Tectonic and geodynamic models of the formation of the Amerasian Basin are discussed. The Arctic margins of the Chukchi region and Northern Alaska have much in common in their Late Jurassic–Early Cretaceous tectonic evolution: (1) Both have a Neoproterozoic basement and a complexly deformed sedimentary cover, with the stage of Elsmere deformations recorded in their tectonic history; (2) the South Anyui and Angayucham ocean basins have a common geologic history from the beginning of formation in the late Paleozoic to the closure at the end of the Early Cretaceous, which allows us to consider them branches of the single Proto-Arctic Ocean, the northern margin of which was passive and the southern margin was active; (3) the dipping of the oceanic and, then, continental lithosphere took place in subduction zones southerly; (4) the collision of the passive and active margins of both basins occurred at the end of the Early Cretaceous and ended in Hauterivian–Barremian time; (5) the collision resulted in thrust–fold structures of northern vergence in the Chukchi fold belt and in the orogen of the Brooks Ridge. A subduction-convective geodynamic model of the formation of the Amerasian Basin is proposed, which is based on seismic-tomography data on the existence of a circulation of matter in the upper mantle beneath the Arctic and East Asia in a horizontally elongated convective cell with a length of several thousand kilometers. This circulation involves the subducted Pacific lithosphere, the material of which moves along the bottom of the upper mantle from the subduction zone toward the continent, forming the lower branch of the cell, and the closing upper branch of the cell forms a reverse flow of matter beneath the lithosphere toward the subduction zone, which is the driving force determining the surface kinematics of crustal blocks and the deformation of the lithosphere. The viscous dragging of the Amerasian lithosphere by the horizontal flow of the upper mantle matter toward the Pacific leads to the separation of the system of blocks of Alaska and the Chukchi region from the Canadian Arctic margin. The resulting scattered deformations can cause a different-scale thinning of the continental crust with the formation of a region of Central Arctic elevation and troughs or with a breakup of the continental crust with subsequent rifting and spreading in the Canadian Basin.

The Arctic region includes the Arctic Ocean and its continental margins. It is usually divided into the West and East Arctic regions based on structural features and differences in geologic history. The boundary between them is typically traced along the Lomonosov Ridge that separates two large sedimentary basins: the Eurasian Basin to the west and the Amerasian Basin to the east. The oceanic crust of the Eurasian basin formed during seafloor spreading that began 56 Ma (Karasik, 1968; Vogt et al., 1979; Taylor et al., 1981; Brozena et al., 2003; Glebovsky et al., 2006; Gaina et al., 2011). The Gakkel mid-oceanic ridge separates the abyssal Nansen and Amundsen basins and is the divergent boundary between the North American and Eurasian tectonic plates. We can consider the spreading model of formation of the Eurasian basin as universally accepted. Its plate tectonic models differ in some details: interpretation of the oldest magnetic anomalies, stages of spreading, definition of the rifting stage, specifics of the relationship between oceanic structures and the continental margin of the Laptev Sea, etc. Significant variations also arise when interpreting seismic-stratigraphic lines across the Eurasian and Amerasian basins – they increase with the distance from the boreholes drilled on the arctic margin of North Alaska.

However, there are very different opinions on the origin of the Amerasian Basin. A full detailed review of the ideas on its formation was reported in (Lawver and Scotese, 1990). Since then, many models have been published, but they did not introduce essentially new approaches or solutions. One exception would be the concept of a back-arc origin of the Canada Basin (Miller et al., 2006, 2018; Pease and Coakley, 2018; Nikishin et al., 2021).

All the various opinions on the formation of the Amerasian Basin can be combined into five independent groups: oceanization of continental crust; captured Pacific oceanic crust; multistage spreading; back-arc basin; rotation hypothesis. The debate concerning the origin of the Amerasian Basin is the focus of this article: here we discuss the different hypotheses and tectonic models of formation and development of the main lithospheric units of the East Arctic in the Late Jurassic–latest Early Cretaceous interval. Specifically, we consider the occurrence of the Canada Basin, the closing of the Angayucham and South Anyui paleo-oceans, and the formation of collisional structures due to the drift of the Alaska and Chukotka continental blocks. In other words, the aim of this article is to develop an uncontroversial tectonic and geodynamic model for the formation of the main types of structures of the East Arctic.

