incorporated into the advancing thrust belt and are at high risk for imbrication, uplift, and erosion during later stages of orogenic activity. The parts of foreland basins most likely to be preserved in front of the orogen typically record distal up-dip deposition relative to the basin axis late in the basin’s development (e.g., DeCelles, 2012). Thus, the oldest sediments sampled in many intact foreland basins are starved or thin sedimentary sequences deposited near the end of tectonism. For this reason, study of the synorogenic deposits preserved in the thrust belt are important for unraveling the relative contributions from transverse and longitudinal dispersal systems in foreland basins through time.

The latest Jurassic and Early Cretaceous Brooks Range fold-thrust belt of northern Alaska is a peripheral orogen with a stack of north-directed thrust sheets, each capped by synorogenic deposits. A foreland basin to the north, the Colville Basin, is filled with a 10-km-thick succession of lithic marine, shelfal, and nonmarine sedimentary rocks deposited on the autochthon. These deposits are largely younger than the synorogenic deposits associated with the thrust sheets in the Brooks Range, indicating that the Colville Basin represents only the last stage of a forward-migrating coupled deformational belt and foreland basin system that existed throughout the Early Cretaceous.

The synorogenic deposits of the western Brooks Range provide a good record of the early history of sedimentation derived from the developing orogenic belt. Additional motivation for this study derives from the fact that the Brooks Range is one of the major orogenic systems bordering the Arctic Ocean basin, and its foreland basin may contain a sedimentary record of some of the events that occurred in the Arctic region before and during construction of the western Arctic Ocean basin by rifting in the Early Cretaceous. The tectonic history of this older part of the Arctic basin (Amerasia Basin) has long been controversial and difficult to resolve (e.g., Lawver et al., 2002; Grantz et al., 2011b; Pease, 2011). This study examines the early foreland basin deposits of the Brooks Range using a combined approach that unites sandstone petrography, detrital zircon U-Pb geochronology, zircon fission-track U-Pb double dating of individual zircons, and field studies to understand the origin of the depositional systems that dispersed sediment into the basin as the orogen evolved. The results have implications for the nature and tectonic significance of the upper plate of the Brooks Range orogen, for the collisional processes that were involved in the deformation and how they changed along strike, and for the timing of exhumation. The results also provide clues about the regional paleogeography and tectonic configuration of the Arctic region during the critical time that the Amerasia Basin was formed and the dispersal pathways for sediment that were derived from the Uralian-Taimyr orogen and deposited on the opposite side of the Arctic Ocean in the synorogenic deposits of the Brooks Range.

GEOLOGIC FRAMEWORK

The Brooks Range fold-thrust belt extends E–W across northern Alaska from the Canadian border to the Chukchi Sea and probably continues farther west beneath the Chukchi Sea as the Herald and Wrangel arches to perhaps as far west as the New Siberian Islands, ∼750 km northwest of Figure 1 (Drachev et al., 2010; Grantz et al., 2011a). The Brooks Range and adjacent Colville Basin in northern Alaska are thought to be the products of north-directed Jurassic and Early Cretaceous arc-continent collision between the south-facing (present coordinates) passive continental margin of the Arctic Alaska terrane and the Jurassic Koyukuk arc terrane of interior Alaska (e.g., Moore et al., 1994). The collision resulted in closing of the Angayucham Ocean and northward emplacement of the Angayucham terrane, an oceanic assemblage composed of a Jurassic ophiolite and subduction complex, and the Koyukuk arc terrane onto Arctic Alaska (Fig. 2A). A similar history is recognized in the Russian Far East, where the closing of the South Anyui Ocean basin culminated in collision of Carboniferous to Jurassic arc terranes and Jurassic and Early Cretaceous ophiolite and subduction complexes of the Alazeya-Oloy fold belt and South Anyui terrane onto the Chukotka terrane to the north (Sokolov et al., 2009; Sokolov, 2010). In western Chukotka, the closure of the South Anyui Ocean may have culminated in collision with the outer margin of Siberia (Miller et al., 2009). The arc-continent collision along the entire length of the orogen was approximately concurrent with rifting of the Arctic Alaska and Chukotka terranes away from their original positions along the northern margin of North America, probably by counterclockwise rotation during formation of the Amerasia Basin in the Jurassic and Early Cretaceous (e.g., Lawver et al., 2002; Grantz et al., 2011b), although other scenarios have been proposed (e.g., Lane, 1997; Miller et al., 2006, 2009; Amato et al., 2009; Dickinson, 2009).

Models for the late Mesozoic collisional events in Alaska and the Russian Far East portray these areas as parts of a complex region of arcs and microcontinental fragments along the northern margin of North America in the ancestral northern Pacific Ocean (e.g., Rubin et al., 1995; Nokleberg et al., 2000). These models correlate the Angayucham and South Anyui oceans as parts of the same late Paleozoic to Mesozoic oceanic basin and the Arctic Alaska and Chukotka terranes as eastern and western parts, respectively, of the same continental margin. Similarities in Neoproterozoic and lower Paleozoic stratigraphy, plutonic belts, and Mesozoic deformational and metamorphic events support correlation of the terranes (e.g., Natal’in et al., 1999; Dumoulin et al., 2002). In this paper, “Arctic Alaska–Chukotka microplate” is used to designate the conjoined Arctic Alaska and Chukotka terranes, but we refer to “Arctic Alaska” and “Chukotka,” respectively, to specify features limited to only those parts of the microplate.

In the western Brooks Range, the forward (northern) part of the orogen is composed of a series of stratigraphically thin thrust sheets consisting of Upper Devonian to Jurassic continental margin sediments, each depositionally overlain by a succession of deep-marine synorogenic clastic deposits (Martin, 1970; Mayfield et al., 1988) (Fig. 3). From bottom up, these allochthons consist of the Endicott Mountains allochthon (EMA), the Picnic Creek allochthon (PCA), Kelly River allochthon (KRA), Ipnavik River allochthon (IRA), and Nuka Ridge allochthon (NRA) (Fig. 2B). A proximal autochthonous succession underlies the allochthons and constitutes basement for the orogen-derived deposits in the foreland. The upper Paleozoic and lower Mesozoic deposits in both the allochthons and the autochthon consist of generally similar, texturally and compositionally mature strata, including Devonian and Lower Mississippian quartz- and chert-rich siliciclastic strata (Endicott Group), Mississippian and Pennsylvanian carbonate platform deposits (Lisburne Group), and Permian and Triassic fluvial to outer continental margin clastic, shale, and siliceous deposits (Sadlerochit Group, Shublik Formation, and Etivluk Group) (Fig. 4). In the Nuka Ridge allochthon, the compositionally mature Mississippian siliciclastic deposits are replaced by coeval arkosic deposits that have yielded abundant zircons with ca. 2.1–2.0 Ga U-Pb ages and microclines with 1.2–1.1 Ga K-Ar ages (Moore et al., 1997). Granitic rocks of this age and cooling history are not known in the Arctic Alaska–Chukotka microplate but have been reported locally in continental terranes in southwestern Alaska and Siberia (Moll-Stalcup et al., 1996; Bradley et al., 2007; MacLean et al., 2009).

Stratigraphic and structural relations indicate that the western Brooks Range constitutes a latest Jurassic and earliest Cretaceous foreland-younging thrust stack (Mayfield et al., 1988). Paleontologic data indicate that synorogenic deposits are as old as latest Jurassic in the higher allochthons but are probably no older than Valanginian (Early Cretaceous) in the Endicott Mountains allochthon (Fig. 4). This suggests that each allochthon was covered by ancestral foreland basin successions derived from the orogenic belt at the time it was located near the front of the orogen. As thrusting propagated northward, foreland deposits were successively detached with their underlying platform successions as allochthons and incorporated into the thrust belt. Estimates of total shortening represented by the Brookian orogen range from 250 to 500 km or more (Moore et al., 1986; Oldow et al., 1987; Mayfield et al., 1988; Young, 2004).

The structurally highest allochthons in the Brooks Range are klippen of the Copter Peak allochthon (CPA) and the overlying Misheguk Mountain allochthon (MMA), which together form the oceanic Angayucham terrane (Moore et al., 1994) (Fig. 2B). The Copter Peak allochthon consists of imbricates of mafic igneous rocks and covering units of chert that are thought to compose an early Mesozoic subduction assemblage, whereas the Misheguk Mountain allochthon consists principally of Middle Jurassic ophiolite (Harris, 2004) that probably represents forearc basement to the Koyukuk arc located farther south in interior Alaska (Fig. 2A).

The hinterland of the Brookian orogen consists of a continental assemblage of Devonian and older quartz schist, marble, and orthogneiss exposed in the southern Brooks Range. Large 395–380 Ma orthogneiss batholiths intrude these rocks in addition to sparse, smaller orthogneiss plutons dated at 750–700 Ma and 971 Ma (e.g., Karl and Aleinikoff, 1990; Aleinikoff et al., 1993). Most of these rocks underwent high-pressure/low-temperature metamorphism between ca. 160 and 120 Ma (Late Jurassic to Early Cretaceous) (Christiansen and Snee, 1994) followed by regional greenschist-facies metamorphism at the time of exhumation, ca. 120–90 Ma (Albian–Cenomanian) (e.g., Miller and Hudson, 1991). Exhumation was accompanied by south-side-down normal faulting along the southern flank of the Brooks Range (e.g., Little et al., 1994) (Fig. 1). These events produced a regional northward tilt that extends across the southern Brooks Range and the frontal part of the orogen to the axis of the Colville Basin (Vogl et al., 2002). Rocks similar to those in the southern Brooks Range are also present in the Seward Peninsula and the Ruby terrane south of the Brooks Range (e.g., Roeske et al., 1995; Amato et al., 2009) suggesting that the hinterland of the Brooks Range orogen probably extends southward into those areas (Fig. 1).

Deformed and metamorphosed Paleozoic and older rocks also constitute most of basement of Chukotka (Bering Strait Geologic Field Party, 1997), although prograde high-pressure metamorphic conditions have not been confirmed in these rocks and a covering succession of allochthons composed of imbricated platform and continental margin sedimentary strata like those in the Brooks Range have not been identified. Deformed and metamorphosed Triassic turbidite clastic rocks (Ustieva unit of Sokolov et al., 2009) instead overlie the Paleozoic rocks in Chukotka. These rocks are overlain by allochthons of ophiolite (the South Anyui terrane) and Upper Jurassic and Lower Cretaceous orogenic strata similar to the Copter Peak allochthon and Misheguk Mountain allochthon in the structurally highest parts of the Brooks Range (Sokolov, 2010).

BROOKIAN SEQUENCE

Ancestral foreland basin strata in the Brooks Range and depositional fill of the Colville foreland basin are composed of litharenites derived from the Brooks Range (Mull, 1985; Bird and Molenaar, 1992). These deposits compose the older part of the upper Mesozoic and Cenozoic Brookian sequence, which is distinguished by its lithic composition and southern provenance from the underlying quartz-rich, northerly derived platform and rift-shoulder deposits of the upper Paleozoic and lower Mesozoic Ellesmerian and Beaufortian sequences (Lerand, 1973; Hubbard et al., 1987).

In the Colville Basin, the early Brookian clastic sediments prograded longitudinally along the basin from west to east during the Aptian and Albian (late Early Cretaceous), although paleocurrent data near the southern margin of the basin indicate significant detritus also came from the Brooks Range to the south (Molenaar, 1988; Bird and Molenaar, 1992; Houseknecht et al., 2009). A major deltaic depocenter in the western part of the basin composed mainly of westerly derived sediments is termed the Corwin Delta, whereas a large depocenter located in the central North Slope composed of south-derived sediment is termed the Umiat Delta (Fisher et al., 1969; Ahlbrandt et al., 1979; Huffman et al., 1988) (Fig. 5).

East-west seismic reflection profiles along the axial part of the basin reveal a spectacular clinoform geometry that has driven the Cretaceous stratigraphic nomenclature employed in the basin (Molenaar, 1988; Bird and Molenaar, 1992). Bottomset seismic facies mark deposits of condensed deep marine mudstone and radioactive shale that are assigned to the Barremian to Campanian Hue Shale (Fig. 2C). Stratigraphically higher foreset seismic facies correspond to deep marine basinal, slope, and outer-shelf mudstone and sandstone turbidites of the Aptian to Cenomanian Torok Formation. Topset facies characterize shallow-marine to non-marine coal-bearing sandstone, mudstone, and conglomerates of the Albian to Cenomanian Nanushuk Formation (Fig. 6A), which are >6 km thick in the western Brooks Range (Mull et al., 2003).

Along the southern margin of the Colville Basin, the Hue Shale is absent and the basal strata consist of coarse-grained shallow-marine deposits assigned to the Fortress Mountain Formation (Molenaar, 1988; Houseknecht and Wartes, 2013) (Figs. 2B and 2C). In the western part of the basin, these strata consist of up to ∼3 km of sand-rich deposits that are characteristically rich in detrital carbonate and white mica (Fig. 6C). This part of the Fortress Mountain Formation is called the Mount Kelly Graywacke Tongue of the Fortress Mountain Formation (Mull, 1985), which for simplicity is referred to as the Mount Kelly Graywacke in this paper. Locally, the Mount Kelly Graywacke is underlain by more than 250 m of thin-bedded, fine-grained sandstone and mudstone that was informally named the lower Brookian shale by Mull (1995). These strata have yielded Hauterivian microfossils and are interpreted to represent the earliest phase of Colville Basin sedimentation in this part of the basin. The Mount Kelly Graywacke is probably Aptian (Mull et al., 2000), but stratigraphic relations allow it to be as old as Hauterivian or Barremian. It is overlain by slope mudstone deposits of the Torok Formation.

The well-developed progradational patterns of the axial part of the Colville Basin depositional fill suggest that its provenance lies in the Lisburne Peninsula, Chukotka Peninsula, and now subsided highlands such as the Herald Arch (Molenaar, 1988). Based on compositional observations, Mull (1985) concluded that the Mount Kelly Graywacke was derived from sources in both the metamorphic hinterland and the allochthons that comprise the fold-thrust belt of the Brooks Range.

Ancestral foreland basin deposits associated with the allochthonous sequences of the Brooks Range are largely assigned to the Okpikruak Formation (Figs. 2B and 2C), which consists of deep-marine micaceous sandstone, shale, polymict conglomerate, and pebbly mudstone (Moore et al., 1986; Young, 2004). In the Red Dog Mining district where it is best described, the Okpikruak Formation comprises successions of dark mudstone and thin-bedded turbidites more than 375 m thick (Young, 2004). Polymict conglomerate is locally common and contains angular to well-rounded clasts of carbonate rock and chert derived from the Lisburne and Etivluk groups, igneous clasts including basalt, granite, gabbro, and ultramafic rocks that may have been derived from the ophiolite of the Misheguk Mountain allochthon and Copter Peak allochthon, and schist, orthogneiss, and other rock types that are not found in any of the allochthons (Young, 2004; De Vera, 2005) (Fig. 6E). Olistostromes are locally common, containing meter- to kilometer-scale blocks of these same lithologies, with blocks of the Etivluk and Lisburne groups being particularly common (Fig. 6F). The olistostromes have both mud-rich and sand-rich matrices and are interpreted to record deposition at sites proximal to locations of active thrusting (De Vera, 2005). Silica-cemented quartz-chert sandstones such as those of the Endicott Group in Endicott Mountains allochthon and Picnic Creek allochthon and in the lower Lisburne Group in Kelly River allochthon evidently are poorly represented or absent as clasts in the conglomerates and olistostromes. Outside of the Red Dog Mining district, the Okpikruak Formation mainly consists of thin- to medium-bedded, fine- to medium-grained sandstone turbidites with local conglomerate bodies and mudstone sequences in all of the continental allochthons (e.g., Curtis et al., 1984, 1990; Mull, 1985) (Fig. 6D). Stratigraphic thicknesses up to 3 km are reported (Mayfield et al., 1988), although these sections have few marker beds and may have been duplicated by unmapped faults.

The oldest known Brookian deposits in the western Brooks Range form a thin-bedded sequence of volcanogenic sandstone and mudstone turbidites that is more than 150 m thick exposed in stream cuts along the Kugururok River (Figs. 2C and 6B). Curtis et al. (1984) reported the presence of Late Jurassic Buchia pelecypods and mapped these rocks as “unit Jw” because of uncertainties about their stratigraphic and structural position. Despite local faulting, field relations of these strata suggest to us that unit Jw turbidites most likely rest on mafic rocks belonging to either the Copter Peak allochthon and/or the Misheguk Mountain allochthon.

METHODS

We collected 17 samples of fine- to medium-grained sandstone from representative parts of the Brookian sequence mostly along a northwest-trending transect extending from the structurally highest parts of the fold-thrust belt in the western Brooks Range to the axial part of the Colville Basin along the Chukchi Sea for sedimentary petrography, detrital zircon U-Pb geochronology, and zircon fission-track dating. We also dated two clasts, a quartz-rich sandstone and a granitic rock, from a conglomeratic debris flow in the Okpikruak Formation of the Picnic Creek allochthon (Figs. 1 and 3 and Table 1). Sample description information is provided in the Supplemental File1.

Thin sections of these sandstone samples and two additional samples from closely related sites were stained for potassium feldspar and point counted (n > 400) using the Gazzi-Dickinson counting technique to minimize compositional dependence on grain size (Ingersoll et al., 1984; Zuffa, 1985) (Tables 2 and 3; Supplemental Table SF1 [see footnote 1]). Grains affected by low-grade diagenesis were counted as framework grains if their origin was identifiable; otherwise they were recorded as diagenetic minerals. Volcanic grains were identified by the presence of volcanic texture; possible volcanic grains that lack volcanic texture were counted as chert if silicic in composition or as argillite if mafic and altered to chlorite. Detrital carbonate grains were included as framework grains on ternary diagrams (Ingersoll et al., 1987), but accessory minerals and fossils were excluded. Point count framework grain types and recalculated modal data are presented in Tables 2 and 3 and raw data are shown in Supplemental Table SF1 (see footnote 1).

Zircon ages in 15 sandstone samples and two conglomerate clasts were determined using a laser ablation–inductively coupled plasma mass spectrometer (LA-ICPMS) at the Washington State University Geoanalytical Laboratory in Pullman, Washington. The U-Pb ages of 60–110 zircon grains were determined for each sample. The U-Pb ages of an additional 30 grains in four of the same sandstones and 60 grains from another sandstone were determined using the Agilent 7700× quadrapole of Donelick Properties in Viola, Idaho. Zircon fission-track ages of the same 30 zircons from the four sandstones were also determined, such that a total of 120 grains were double dated by U-Pb and zircon fission-track methods.

Analytical techniques for the U-Pb geochronology and zircon fission-track dating are presented in the Supplemental File and Supplemental Tables SF2, SF3, and SF5 (see footnote 1), and the complete U-Pb geochronologic results are reported in Excel format in Moore (2014). For the purpose of discussion, zircon U-Pb ages determined on the two instruments were combined without bias, resulting in 80–129 zircons being dated from each sample and a total of 1888 zircons dated in this study. Histograms of U-Pb ages, relative probability curves, and weighted mean U-Pb ages were calculated using Isoplot 3.0 (Ludwig, 2005); cumulative probability curves and Kolmogorov-Smirnoff (K-S) statistics (Press et al., 1986; Guynn and Gehrels, 2010) were determined using the plotting program of Gehrels (2010). The geologic time scale of Walker et al. (2013) is used in this paper.

We have organized the samples into five groups based on structural and stratigraphic position for the purpose of interpretation (Table 1). The two samples from unit Jw are grouped together because of their apparent depositional position on the allochthonous oceanic rocks of the Angayucham terrane. The samples from the Ipnavik River allochthon, Kelly River allochthon, and Picnic Creek allochthon are combined together as a “high-allochthon” group, whereas the samples from the Endicott Mountains allochthon and Endicott Mountains allochthon(?) are grouped together as a “low-allochthon” group. For the samples from the Colville Basin, we have followed the classification scheme of DeCelles and Giles (1996) for foreland basin deposits and grouped together the samples from the lower Brookian shale and the Mount Kelly Graywacke on the southern flank of the basin as a “wedgetop” group, whereas the three samples from the Nanushuk Formation are grouped together as a “foredeep” group.

SANDSTONE PETROGRAPHY

Brookian sandstones range in composition from litharenite to feldspathic litharenites, sublitharenites, and calclitharenites. The quartz-rich sandstone clast from the debris flow in the Okpikruak Formation, in contrast, is a subarkose. Framework grains identified and counted in these sandstones include monocrystalline quartz, polycrystalline quartz, chert, plagioclase, potassium feldspar (perthite and microcline), shale, siltstone, extrabasinal carbonate, mafic and intermediate volcanic, metamorphic, serpentinite, diabasic, and granitic lithic grains (Table 2). Interstitial material in most samples is almost entirely pseudomatrix, although zeolite cement is present in samples 06CP-07 and 06TM-08. Quartz overgrowths and cement are common in sample 06TM-29B. Sandstones show evidence of diagenetic alteration including creation of one or more of the following: calcite, clay minerals, quartz, zeolites, and alteration of plagioclase to albite. Sandstone textures and grain compositions are nonetheless typically identifiable, although small patches of samples 06CP-07 and 06TM-08 are overprinted by laumontite to such an extent that grains could not be identified with certainty. These areas were avoided during point counting.