To the east of Lomonosov Ridge is the area of Central Arctic uplifts and basins, which includes the Alpha-Mendeleev Rise, the Chukchi plateau, the Northwind Ridge, the Podvodnikov and Makarov basins (Fig. 1). The Canada Basin is located to the east and includes an area of oceanic crust. Opinions on the types of crust in the Central Arctic uplifts and basins region changed significantly with time. The continental nature of Lomonosov Ridge and the Chukchi plateau is not in doubt (Weber and Sweeney, 1985; Jokat et al., 1992; Grantz et al., 1998; Petrov et al., 2016; Kaminsky, 2017; Poselov et al., 2019). The evolving points of view on the origin of the Alpha and Mendeleev rises changed from a seafloor spreading (Hall, 1973) and oceanic plateau model (Forsyth et al., 1986; Jackson et al., 1986; Lawver et al., 2002; Jokat, 2003; Grantz et al., 2011) to thinned continental crust (Taylor et al., 1981; Zonenshain et al., 1990; Lebedeva-Ivanova et al., 2006; Golonka, 2011; Laverov et al., 2013; Kaminsky, 2017; Poselov et al., 2019). Underwater sampling of Mendeleev Rise in the “Arktika 2012” scientific voyage (Morozov et al., 2013; Vernikovsky et al., 2014) and from a submersible (Skolotnev et al., 2017, 2019) confirmed its continental nature.

The units of the East Arctic considered in this article are the Amerasian Basin, the huge shelf with the New Siberian Islands archipelago and Wrangel Island, as well as the continental margin of the East Siberian, Chukchi, Laptev, and Beaufort seas. Most of the shelf and continental margin is composed of the Chukotka (New Siberian–Chukotka) folded area and the structures of North Alaska.

The hypothesis of oceanization of continental crust (Shatsky, 1963; Pusharovsky, 1976), the model of captured ancient early Mesozoic (Churkin and Trexler, 1980; Karasik, 1980) or Paleozoic oceanic crust (Rowley and Lottes, 1988), as well as the model of multistage spreading (Gurevich et al., 2003, 2005) do not have supporters at this time.

The models in which the Canada Basin is regarded as a back-arc basin, despite being defended in recent publications (Miller et al., 2006, 2018; Gottlieb et al., 2018; Nikishin et al., 2021), does not stand up to scrutiny for the following reasons. Firstly, the distance between the Koyukuk island arc subduction zone (North Alaska) and the Canada Basin in the present-day setting is approximately 1000 km not counting the shortening caused by fold-and-thrust deformation on the Brooks Range. There are no known actualistic models for this. The Sea of Japan and South China Sea back-arc basins are at a distance of 500–600 km from the subduction zone. Secondly, Upper Devonian–Jurassic sediments of the Arctic Alaska terrane accumulated on the passive margin of the Angayucham Ocean, which excludes a northern dip of the subduction zone towards the Canada Basin. Thirdly, according to American researchers, the subduction zone of the Koyukuk island arc dipped southward (Plafker and Berg, 1994; Harris, 1995; Moore et al., 1997; Nokleberg et al., 2000; Lemonnier, 2015).

The rotation hypothesis is the most popular and supported by most researchers studying the Arctic. It was proposed in (Carey, 1955) and became the basis for various models (Embry, 1990; Lawver et al., 2002; Grantz et al., 2011). According to the rotation hypothesis, the large Arctic Alaska–Chukotka continental block broke off from Arctic Canada and accreted to North America and Eurasia while rotating counterclockwise. The rotation pole was in the Mackenzie River delta. This caused the closing of the South Anyui and Angayucham oceans and the formation of the Canada Basin with oceanic crust. Subsequent publications (Grantz et al., 2011) proposed a two-stage rotation model – formation of transitional crust ca. 195–131 Ma followed by oceanic crust spreading ca. 131–127.5 Ma.