Point-count data presented in Table 3 reveal that lithic grains in Colville foredeep samples are dominantly siliciclastic sedimentary, chert, and detrital carbonate grains, whereas the wedgetop group is dominated by detrital carbonate grains. Coarse white mica and metamorphic lithic grains are also common in the wedgetop deposits. Lithic grains in samples from the Okpikruak Formation in the lower allochthons include abundant siliciclastic sedimentary and chert grains and lesser proportions of detrital carbonate and metamorphic grains, whereas the Okpikruak in higher allochthons is marked by highly variable proportions of chert, siliciclastic sedimentary, detrital carbonate, and volcanic grains. Serpentinite, gabbro, pyroxene, and epidote are common in several of the samples from the Kelly River allochthon and Picnic Creek allochthon, indicating contributions from a nearby ophiolitic source. Lithic grains in unit Jw are almost entirely volcanic fragments. The quartz-rich sandstone clast from the olistostrome in the Okpikruak Formation contains few lithic fragments and is dominated by a mature assemblage of quartz and potassium feldspar.

Figure 7 shows the sandstone petrographic data plotted on ternary diagrams after Dickinson (1985) and Ingersoll and Suczek (1979). We include for comparison data published by previous workers, which were produced using methodologies different from the Gazzi-Dickinson method employed in the present study and are likely to be biased toward somewhat higher proportions of rock fragments.

On the total quartz-feldspar-lithic fragment (QtFL) ternary plot (Fig. 7A), most samples plot along two linear trends, one consisting mainly of the foredeep, wedgetop, and low-allochthon samples having a relatively higher proportion of siliceous grains (line A in Fig. 7A), and the other consisting of the Jw and most high-allochthon samples at relatively higher proportions of feldspar and lithic grains (line B in Fig. 7A). The samples in trend A all plot in the recycled orogenic field, whereas those in trend B plot in the undissected to dissected arc fields. This suggests that two source areas may be represented in the data, one consisting of recycled detritus of sedimentary or metasedimentary origin (trend A) and the other consisting of debris from a volcanic arc complex (trend B). The sandstone clasts in the Okpikruak Formation plot in the craton interior and basement uplift fields in stark contrast to all of the other samples in the study.

On the monocrystalline quartz-feldspar-total lithic fragment (QmFLt) diagram (Fig. 7B), the samples fall along similar trends emanating from the Lt pole. Most of the samples from the low allochthons and foredeep and wedgetop deposits of the Colville Basin plot in the lithic recycled, transitional recycled, and mixed recycled-dissected arc fields along a trend of decreasing lithic grains (trend A), whereas those from most of the higher allochthon and the unit Jw samples plot across the transitional arc and dissected arc fields (trend B). As in the QtFL diagram, the clasts from the Okpikruak conglomerates plot separately in the craton interior and basement uplift fields.

On the LvLmLs plot, the samples from Colville Basin and the lower allochthons fall in the “rifted continental margins” field near the sedimentary (Ls) pole on the lithic grain diagram (Fig. 7C), whereas those from the higher allochthons and Jw plot mostly at higher proportions of volcanic component (Lv) in or near the “mixed magmatic arcs and rifted continental margins” field. The samples from unit Jw plot very close to the “magmatic arcs” field, indicating that they have volcanic compositions close to those of magmatic arcs.

The monocrystalline quartz-plagioclase-feldspar (PQmK) ternary diagram (Fig. 7D) demonstrates that most Colville Basin and Okpikruak Formation samples have little to no potassium feldspar, with only three samples containing more than 1%.

These diagrams show that most samples from the Okpikruak Formation in the higher allochthons and unit Jw are relatively enriched in volcanic and felsic components, indicating strong contributions from the Angayucham and Koyukuk terranes, which together form the upper plate of the Brookian orogen. Samples from the lower allochthons and most of those from the Colville Basin, on the other hand, contain a much lower proportion of arc material and consist instead mainly of mixtures of continental and recycled grain types.

The overall pattern suggests that the Late Jurassic to Albian part of the Brookian sequence of the western Brooks Range represents an unroofing sequence. The oldest deposits, consisting of the syntectonic deposits of the Okpikruak Formation and unit Jw in the higher allochthons, represent erosion primarily of the Angayucham and Koyukuk terranes in the structurally higher parts of the Brookian orogen. The Okpikruak sandstones in the lower allochthons and most debris in the Colville Basin, on the other hand, consist mostly of debris derived from a sedimentary or metasedimentary orogenic area, which is consistent with derivation from the Paleozoic and older rocks such as those in the Arctic Alaska–Chukotka microplate.

DETRITAL ZIRCON U-PB GEOCHRONOLOGY

Brookian Sandstones

Zircons dated in the Brookian sandstones range in age from 3116 Ma to 111 Ma (Table 4). Plotted graphically (Fig. 8), the zircon ages fall into three major age populations, one ranging from 180 to 140 Ma (Jurassic and earliest Cretaceous), another from 542 to 200 Ma (Paleozoic and Triassic), and the third from 2000 to 1750 Ma (Paleoproterozoic). Modest, broadly distributed populations of Neoproterozoic and Archean zircons (900–600 Ma and 2800–2400 Ma, respectively) can also be recognized in the data set. The Paleozoic–Triassic population is the largest and displays a broad asymmetric distribution with peaks at 329 Ma and 237 Ma. Because of the regional sub-Mississippian unconformity in the Arctic Alaska–Chukotka microplate, we divided the Paleozoic–Triassic population into an early Paleozoic (542–359 Ma) subpopulation and a late Paleozoic and Triassic (359–200 Ma) subpopulation. This division also separates the younger, most voluminous part of the Paleozoic–Triassic population from its older, long tail composed of zircons that decrease in abundance with increasing age. Ages of the four resulting population groupings are shown as color bands on Figures 8–12.

Age distributions of zircons in individual samples (Table 4 and Supplemental Fig. SF1 [see footnote 1]) reveal that all 15 sandstone samples consist dominantly of some or all of the population groups defined above (Fig. 9). Most samples contain few or no zircons that approximate their stratigraphic age. Exceptions to this include unit Jw, which contains abundant Late Jurassic grains as young as 148 Ma (Tithonian) at the Tithonian Buchia fossil locality reported by Curtis et al. (1984) (sample 06CP-07). In addition, samples from the Nanushuk Formation at Corwin Bluff contain sparse zircons that may approximate the stratigraphic age of this unit, including a Barremian zircon (128 ± 4 Ma) in sample TM-01 and three Aptian zircons (weighted mean average = 115 ± 4 Ma) in sample TM-02. These ages are consistent with the 120 Ma age for the Nanushuk at Corwin Bluff estimated from seismic data by Schenk et al. (2012) (Fig. 5), but are significantly older than the mainly middle and late Albian age estimated from plant fossils in the section (Smiley, 1969; Spicer and Herman, 2010) (see the Supplemental File [see footnote 1] for details about the stratigraphic positions of these samples).

Other exceptions are the lower Brookian shale from the base of the southern flank of the Colville Basin, which contains Barremian zircons (sample 04TM-141, five grains, weighted mean average = 128 ± 4 Ma) and allochthonous Brookian strata from the Lisburne Peninsula, which contain a Valanginian zircon (sample 94TM-47, 135 ± 3 Ma). The Barremian zircons in the lower Brookian shale are younger than the Hauterivian age reported by Mull (1995) from paleontologic data. With the obvious exception of unit Jw, we conclude that lower Brookian strata received only small contributions of zircons from contemporaneous arc volcanic sources until at least the Aptian.

Aside from a single spike in zircon ages at 155 Ma (i.e., the Jurassic and earliest Cretaceous population), there are few other peaks in zircon ages that are present in most or all samples. This could result from the small statistical size (∼100 grains) of zircon populations analyzed from Brookian deposits. To increase the statistical size of the various zircon populations and in an attempt to reveal regional trends by age and structural-stratigraphic position, we grouped samples into the five structural-stratigraphic sample groups distinguished earlier in this paper (Figs. 10 and 11).

Samples from unit Jw consist almost exclusively of Jurassic and earliest Cretaceous zircons (89%, Table 4) that range in age from 180 to 140 Ma with a maximum probability of 155 Ma (Fig. 10). Sample 06TM-08 (but not sample 06CP-07) also contains a significant number of younger zircons with a peak at 129 Ma (Barremian) (not shown in Fig. 8). Nearly all of these grains have high U contents (>3000 ppm) and therefore their ages are judged to be spurious due to Pb loss. These zircon ages consequently are omitted from further consideration in this report (see Supplemental File [see footnote 1] for a discussion of these grains). Pre-Jurassic zircons in the unit Jw sandstones include Silurian to Carboniferous (435–300 Ma), early and middle Neoproterozoic (800–600 Ma), and early and middle Paleoproterozoic (2100–1750 Ma) zircons, but only the Paleoproterozoic zircons fit one of the zircon population groups defined in Figure 8. The high-allochthon group also contains zircons of the Jurassic and earliest Cretaceous population (13%), but zircons of this age decrease in abundance in the low allochthons and in the wedgetop and foredeep groups (both 4%–5%, Table 4 and Fig. 11). A notable exception to this trend is sample 03CP-25 from the Ipnavik River allochthon of the high-allochthon sample group, which yielded no Jurassic zircons. Volcanic detritus in thin sections from this locality suggest this sample was sourced mainly from zircon-poor mafic volcanic rocks and chert of the Copter Peak allochthon rather than from the Misheguk Mountain allochthon and Koyukuk terranes, which contain more zircon-bearing rocks. More analyses of sandstones from the Ipnavik River allochthon are required to assess the significance of the absence of Jurassic and earliest Cretaceous zircons in this allochthon. The overall pattern is that zircons of Jurassic and earliest Cretaceous age decrease in abundance in structurally lower and younger Brookian deposits.

Archean (>2.5 Ga) and Paleoproterozoic (1.6–2.5 Ga) zircons, on the other hand, increase in abundance from 23% in the high allochthon sample groups to ∼35% of the wedgetop and foredeep groups (Table 4). The maximum probability in the Paleoproterozoic zircon population occurs at ca. 1.8 Ga in the allochthon and wedgetop groups but is somewhat older (1.9 Ga) in the foredeep group (Fig. 11). Well-defined Paleoproterozoic and Archean zircon age peaks occur at 2.46 Ga in the low allochthons and 2.68 Ga in the foredeep sample group, but such a peak is not present in the high allochthons. Although they are recognized in some samples (e.g., 10TM-82B, 10TM-78D, and TM-01, Supplemental Figs. SF1–SF4 [see footnote 1]), zircons of 2.0–2.1 Ga age are not common in any of the structural-stratigraphic groups, indicating that the distinctive arkosic middle Paleozoic sandstones of the Nuka Ridge allochthon did not form a significant source for Brookian deposits in the western Brooks Range.

In contrast to the other zircon population groups, the Paleozoic–Triassic population displays few obvious trends by structural and stratigraphic position. Early Paleozoic zircons decrease in abundance from the high-allochthon sample group (23%) to the foredeep sample group (13%), with peaks in maximum probability occurring at ca. 425–415 Ma in the wedgetop and foredeep sample groups and smaller peaks at 480–460 Ma in the allochthon sample group. Triassic and late Paleozoic zircons have significant peaks at ca. 310–345 Ma in all of the sample groups and ca. 230–240 Ma in all but the high-allochthon sample group. The high-allochthon sample group is the one sample group (aside from unit Jw) in which Triassic zircons are not common.

In order to evaluate the similarity of the zircon populations, probability (P) values were calculated from K-S statistics for the individual samples from the allochthon, wedgetop, and foredeep sample groups. K-S statistics test the hypothesis that two detrital zircon age populations were not derived from the same parent population (Guynn and Gehrels, 2010). For cases in which there is a greater than 95% probability that two populations were not derived from the same parent population or share the same sediment source P < 0.05, whereas for two populations having statistically identical populations P = 1. The K-S statistics for all of the sandstone samples we dated are shown in Table 5. Zircons that are Jurassic or younger have been omitted because K-S statistics are sensitive to variations in the abundances of zircons of the same age and the previously noted variations in the abundance of the Jurassic and Cretaceous zircons could mask other trends that might be present.

For Triassic and older zircons, 47 of the sample pairs have P values >0.05. Only sample 03CP-25 from the Ipnavik River allochthon has P < 0.05 for most sample pairs, suggesting that this sample is dissimilar to the others and might have a somewhat different provenance. This might be explained by its lack of Triassic zircons and a peak in the Paleoproterozoic at 1.66 Ga (Supplemental Fig. SF4 [see footnote 1]), an age younger than the range of the Paleoproterozoic zircon population group defined above. If this sample is omitted from consideration, 65% of the sample pairs have P > 0.05 and 8% have P > 0.9, indicating a significant to very high probability that most of the samples were derived from the same Triassic parent population or the same Triassic sediment source area.

Similarities in the zircon populations are evident in the cumulative probability plots shown in Figure 12. These plots also show that most of the variation between samples is due to variation in the abundance of Phanerozoic and Paleoproterozoic zircon populations and that Neoproterozoic and Mesoproterozoic zircons are relatively minor components. Exceptions include samples 94TM-47 (Endicott Mountains allochthon) and 04TM-143 (wedgetop), which display steeper slopes between ca. 800–1500 Ma. Although having the unusual characteristics described above, sample 03CP-25 (Ipnavik River allochthon) displays a cumulative probability curve that is generally similar to those of the other samples from the Okpikruak Formation, suggesting that its origin may reflect local disparities in the geology of the source region rather than an entirely different provenance area. A general shift to higher proportions of Paleoproterozoic zircons in the samples from the Colville Basin relative to the samples from the Okpikruak Formation (Fig. 12B) could indicate erosion in the source area toward older rocks with time. The two samples from unit Jw, on the other hand, clearly reflect a different source area because of the dominance of Jurassic and earliest Cretaceous zircons in these rocks (Fig. 12A).

Conglomerate Clasts

A cobble of quartz-rich subarkosic sandstone and a boulder of a granodiorite were collected for U-Pb age dating from a conglomeratic debris-flow deposit at location 06TM-29 in the Picnic Creek allochthon. In the field, these clasts were hypothesized to have come from basement rocks and overlying lower Mississippian clastic rocks (Endicott Group) of the Arctic Alaska terrane in the structural core of the Brooks Range.

Detrital zircon U-Pb ages from the quartz-rich sandstone clast (sample 06TM-29B) reveal a dominant population of zircons distributed between 415 and 360 Ma with a peak in probability at 412 Ma and a bimodal subordinate Paleoproterozoic population having peaks at 1771 and 1938 Ma (Figs. 13A and 13B). The sample contains only one zircon younger than 360 Ma (280 Ma), whose age we suspect has been altered due to Pb loss prior to deposition because of its high U content (>4500 ppm). For this reason, the youngest zircon age is disregarded and the depositional age of the rock is considered to be Mississippian or younger (<359 Ma). The distribution of zircon ages for this clast is similar to that of the Devonian and older zircons in the sandstone (sample 06TM-29A) collected from the same locality (Table 6; Supplemental Fig. SF1 [see footnote 1]). This suggests that the quartz-rich sandstone clast may be representative of the provenance for the Devonian and older zircons in most of the Brookian sandstones and that those zircons probably were recycled from older sedimentary rocks.

The zircon ages from the granitic clast (sample 06TM-29D) fall between 170 and 150 Ma and yield a weighted mean age of 159.8 ± 0.6 Ma (Late Jurassic) (Figs. 13C and 13D). This age is coincident with many of the zircon ages of unit Jw and indicates that the granodiorite was likely derived from a granitic body in the Angayucham or Koyukuk terranes rather than the basement rocks of the Arctic Alaska–Chukotka microplate.

ZIRCON FISSION-TRACK AND U-PB DOUBLE DATING

The thermochronologic history of a sandstone in the source region sometimes can be constrained by dating the same zircon grains using both U-Pb and zircon fission-track (ZFT) methods (i.e., “double dating”). The U-Pb isotopic system closes at temperatures of >700 °C in most natural igneous and metamorphic rocks (Hanchar and Watson, 2003), whereas the ZFT system closes at temperatures of ∼200–250 °C (Lee et al., 1997; Bernet, 2009). If U-Pb and ZFT ages are plotted against each other, zircon grains that cooled rapidly at the time of crystallization should have nearly identical (i.e., concordant) U-Pb and ZFT ages within analytical uncertainty (commonly substantial for ZFT ages) and thus fall on a line with a slope equal to 1 (Fig. 14). At the other extreme, zircons that reside at depths with temperatures >250 °C for a significant period after crystallization and later cool below the closure temperature for ZFT will fall on lines having slopes less than 1. If the zircon grains of a sample have a variety of crystallization ages (as in many sandstones) and the sample cooled through the ZFT closure temperature at a significantly later time, then the zircons will all plot at the same ZFT age regardless of the U-Pb age of the individual zircons (i.e., at a slope of zero). In that situation, if all of the zircons have ZFT ages that are greater than the depositional age of the sandstone, then the cooling must have occurred in the source area; whereas, if the ZFT age is younger than the depositional age, then the sandstone itself cooled through the closure temperature for ZFT following deposition.

Thirty zircon grains from each of four of our samples were dated using both the U-Pb and ZFT methods (Supplemental Table SF5 [see footnote 1]) and the ages plotted against each other in Figure 14. These include two samples from the upper allochthons (samples 03CP-25 from the Ipnavik River allochthon and 10TM-78D from the Picnic Creek allochthon, respectively), one sample from the wedgetop deposits (sample 04TM-141), and one from the foredeep deposits (sample TM-02).

All four samples display generally similar distributions on ZFT versus U-Pb cross plots, indicating that the ages do not vary by the amount of stratigraphic and/or tectonic burial. In addition, all of the samples contain many grains that have ZFT ages older than the depositional age, suggesting that some ZFT ages are provenance ages. In all samples, the zircons with Paleoproterozoic U-Pb ages yield ZFT ages that are significantly younger than their U-Pb crystallization ages, with most of the ZFT ages falling between 100 and 250 Ma and all being younger than 450 Ma. This suggests that the Paleoproterozoic grains cooled through the closure temperature for ZFT and were mostly to fully annealed prior to deposition, many probably in the Mesozoic at the time of the Brookian orogeny, shown by a green band at ZFT cooling ages of 160 to ca. 100 Ma, although an older period(s) of cooling such as during the Uralian-Taimyr orogeny (orange band) is also permissible due to the U-Pb crystallization ages and the large uncertainties of ZFT ages.

To better resolve the timing of cooling and to verify that the source area of the Brookian sandstones experienced Brookian cooling, we plotted the ZFT and U-Pb data on relative probability plots (Fig. 14). Only double-dated zircons having Triassic and older (>200 Ma) U-Pb ages are included in these plots because younger grains could not show evidence of pre-Brookian cooling events. The plots clearly show that the time of maximum probability of cooling is considerably younger than the U-Pb crystallization ages of the zircons and occurred ∼20 m.y. or less before the time of deposition of the samples. For three of the four samples, the time of peak cooling corresponds with the time of the Brookian deformational event. All of the samples, however, have asymmetric ZFT cooling curves with longer tails toward older ages and central (mean) cooling ages that are displaced to older ages from their respective peaks of maximum probability. This suggests that the ZFT data may reflect a mixing of a subordinate component of unannealed or partly annealed zircons having late Paleozoic and Triassic ZFT ages with the dominant population of Brookian ZFT ages. The somewhat older peak cooling age in sample 03CP-25 indicates that a relatively larger number of Paleozoic and Triassic cooling ages are present in this sample, causing the time of maximum probability of cooling to reflect an intermediate age. Analysis of the ZFT data using a deconvolution program supports the interpretation that the samples cooled initially in the Pennsylvanian, Permian, and Triassic at the time of the Uralian-Taimyr orogeny and later during the Late Jurassic and Early Cretaceous at the Brookian orogeny (Fig. 14).

The zircon thermochronologic data indicate that the Brookian deposits were derived after 7–10 km of exhumation (assuming a geothermal gradient of 25–35 °C/km) from either (1) a pre-Jurassic source area having a complex history of multiple cooling events including protracted cooling that lasted into the Early Cretaceous (i.e., first-cycle sedimentary deposits) or (2) a recycled sedimentary source area consisting of debris shed into a basin from a Paleozoic and Triassic deformational belt and then uplifted, eroded, and redeposited in the Early Cretaceous as part of the Brookian deformation event (i.e., second-cycle sedimentary deposits).