The rotation hypothesis is also supported by comparative analysis of Paleozoic and Mesozoic formations of Arctic Canada and North Alaska. Paleomagnetic data (Halgedahl and Jarrard, 1987) were obtained from core samples of the Berriasian–Valanginian (145.0–134.7 Ma) Kuparuk Formation from two boreholes on the northern slope of Alaska. Calculations show that the Arctic Alaska plate rotated approximately 70 degrees counterclockwise, which also supports the rotation hypothesis.

Moreover, the time of opening of the Canada Basin is limited by the period from 130 to 100 Ma (Halgedahl and Jarrard, 1987). Indeed, Embry and Dixon (1990) supposed a Hauterivian (134.7–130.8 Ma) age for this process, and linked the cessation of seafloor spreading to a regional Cenomanian unconformity (100.5–93.9 Ma). The discovery of traces of Ellesmere deformation in the Chukotka folded area (Natal’in et al., 1999; Lane et al., 2015; Luchitskaya et al., 2015; Verzhbitsky et al., 2015; Prokopiev et al., 2018) pointed to the possibility of the Chukotka structures breaking from Arctic Canada, and thus removed some objections against the rotation hypothesis.

At the same time, the rotation models were confronted with a spatial problem (Lane, 1997; Miller et al., 2006, 2018; Kuzmichev, 2009), therefore, they chose different interpretations for the western margin of the drifting and rotating Arctic Alaska–Chukotka microplate. Some researchers argued that the drift took place along a transform fault along the Amerasian border of Lomonosov Ridge (Embry, 1990; Embry and Dixon, 1990; Grantz et al., 1998, 2011; Lawver et al., 2002), others considered sliding along the eastern margin of the Alpha-Mendeleev Rise (Laverov et al., 2013) or along the Northwind Ridge margin (Nikishin et al., 2021).

Views on the oceanic nature of the uplift and on the oceanic crust of Makarov Basin as a continuation of seafloor spreading in the Canada Basin corresponded to the rotation hypothesis. However, validation of the continental nature of the Alpha-Mendeleev Rise crust and the formation of the Makarov basin in the Cretaceous–Cenozoic (Miller and Verzhbitsky, 2009) limited the area of oceanic crust to the southern part of the Canada Basin. Consequently, models with a rotating Arctic Alaska–Chukotka microplate became untenable.

The tectonic history of the Amerasian Basin cannot be reconstructed without knowing and accounting for the regional geology of adjacent continental margins of Northeast Asia and North America.

Geographically, the arctic margin of Chukotka occupies the continental margin and shelf of the East Siberian and Chukchi seas from the New Siberian Islands archipelago to Wrangel Island and Herald Island. Tectonically, this region is composed of Mesozoic structures of the Chukotka folded area including the New Siberia–Wrangel (East Arctic), the Anyui–Chukotka and the South Anyui fold systems (Fig. 2).

The northern boundary of the Chukotka Mesozoides is along the front of the Wrangel–Herald thrust (Fig. 2). Northerly of the thrust, the structures of the Arctic Alaska–Chukotka microplate did not undergo Mesozoic deformation (Pease, 2011; Pease et al., 2014) that took place at the end of the Early Cretaceous epoch during the Chukotka orogeny. Along its southern boundary is the South Anyui fold system that is considered a suture zone (Parfenov and Natal’in, 1977; Natal’in, 1984; Parfenov et al., 1993; Sokolov et al., 2002).