WERE LOWER BROOKIAN SANDSTONES SOURCED FROM THE PALEOZOIC AND OLDER PLATFORM COVER AND BASEMENT OF THE ARCTIC ALASKA–CHUKOTKA MICROPLATE?

The presence of carbonate, metamorphic, and volcanic lithic grains, coupled with evidence for paleoflow from the west and southwest, have led to the interpretation that the foreland basin deposits of the western Colville Basin were derived from highlands in the Lisburne Peninsula, Chukchi Sea, and Chukotka (Molenaar, 1988; Mull, 1985; Bartsch-Winkler and Huffman, 1988; Molenaar et al., 1988). Olistoliths and conglomerate clasts of Devonian and Mississippian carbonate rocks, chert, mafic igneous rocks, and granitic rocks in the Okpikruak Formation point to local source areas in nearby allochthons of the Brooks Range thrust belt (Mull, 1985; Crane, 1987). Glaucophane, chloritoid, garnet, and white mica in the Nanushuk Formation and metamorphic lithic grains and white mica in the Mount Kelly Graywacke have been interpreted as evidence that the metamorphic core of the Brooks Range in the southern Brooks Range and Seward Peninsula may have been a source area for Brookian sandstones (Mull, 1985; Till, 1992).

Our compositional data suggest that an unroofing succession exists in Brookian sandstones with detritus shed from the Angayucham and Koyukuk terranes being most abundant in Brookian deposits from the oldest and structurally highest allochthons, whereas detritus shed from older and more quartz-rich sources becomes increasingly abundant in the younger Brookian deposits of structurally low allochthons and the Colville Basin. The shift to higher proportions of older zircon grains in younger Brookian sedimentary strata appears to support this interpretation. An unroofing succession with these characteristics implies that the structural cover and basement of the Arctic Alaska–Chukotka microplate may have become increasingly exhumed and eroded as deposition in the Brookian foreland basin progressed. The detrital zircon data can be used to test this idea.

An important finding of this study is that detrital zircon age populations of the Brookian deposits from all but unit Jw contain abundant late Paleozoic and Triassic zircons. Zircons of this age are unlikely, however, to have been derived from older rocks in northern Alaska, because during the late Paleozoic this area was characterized by deposition of strata in a starved proximal to distal passive continental margin setting (e.g., Moore et al., 1994). These strata (the Ellesmerian sequence) were deposited during a 200 m.y. period of tectonic quiescence that extended from the latest Devonian to the Middle Jurassic. In Chukotka and Wrangel Island, the period of tectonic quiescence ended with deposition of lithic-rich turbidites and intrusion of mafic sills in the Triassic. These observations indicate that zircon-rich late Paleozoic and Triassic igneous or metamorphic rocks are unlikely to ever have existed in the Arctic Alaska–Chukotka microplate and that these zircons instead must have had their ultimate origin in a source area outside of the Arctic Alaska–Chukotka microplate.

Although stratigraphic relations indicate that the ultimate source area for the Upper Paleozoic and Triassic zircons lies outside of the Arctic Alaska–Chukotka microplate, the origin of Devonian and older zircons is less certain because pre-Jurassic strata in the Arctic Alaska–Chukotka microplate include abundant zircons of this age (e.g., Amato et al., 2009; Miller et al., 2010; Moore, 2010; Dumoulin et al., 2013). To test whether lower Paleozoic and older rocks of the Arctic Alaska–Chukotka microplate might have supplied Devonian and older zircons to the Brookian deposits, we compare zircons older than 359 Ma (end of Devonian) in the Brookian sequence to pre-Mississippian zircons from prospective source rocks in the Arctic Alaska–Chukotka microplate on a cumulative probability plot (Fig. 12C). The Arctic Alaska–Chukotka microplate samples are from various structural positions including: (1) parautochthonous upper Paleozoic rocks from the frontal part of the Brookian thrust belt in the Lisburne Peninsula and Wrangel Island (Miller et al., 2010); (2) Mississippian clastic rocks from the Ellesmerian sequence in the Endicott Mountains allochthon in the central part of the Brooks Range (Dumoulin et al., 2013), and (3) middle Paleozoic and older rocks from the metamorphic hinterland of the orogen in the Seward Peninsula (Amato et al., 2009). Detrital zircon ages of the Paleozoic rocks sampled on Wrangel Island are representative of parautochthonous basement rocks present across a wide area of Chukotka (Gottlieb et al., 2012), and detrital zircon ages of Mississippian clastic rocks from the Endicott Mountains allochthon are representative of the zircon ages of upper Paleozoic strata in the other allochthons of the western Brooks Range (Moore, 2010). Detrital zircon ages from the Seward Peninsula samples are similar to those in the metamorphic core of the orogen in the southern Brooks Range (Amato et al., 2009). Also included in Figure 12C are zircon data from the quartz-rich sandstone clast in the Okpikruak Formation of the Picnic Creek allochthon (sample 06TM-29B) and the detrital zircon data of Wartes (2008) from the Brookian foredeep and wedgetop deposits in the central Brooks Range.

The light-shaded field in Figure 12C shows the distribution of the cumulative curves for Okpikruak and Colville Basin samples. The plot affirms that these samples are enriched in early Paleozoic and Paleoproterozoic zircons and relatively depleted in late Neoproterozoic and Mesoproterozoic zircons. In contrast, zircons from the samples from the Paleozoic and older rocks of the Arctic Alaska–Chukotka microplate define a field (dark shaded in Fig. 12C) that is enriched in Neoproterozoic and Mesoproterozoic zircons and relatively depleted in Paleoproterozoic zircons. The contrast between these fields suggests that Devonian and older zircons from early Brookian sandstones in the western Brooks Range were not derived from lower Paleozoic and older source areas in the Arctic Alaska–Chukotka microplate.

This conclusion is substantiated by K-S statistics for pre-Mississippian zircons in the Okpikruak Formation and Colville Basin and the possible source areas of similar or older age in the Arctic Alaska–Chukotka microplate in Alaska and Chukotka (Table 6), which reveal a high probability of dissimilarity (i.e., P < 0.05) in most Brookian and Arctic Alaska–Chukotka microplate comparisons. Samples 94TM-47 from the Lisburne Peninsula and to a lesser extent 04TM-143 from the Mount Kelly Graywacke have transitional characteristics that suggest they could include some sediment derived from the Arctic Alaska–Chukotka microplate. The quartz-rich sandstone clast from the Picnic Creek allochthon displays a zircon population that is dissimilar to those of the Paleozoic and older Arctic Alaska–Chukotka microplate samples (Fig. 12C) and is instead more similar to those of the lower Brookian sandstones (e.g., the high-allochthon group in Table 6), suggesting that it might share the same provenance.

Detrital zircon data of Wartes (2008) provide an opportunity to examine whether the Devonian and older zircon populations in the western Brooks Range extend into the central Brooks Range, ∼500 km to the east. Wartes (2008) reported detrital zircon analyses from two locations in the Nanushuk Formation and three locations in the Fortress Mountain Formation, representing foredeep and wedgetop stratigraphic positions, respectively. Nanushuk deposits in the central Brooks Range are younger than those from the western Brooks Range because deposition in the Nanushuk was strongly progradational from west to east (Fig. 5). The Fortress Mountain in the central Brooks Range is clearly stratigraphically older than the Nanushuk in the same area (Wartes, 2008), but the ages of these deposits relative to the lower Brookian samples in the western Brooks Range are uncertain.

A composite cumulative probability curve for the Fortress Mountain Formation (wedgetop) samples from the central Brooks Range (Fig. 12B) shows that they are quite similar to the Brookian samples from the western Brooks Range, whereas the composite curve for the Nanushuk Formation foredeep samples in the central Brooks Range is less segmented and quite different from the western Brooks Range samples due to differences in the relative abundances of their Paleozoic and Proterozoic zircon populations.

The cumulative probability curves for Devonian and older zircons (Fig. 12C) show that the Fortress Mountain Formation (wedgetop) samples from the central Brooks Range plot near the average of the Brookian sandstones from the western Brooks Range. This suggests that the Fortress Mountain Formation of the central Brooks Range had the same provenance as the Brookian deposits of the western Brooks Range or could have been recycled from the erosion of older Brookian deposits (i.e., the Okpikruak Formation). The Nanushuk Formation foredeep deposits of the central Brooks Range, in contrast, display a steep slope that plots in the field of the Arctic Alaska–Chukotka microplate samples in Figure 12C. This suggests that Nanushuk deposits in the central Brooks Range may have been derived from erosion of the Paleozoic and older rocks of the Arctic Alaska–Chukotka microplate.

These interpretations are supported by K-S statistics. Table 6 compares pre-Mississippian zircons in the samples of Wartes (2008) from the central Brooks Range with samples from various parts of the Arctic Alaska–Chukotka microplate as well as the lower Brookian deposits of the western Brooks Range and to samples from possible source areas in the Arctic Alaska–Chukotka microplate. The K-S statistics suggest similarities between the foredeep samples (i.e., Nanushuk Formation) of the central Brooks Range and possible source areas in Arctic Alaska–Chukotka microplate basement and Paleozoic cover rocks in Wrangel Island, the Lisburne Peninsula, and in the thrust sheets of the Brooks Range (P = 0.21, 0.49, and 0.6, respectively) but are consistently dissimilar to Devonian and older zircons in the lower Brookian deposits of the western Brooks Range (P < 0.05). The pre-Mississippian zircons from the wedgetop deposits (Fortress Mountain Formation) in the central Brooks Range, in contrast, are dissimilar to any of the possible source localities in the basement and Paleozoic cover of the Arctic Alaska–Chukotka microplate (P < 0.05). Instead, they have similarities to the pre-Mississippian zircons in the Okpikruak Formation of the lower allochthons (P = 0.59) and wedgetop deposits (P = 0.29) of the western Brooks Range.

The foredeep samples of Wartes (2008) from the central Brooks Range were collected from proximal parts of the Umiat Delta, which is highly enriched in phyllite and schist fragments (Bartsch-Winkler and Huffman, 1988). In contrast, Bartsch-Winkler and Huffman (1988) noted that sandstones of the Corwin Delta contain few of these grain types and are instead sedimentary litharenites. Till (1992) showed that sandstones having metamorphic lithic compositions similar to those from the Umiat Delta also yield detrital glaucophane in heavy-mineral separates, whereas glaucophane was not reported in heavy-mineral separates from Corwin Delta sandstones. She suggested that the first appearance of detrital glaucophane in Brookian sandstones marks the time of onset of erosion of the core of the Brooks Range and that this change in heavy-mineral compositions occurred during the mid-Albian. Schenk et al. (2012) estimated from seismic data that deposition of strata in the central Colville Basin near the location of the Umiat Delta occurred at ca. 110–105 Ma (late early to early late Albian; Ogg and Ogg, 2008), whereas two zircons in the samples dated by Wartes (2008) yield ages of 90–89 Ma (Turonian), suggesting that the change in provenance may have occurred as late as Late Cretaceous time. These observations suggest that the Arctic Alaska–Chukotka microplate did not begin contributing detritus into Brookian sandstones before the mid-Albian.

This interpretation is supported by Ar-Ar cooling ages from metamorphic rocks in the present-day southern Brooks Range, which indicate that, following high-pressure metamorphism at depths of 10–40 km in the Early Cretaceous, these rocks passed through closure temperatures for hornblende (∼500 °C) at 105–103 Ma and white mica (350–420 °C) at 100–90 Ma (Vogl et al., 2002). These cooling ages indicate that the onset of uplift-related cooling in the metamorphic core of the range approximately coincided with the Albian time of appearance of zircons of Arctic Alaska–Chukotka microplate affinity in the Colville Basin foredeep. Cooling in the metamorphic rocks of the southern Brooks Range at this time has been ascribed to postcollisional extensional exhumation (e.g., Miller and Hudson, 1991; Vogl et al., 2002).

PROVENANCE OF EARLY BROOKIAN SANDSTONES

The abundance of Jurassic zircons in the lower Brookian strata and ophiolitic detritus (clinopyroxene, epidote, serpentinite, and amphibole) in the higher allochthons (Kelly River allochthon and Picnic Creek allochthon) strongly support the interpretation that the lower Brookian strata were at least partially sourced from the Angayucham terrane. Dikes and stocks of plagiogranite and biotite tonalite are locally present in the ophiolitic sequences in the Brooks Range, and similar rock types could have been the source for the boulder of 160 Ma granodiorite (sample 06TM-29D) in the Picnic Creek allochthon. Ophiolitic and arc strata of similar age reportedly form klippen at structurally high levels in eastern Chukotka (Vel’may terrane) and are present along the South Anyui zone (Fig. 1) at the southern margin of Chukotka (Sokolov, 2010), providing evidence that thrust sheets of arc-related Jurassic oceanic rocks may have once extended across much or all of the Arctic Alaska–Chukotka microplate.

Provenance of the pre-Jurassic zircons in the lower Brookian sandstones is more difficult to determine. In addition to the detrital zircons in these sandstones, there are several other attributes of the lower Brookian deposits that are relevant to understanding the provenance of this detritus. These are: (1) lower Brookian deposits contain sparse olistostromes and conglomeratic intervals that include chert, carbonate, mafic igneous, granitic, quartz-bearing schist, and quartzite clasts ranging up to hill size (e.g., Chapman and Sable, 1960; Mull, 1985) and that point to nearby source areas; (2) wedgetop deposits in the central Brooks Range region contain locally abundant tonalite and silicic volcanic cobbles that yield U-Pb ages of 274–253 Ma and suggest a proximal source of Permian magmatic rocks (Wartes et al., 2006); (3) a sandstone from the Okpikruak Formation in the central Brooks Range contains detrital white mica that yields Pennsylvanian 40Ar/39Ar ages (Toro et al. 1998) and crossite that is likely to have been derived from a non–Arctic Alaska–Chukotka microplate source area (Till, 1992); and (4) the quartz-rich sandstone clast from a debris-flow deposit in the Picnic Creek allochthon (sample 06TM-29B), which suggests that many Devonian and older zircons in the Brookian sequence may have been derived from recycled middle Paleozoic deposits having a different provenance than those of the Arctic Alaska–Chukotka microplate. Finally, the ZFT thermochronology reported here coincides with the times of both Uralian-Taimyr and Brookian deformation.

Rocks in three source areas could explain some or all of these observations. These source areas include: (1) the Triassic turbidite sequence that rests on the Arctic Alaska–Chukotka microplate in Chukotka, (2) the accreted continental, arc, and oceanic terranes of the South Anyui-Angayucham oceans, and (3) the synorogenic deposits of western Chukotka. Evidence for and against these scenarios is discussed below.

Triassic Turbidite Sequence of the Russian Far East

A sequence of fine-grained, thin-bedded Triassic turbidites of Triassic age was deposited across Carboniferous and Permian carbonate platform sequences in the Arctic Alaska–Chukotka microplate in the Russian Far East (Sokolov et al., 2002; Tuchkova et al., 2009; Sokolov, 2010) (Fig. 1). These deposits have low sandstone-shale ratios and consist of prodelta sediments, distal turbidites, and contourites (Tuchkova et al., 2009). Sedimentation in the Early Triassic was accompanied by intrusions of mafic dikes and sills that suggest extensional tectonics (Miller et al., 2006) followed later in the Triassic by deposition of passive margin strata (Sokolov et al., 2009). Sandstones in the sequence are micaceous, lithic-rich, feldspathic quartz arenites that were shed from metamorphic and volcanic source areas (Tuchkova et al., 2009). These rocks are generally weakly metamorphosed, tightly folded, and record as many as three periods of deformation during the Jurassic and Early Cretaceous (Sokolov et al., 2009). Paleoflow is thought to be from the north (Sokolov, 2010), although some uncertainty exists due to the complex deformation (S.D. Sokolov, 2013, written commun.). Reported thicknesses for the turbidite sequence range from 0.8 to 1.5 km, and structural thicknesses of up to 5 km are estimated for the unit in central Chukotka (Kos’ko et al., 1993; Miller et al., 2006, 2010; Tuchkova et al., 2009).

Detrital zircon U-Pb analyses from sandstones of the Triassic turbidite unit in western Chukotka, Wrangel Island, and a thin unit of compositionally similar shelf sandstones in the Lisburne Peninsula (Moore et al., 2002) reveal generally similar distributions of zircon ages that feature significant populations at 350–230 Ma, 500–400 Ma, and a smaller, variable population at 2.0–1.7 Ga (Miller et al., 2006, 2010). On the basis of these ages, Miller et al. (2006) suggested that the source area for the Triassic turbidites was in the Ural Mountains, Taimyr, and the Siberian traps volcanic field and that the sediment was transported through fluvial drainages into deep-marine depocenters along the southern margin of the Arctic Alaska–Chukotka microplate prior to opening of the Arctic Basin.

Figure 12D shows a cumulative probability plot that compares the ages of Triassic and older zircons from the lower Brookian sandstones of the western Brooks Range to those in the Triassic turbidites in western Chukotka, Wrangel Island, and the Lisburne Peninsula. The diagram shows that ages of the Triassic and older zircons from lower Brookian sandstones are quite similar to those in the Triassic turbidites, varying mostly in the relative proportions of Phanerozoic and Paleoproterozoic zircons. The detrital zircon ages from the Okpikruak Formation in the Brooks Range are particularly similar to those in the Triassic turbidites from Wrangel Island. Although representing similar zircon populations, the sandstones from foredeep and wedgetop deposits in the western Colville Basin are richer in Paleoproterozoic zircons than the Triassic turbidites, indicating that these sandstones have a similar but somewhat different provenance.

Table 5 presents K-S statistics for zircon populations from Triassic turbidites in Chukotka, Wrangel Island, and the Lisburne Peninsula compared to our data from Triassic and older zircons from the Colville Basin and Okpikruak Formation in the western Brooks Range. For all Triassic deposits, the comparison yields values of P > 0.05 for 49% of the comparisons against the Brookian samples. For just the Wrangel Island data, P > 0.05 for 62% of the comparisons with the Brookian samples. (Table 5). These results support the interpretation that many or most zircons in the lower Brookian deposits could have been derived from the Triassic turbidite unit, especially in the Wrangel Island area.

If the Brookian deposits in the Brooks Range were sourced from the Triassic turbidite unit and that unit is composed of detritus shed from earlier erosion of the Uralian-Taimyr orogenic belt as proposed by Miller et al. (2006) and supported by our ZFT thermochronologic data, then it is not possible that the dominant, Jurassic and Early Cretaceous, episode of cooling occurred in the Uralian-Taimyr source area. Instead, detritus from the Uralian-Taimyr source area must have cooled and been uplifted by the Early Triassic when it was eroded and deposited into the Triassic turbidite basin. Following deposition, a second period of cooling occurred during the Late Jurassic and Early Cretaceous Brookian orogeny. This would indicate that the late Paleozoic and Triassic zircon grains in the Brookian deposits of the western Brooks Range are second-cycle sedimentary detritus (Fig. 15).

Accreted Continental, Arc, and Oceanic Terranes of the Angayucham and South Anyui Oceans

The record of the closure of the South Anyui and Angayucham oceans is preserved in the South Anyui zone in the Russian Far East, where previously assembled late Paleozoic and Triassic ophiolite and volcanic arc and continental terranes of Siberian affinity (e.g., the Omolon terrane) form the basement for a Late Jurassic and Early Cretaceous arc (the composite Alazeya-Oloy arc terrane) that was thrust northward onto the Arctic Alaska–Chukotka microplate during the Late Jurassic and earliest Cretaceous (Sokolov, 2010). The thrust sheets of the upper plate of the orogen hold a variety of late Paleozoic and early Mesozoic igneous rocks that could have provided the source for the zircon populations in the Brookian deposits. The early Paleozoic and older zircons in the Brookian deposits may have been derived from erosion of the continental terranes that form the basement of some of the arc terranes, whereas ophiolitic and volcanic clastic detritus in the sandstones and conglomerates in the Picnic Creek allochthon, Kelly River allochthon, and Ipnavik River allochthon could have been derived from arc and forearc terranes in the upper plate.

If the primary provenance for lower Brookian deposits lay entirely in the upper plate rocks, however, then the composition of the resulting Brookian deposits would be expected to vary spatially and temporally depending on the volume and extent of the specific terranes that were available to erosion. Instead, the generally constant and broad distributions of sandstone composition and detrital zircons suggest recycling of a largely sedimentary provenance area. In addition, a continental terrane large enough to contribute most or all of the Paleozoic and Precambrian zircons with the observed mix of ages has not been recognized in the orogenic belt. Finally, it seems unlikely that terranes in the upper plate of the Brookian orogen could have undergone sufficient burial during the Jurassic and Early Cretaceous to set ZFT ages because of their high structural positions in the orogen.

Synorogenic Deposits of Western Chukotka

Upper Jurassic and Lower Cretaceous synorogenic strata rest unconformably on the fine-grained Triassic turbidite deposits throughout Chukotka. The Jurassic and Cretaceous deposits consist of thin- to thick-bedded turbidites, tuffaceous turbidites, polymictic conglomerates, and olistostromes that were shed northward from the South Anyui collisional zone along the southern margin of Chukotka (Sokolov et al., 2002, 2009; Bondarenko et al., 2003). Their lithologic character, age, and involvement in thrusting suggest that these deposits are correlative with the allochthonous lower Brookian predecessor foreland basin deposits of the western Brooks Range (i.e., the Okpikruak Formation).

Although broadly similar to the major zircon age populations of the lower Brookian deposits, the detrital zircon ages from syntectonic deposits in western Chukotka differ in that they are dominated by Paleoproterozoic zircons and contain a much lower proportion of lower Paleozoic, Neoproterozoic, and Mesoproterozoic zircons. As a result, the western Chukotka deposits contain a distinctly different age distribution (Fig. 12B) that suggests a source terrane enriched in Precambrian crystalline rocks. These data support the interpretation that the Jurassic and Early Cretaceous deformation of western Chukotka may have involved parts of the continental margin of Siberia or blocks that had previously rifted away from the Siberian margin such as the Omolon terrane (Miller et al., 2009).

The large difference between the detrital zircon populations of the Triassic turbidite unit and the covering Upper Jurassic and pre-Aptian syntectonic strata in western Chukotka (Figs. 12B versus 12D) shows that the Triassic deposits in this area probably were not a primary source area for the overlying synorogenic deposits. For similar reasons, the western Chukotka synorogenic deposits seem unlikely to constitute an up-dip part of the foreland basin depositional system in western Alaska. It remains possible, however, that by the Aptian and Albian, some sediment from western Chukotka may have been funneled eastward down the foreland basin into the Colville Basin. If present as a small fraction, the addition of Siberian affinity zircons could have increased the proportion of Paleoproterozoic zircons in Colville Basin sandstones without otherwise altering the detrital zircon age distribution in the sediment.

DISCUSSION

Compositional and detrital zircon data from the lower Brookian deposits of the western Brooks Range document changes in the provenance of the sediment shed from the Brookian orogen. The oldest Brookian deposits are the Upper Jurassic thin-bedded volcaniclastic turbidites of unit Jw, which were deposited on volcanic rocks of either the Copter Peak allochthon or the Misheguk Mountain allochthon, which compose the Angayucham terrane. Unit Jw deposits occupy magmatic arc fields on sandstone compositional diagrams (Figure 7) and contain an almost unimodal population of Late Jurassic zircons that together point to derivation from the unroofing of a Late Jurassic volcanic arc. A small proportion of early Paleozoic, Neoproterozoic, and Paleoproterozoic zircons suggest that the arc may have formed on continental rocks or was close to a source of continental sediment. This suggests that unit Jw originated as a forearc basin succession that was deposited either directly on ophiolite basement (if underlain by the Misheguk Mountain allochthon) or on the hindward part of the subduction complex (if underlain by the Copter Peak allochthon). Although unit Jw strata are included in the Brookian sequence because of their lithic composition and derivation from a southern source area, our new data suggest that the unit is best viewed as a lithologic component of the Angayucham terrane.

Brookian deposits from allochthons of the Arctic Alaska–Chukotka microplate in the western Brooks Range have sandstone compositions that range from magmatic arc to recycled orogen composition (Fig. 7), probably reflecting a decline over time of the amount of detritus derived from the Angayucham and/or Koyukuk terranes relative to that derived from the Triassic turbidite unit that overlies the Arctic Alaska–Chukotka microplate in Chukotka. Relatively small contributions of sediment from continental sources rich in zircon such as the Triassic turbidite unit, however, can have outsized importance in detrital zircon populations relative to those of oceanic and volcanic arc sources. The quartz-rich clasts (e.g., sample 06TM-29B) in conglomerates and olistostromes otherwise dominated by mafic and intermediate clasts suggest that continental rocks might have been present in the basement of the Angayucham and/or Koyukuk terrane. Alternatively, the clasts might have been derived from coarse-grained channel deposits in the Triassic turbidite unit that are no longer present because of subsequent erosion.

Brookian deposits of the low allochthons contain locally numerous olistostromes that contain large blocks and clasts of Permian and Triassic chert from the Etivluk Group and limestone from the Lisburne Group of the Arctic Alaska–Chukotka microplate (Martin, 1970; De Vera, 2005). Significant populations of zircons with ages indicating erosion of Endicott Group strata during the Early Cretaceous are not present in our samples. This suggests that active thrusts during this time were rooted in the shale-rich detachments that overlie the clastic rocks of the Endicott Group (e.g., Wallace et al., 1997) (Fig. 4).

The provenance of the wedgetop deposits in the western Colville Basin (i.e., the Mount Kelly Graywacke) is uncertain. Although their composition is rich in detrital carbonate and subordinate metamorphic grains and white mica that seems to point to a source area in the metamorphic rocks of the southern Brooks Range, the detrital zircon U-Pb ages suggest these deposits were more likely derived principally from ancestral foreland basin deposits and the carbonate rocks of the Lisburne Group in the allochthons of the Arctic Alaska–Chukotka microplate, most likely the Kelly River allochthon.

The lower Brookian sandstones from foredeep deposits of the Colville Basin in the western Brooks Range contain Paleoproterozoic zircons that are generally more abundant than those in the Triassic turbidite unit of central Chukotka (Fig. 12D). This suggests that a secondary source area enriched in Paleoproterozoic zircons may have contributed sediment to the Colville Basin. Longitudinal transport of the late synorogenic sediments from western Chukotka, where sandstone with higher abundances of Paleoproterozoic zircons are present (Miller et al., 2009), could explain this observation. Paleoproterozoic zircon populations, however, are also locally abundant in the Triassic turbidite unit from Wrangel Island (e.g., sample ELM06WR518 of Miller et al., 2010), suggesting that Brookian samples with abundant Paleoproterozoic grains may simply reflect compositional variations within the Triassic turbidite unit source area. We prefer this interpretation because the introduction of significant amounts of sediment derived from Siberian sources probably also would have increased the proportion of K-feldspar in the Colville Basin deposits, an increase not supported by our sandstone compositional data.

Following progradation of the Corwin Delta eastward along the axis of the Colville Basin into the central North Slope, the Umiat Delta became the major input point for sediment into the Colville Basin foredeep. The sediment of the Umiat Delta reflects a major change to a source in the metamorphic hinterland of the Brooks Range in the Arctic Alaska–Chukotka microplate. It seems likely that a similar change of provenance may have occurred in the western Brooks Range as well, but its stratigraphic record may be missing due to subsequent erosion across the western North Slope (Houseknecht et al., 2011). If arrival of sediment from the hinterland of the Arctic Alaska–Chukotka microplate was synchronous throughout the foreland region, it would indicate denudation was caused by regional tectonic changes. Alternatively, the arrival of sediment from the hinterland of the Arctic Alaska–Chukotka microplate may have varied along strike due to varying amounts of erosion along the orogenic belt, or because it was sequestered in intermontane basins until erosion provided sedimentary pathways into the foreland basin in the Albian.

New data presented here support the earlier interpretation that most lower Brookian sediment in the western Brooks Range was sourced from areas underlain by the Triassic turbidite unit in the Herald Arch, eastern Chukotka, and present-day Chukchi Sea (Fig. 1). Despite having comparable zircon populations, the present-day thickness and distribution of the Triassic turbidite unit in Chukotka seem inadequate to explain the voluminous lower Brookian deposits in northern Alaska. This suggests that substantial thicknesses of the Triassic strata must have once been present but subsequently were removed by erosion in the source area. This is supported by the deformation and sub–greenschist-facies metamorphism of the Triassic turbidite unit and the ZFT data that indicate burial to depths of as much as 7–10 km, providing evidence for a highland region that consisted of structurally thickened parts of the same unit. The highlands probably extended southward from the Herald-Wrangel thrust across eastern Chukotka to the South Anyui zone and perhaps southward to the Seward Peninsula, although subsequent extensional deformation during the mid-Cretaceous in the Brooks Range and in the Tertiary in the Hope Basin (Tolson, 1987; Dumitru et al., 1995; Klemperer et al., 2002; Miller and Akinin, 2008) may have substantially broadened the modern north-south extent of the former highland.

Based on relations in the South Anyui zone along the southwestern margin of Chukotka, Sokolov et al. (2009) reported that the Triassic turbidite unit occupies a structural position above the parautochthonous basement and platform sedimentary cover of the Arctic Alaska–Chukotka microplate and below the upper Paleozoic and lower Mesozoic ophiolitic, island-arc, and sedimentary rocks of the Russian Far East. This stacking succession suggests that the Triassic turbidite unit occupies a structural level that is equivalent to that of the allochthonous sedimentary cover sequences of Arctic Alaska–Chukotka microplate in the Brooks Range (i.e., the Endicott Mountains allochthon, Picnic Creek allochthon, Kelly River allochthon, and Ipnavik River allochthon), with the arc and ophiolite thrust sheets of the South Anyui terrane being correlative with Angayucham terrane (i.e., the Copter Peak allochthon and Misheguk Mountain allochthon) in Alaska. The Alazeya-Oloi terrane, for example, is thought to comprise remnants of the late Paleozoic to Early Cretaceous arc and subduction complex much like the Koyukuk arc terrane, Misheguk Mountain allochthon, and Copter Peak allochthon represent in Alaska, with both arc-forearc systems having been thrust onto the Arctic Alaska–Chukotka microplate during the Brookian structural event (Sokolov, 2010).

The lateral change in facies from starved distal shelf deposits in the Brooks Range (Fig. 4) to a deep-water sedimentary prism composed of Triassic thin-bedded prodelta turbidites in the Chukchi Sea and Chukotka may have influenced the formation of the highland area that was the main source of lower Brookian sediments. Because the turbidite succession in the Chukchi Sea and Chukotka was significantly thicker than the distal shelf deposits in Alaska, equal amounts of shortening applied to both areas would have resulted in a significantly thicker structural stack in Chukotka than in Alaska (Fig. 16). Erosion of the resulting culmination would have begun in the structurally higher oceanic rocks followed by increasing amounts of erosion of the underlying turbidite unit as the culmination grew and was dissected.

The deep-marine sedimentary environment of the Triassic turbidite unit, coupled with mafic dikes and sills in the older part of the sequence and the deposition of the turbidite section on the older extensionally deformed platform deposits and Neoproterozoic crystalline basement of the Arctic Alaska–Chukotka microplate, suggest that the Triassic turbidite unit was deposited on tectonically thinned continental crust along the rifted southern margin of the Arctic Alaska–Chukotka microplate. One interpretation is that the turbidite unit formed part of a large deltaic, prodeltaic, and basinal turbidite-fan complex that emanated from a fluvial distribution system shed southward across the Arctic region prior to formation of the Jurassic and Cretaceous Amerasia and Cenozoic Eurasia ocean basins (Miller et al., 2013). This interpretation is also consistent with south-directed paleocurrent flow of the turbidites. The Arctic restorations of Miller et al. (2006, 2010) and Amato et al. (2009) placed the Uralian and Taimyr orogens northeast of the Arctic Alaska–Chukotka microplate (in present-day coordinates) in the Triassic (Fig. 17A).

An alternative interpretation favored here is that the South Anyui–Angayucham oceans included the northern and eastern part of the Uralian Ocean, which separated Baltica and Siberia in the late Paleozoic but was closed by the Uralian-Taimyr orogenic event in the Late Permian and Triassic (e.g., Sokolov et al., 2009; Miller et al., 2011, their fig. 4). We suggest that by the middle Mesozoic, the South Anyui Ocean was a remnant ocean basin bordered by highly extended and transitional crust of the Arctic Alaska–Chukotka microplate on the northeast and an active margin of late Paleozoic and early Mesozoic island arcs along the margin of Siberia on the southwest (Fig. 17B). Reconstructions of the Uralian orogen indicate that this orogen closed from south to north during the Carboniferous to the Triassic, with the youngest part of the closure having taken place in Taimyr in the Triassic (Torsvik and Andersen, 2002). Sediment, perhaps including coarse conglomeratic material, was shed from the orogen westward into the Uralian-Taimyr foreland basin and then down the basin axis into deltaic and prodeltaic turbidite fans constructed on distal turbidite deposits in the nearby northwestern apex of the South Anyui Ocean basin. Thrusting of the Siberian margin onto the Arctic Alaska–Chukotka microplate margin during closure of the South Anyui Ocean in the Jurassic and Early Cretaceous would have resulted in collapse of the remnant basin and accretion of the Triassic turbidite unit onto the Arctic Alaska–Chukotka microplate beneath the subducting arc systems that existed outboard of the margin of Siberia. This interpretation explains the presence of the Triassic turbidite unit along the length of the Arctic Alaska–Chukotka microplate in Chukotka and is consistent with the models of Sokolov et al. (2002) and Nokleberg et al. (2000) for the evolution of the South Anyui and Angayucham ocean basins. It also agrees with the models for opening of the Amerasian Basin of Lawver et al. (2002) and Grantz et al. (2011b), which feature counterclockwise rotation of the Arctic Alaska–Chukotka microplate away from the Canadian margin in the Jurassic and Early Cretaceous and restore the western end of the Arctic Alaska–Chukotka microplate against Eurasia in the vicinity of the Kara Sea prior to opening of the basin.

This study provides the first direct evidence for a connection between the Alaskan and Russian parts of the Arctic Alaska–Chukotka microplate. Previous correlations suggested such a connection based on similarities in lower Paleozoic stratigraphy, and tectonic history. The depositional transfer of detritus from the Triassic turbidites in Chukotka to the Brookian foreland basin system in Arctic Alaska validates the correlation back to the earliest Cretaceous and confirms that the Late Jurassic and Early Cretaceous deformational events of Arctic Alaska and Chukotka are parts of the same orogenic event.

Our study further suggests that foreland basin deposition throughout the early evolution of the Brooks Range was increasingly dominated by longitudinal sediment dispersal systems that emanated from the eastern Chukotka–Herald Arch–Chukchi Sea region (Fig. 18). Although there is clear evidence of contributions of sediment from local, possibly submarine source areas in the Brooks Range, the abundance of Paleozoic and Triassic zircons in lower Brookian sandstones extending at least as far east as the central Brooks Range suggests that longitudinal systems may have deposited sediment along the orogen for 750 km or more. In addition, the abundance of these deposits at distal positions suggests that early deposited foreland basin strata were structurally incorporated into the thrust belt where they became important contributors of recycled sediment for younger foreland basin deposits as the thrust system propagated toward the foreland. As more sediment accumulated and became involved in the thrusting, larger volumes of syntectonic sediment were contributed into the foreland basin by recycling, thus reducing the proportion of the sediment contributed from pre-orogenic rocks. This process drove sandstone compositions to more sedimentary lithic compositions and to detrital zircon age distributions like those of the primary source area in Chukotka as the foreland basin evolved. The dominance of longitudinal sediment dispersal systems and local syntectonic recycling did not change until a major transverse sediment dispersal system, the Umiat Delta, developed from a source area in the metamorphic hinterland of the orogen in the Albian. This system shifted the sandstone compositions to metamorphic lithic compositions and to detrital zircon age distributions of the Paleozoic and older rocks of the Arctic Alaska–Chukotka microplate. The change to a transverse sediment dispersal system may mark a major reorganization in the orogenic belt caused by the shift from collisional to extensional tectonics in the Brooks Range in the Albian.

CONCLUSIONS

(1) The latest Jurassic to late Early Cretaceous western Brooks Range is a thin-skinned fold-and-thrust belt with a well-developed foreland basin and remnants of ancestral foreland basin deposits involved as syntectonic deposits at various structural levels.

(2) Sandstones from the highest structural level (unit Jw) consist almost entirely of volcanic arc detritus with a near unimodal population of detrital zircons having ages of 150–170 Ma (Late and Middle Jurassic). These deposits are interpreted as forearc basin deposits in the Angayucham terrane, which in the western Brooks Range were deposited on ophiolite and/or subduction complex rocks that form the upper plate of the Brooks Range orogen.

(3) Detritus and zircons shed from local sources in the Angayucham terrane are abundant in sandstones, conglomerates, and olistostromes in synorogenic sandstones in the immediately underlying thrust sheets, (i.e., the Ipnavik River, Kelly River, and Picnic Creek allochthons) but are less common in the later-emplaced and now lower thrust sheets of the Endicott Mountains allochthon and in the Colville Basin.

(4) Quartz, chert, clastic sedimentary, carbonate, and to a lesser extent metamorphic lithic grains are common in synorogenic sandstones of the high allochthons and dominate the lower Brookian sandstones of the Endicott Mountains allochthon and Colville Basin. These sandstones yield detrital zircon ages that feature large populations at 359–200 Ma (Triassic and late Paleozoic), subordinate populations at 542–359 (early Paleozoic), and lesser populations at 2.0–1.75 Ga (Paleoproterozoic).

(5) Although source areas for lower Brookian deposits have often been assumed to lie in the Paleozoic and older basement rocks of the metamorphic hinterland of the Brookian orogen, the Arctic Alaska-Chukotka microplate in Alaska contains virtually no rocks that could provide the Triassic and late Paleozoic zircons found in the lower Brookian deposits of the western Brooks Range. In addition, the Paleozoic and older rocks of the orogen contain distinctly different proportions of early Paleozoic and older zircons than those found in the lower Brookian strata. The first appearance of detritus having zircon age populations characteristic of the Arctic Alaska–Chukotka microplate in the Brookian sequence occurs in the mid-Albian metamorphic lithic sandstones in the Umiat Delta in the central Brooks Range.

(6) The dominant Paleozoic to Triassic zircon populations in the Brookian sandstones are very similar to those in the Triassic turbidite unit in Chukotka and Wrangel Island and, along with facies and paleoflow data from the Colville Basin, support earlier interpretations of a significant highland area located in eastern Chukotka, the Herald Arch, and the present-day southern Chukchi Sea.

(7) Zircon fission-track double dating supports the interpretation that the Triassic turbidite unit in Chukotka was derived from the exhumation and erosion of upper Paleozoic and Triassic rocks in the Uralian-Taimyr orogenic belt and provides evidence that these Triassic strata were later buried to depths of as much as 7–10 km and exhumed as part of the Brookian orogenic belt in the Jurassic and Early Cretaceous.

(8) The Triassic turbidite unit is proposed to have been deposited in a marginal basin setting in the South Anyui Ocean and thrust onto the Chukotka part of the Arctic Alaska–Chukotka microplate during the Late Jurassic and Early Cretaceous Brookian orogenic event. The turbidite unit was structurally thickened by the deformation, producing a regional culmination in the area of eastern Chukotka, Herald Arch, and southern Chukchi Sea that became the source area for much of the detritus in the foreland basin deposits of the western Brooks Range.

(9) The presence of Triassic and Paleozoic zircon populations in lower Brookian deposits at most structural levels and throughout the development of the collisional phase of the Brookian orogen requires that sediment dispersal occurred mainly by axial (longitudinal) flow over distances in excess of 750 km along the Brookian foreland basin. Recycling of sediment due to structural imbrication of ancestral foreland basin deposits with underlying deformed platform deposits at the thrust front produced olistostromes and local sedimentary systems that also contributed sediment to the foreland basin system.

(10) The transfer of clastic detritus from source areas in the Triassic turbidite unit of eastern Chukotka, the Herald Arch, and the southern Chukchi Sea into the Brookian foreland basin system in northern Alaska confirms previous correlations of the Brookian orogen from northern Alaska across the Chukchi Sea and into the Russian Far East.

This study benefited from discussions with Ken Bird, Dwight Bradley, Julie Dumoulin, Elizabeth Miller, David Houseknecht, Chad Hults, Karen Kelley, Richard Lease, Gil Mull, Alison Till, Marwan Wartes, and Joe Wooden. We thank Margaret Donelick and Jim McMillan for sample preparation and Charles Knaack for technical assistance with LA-ICP-MS data collection procedures. The manuscript was sharpened and significantly improved by the thoughtful comments and advice of U.S. Geological Survey reviewers Kenneth J. Bird and Richard Lease and Geosphere reviewers Jaime Toro and Tim Lawton. Art Grantz, Paul Stone, Sarah Nagorsen, and Mike Diggles kindly read and edited the manuscript. The work was funded by the Energy Resources and Mineral Resources programs of the U.S. Geological Survey. We thank Teck, the owner and operator of the Red Dog Mine, for graciously allowing us to use their facilities for our research. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

1Supplemental File. The Supplemental File contains sample locality information and descriptions, discussions of the U-Pb and zircon fission-track analytical techniques employed in the study, tables reporting the raw sandstone point-count data, U-Pb zircon age data and zircon fission-track age data, and figures showing relative probability plots for U-Pb detrital zircon samples, graphical evidence for Pb loss in sample 06TM-08, and modeling of the zircon fission-track data. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES01043.S1 or the full-text article on www.gsapubs.org to view the Supplemental File.