The New SiberianWrangel and AnyuiChukotka fold systems have a relatively simple two-part structure: a Neoproterozoic (possibly Meso-Neoproterozoic) metamorphic basement (Natal’in et al., 1999; Kos’ko, 2007; Korago et al., 2014; Sokolov et al., 2020) and an intricately deformed Paleozoic–Mesozoic cover (Kos’ko et al., 1993; Natal’in et al., 1999; Geodynamics…, 2006; Kos’ko, 2007; Vernikovsky et al., 2013b; Danukalova et al., 2015). The metamorphic basement is exposed on Wrangel Island (Kameneva and Il’chenko 1978; Kos’ko et al., 1993; Sokolov et al., 2015) and on Chukotka in the Bering terrane (Zhulanova, 1990; Natal’in et al., 1999; Geodynamics..., 2006) and the Velitkina uplift (Gottlieb et al., 2017). There are unconformities in the sedimentary cover at the base of the Upper Devonian system on Kotelny Island (Kos’ko, 2007; Prokopiev et al., 2018), and the bases of the lower and middle Carboniferous system on Wrangel Island (Kos’ko et al., 1993; Sokolov et al., 2015). The unconformities resulted from various events during the Ellesmere orogeny that manifested on Chukotka as ages of the granites of the Kiberov and Kuekvun’ plutons: 352–359, 362 and 364 Ma (Lane et al., 2015; Luchitskaya et al., 2015), and of the orthogneiss of the Koolen’ metamorphic dome in eastern Chukotka – 369 and 375 Ma, where Devonian andesites and tuffs also occur (Natal’in et al., 1999). The Late Devonian–early Carboniferous magmatic bodies formed in a suprasubduction, continental margin geodynamic setting (Natal’in et al., 1999; Luchitskaya et al., 2015).

From a paleotectonic standpoint, the New Siberian–Wrangel and Anyui–Chukotka fold systems formed the Chukotka microcontinent, which was included in the Arctic Alaska–Chukotka microplate (Churkin and Trexler, 1981; Till, 2016; Miller et al., 2018), or was part of Arctida (Zonenshain and Natapov, 1987; Vernikovsky et al., 2013a; Metelkin et al., 2015), or the Bennett-Borovia continental block (Natal’in et al., 1999). Recent geochronological data for the New Siberian Islands (Akinin et al., 2015; Ershova et al., 2016; Metelkin et al., 2016) and results of underwater sampling (Morozov et al., 2013; Skolotnev et al., 2019) show that the Mendeleev Rise can be included into the Arctic Alaska–Chukotka microplate (Sokolov et al., 2020).

The South Anyui fold system (SAFS) is regarded from a structural point of view as the suture that connects the Verkhoyansk–Kolyma and Chukotka folded areas (Fig. 2). It formed in place of the paleo-ocean that closed due to the Early Cretaceous collision. Along its southern boundary, the structures of the Alazeya–Oloi fold system are located – these were formed mainly by island-arc terranes of a wide age range from the Devonian to the end of the Early Cretaceous (Parfenov et al., 1993; Parfenov and Kuzmin, 2001; Geodynamics…, 2006). In the west, the SAFS formations crop out on Great Lyakhovsky Island and terminate on the continent as the “Khrom” loop, which is visible in magnetic anomalies under the Cenozoic cover (Khain et al., 2009; Kuzmichev, 2009). It continues in the east as the formations of the Velmay terrane of East Chukotka (Sokolov et al., 2015; Ledneva et al., 2016) and the Kobuk suture in Alaska (Fig. 2) (Plafker and Berg, 1994).

The SAFS consists of oceanic and island-arc formations, turbidites and ophiolites formed in the Paleozoic–Mesozoic ocean. Two stages are identified in the tectonic development of the ocean: late Paleozoic–early Mesozoic and Late Jurassic–Early Cretaceous (Sokolov et al., 2015; Ganelin, 2017).

During the late Paleozoic–early Mesozoic interval, the Proto-Arctic Ocean was located between the Siberia and Laurentia continental units. This was an extensive paleoocean consisting of two basins: South Anyui and Angayucham. In the late Paleozoic time prior to the collision of the Kara microcontinent and Siberia, it connected the Paleo-Pacific to the Ural paleo-ocean (Zonenshain and Natapov, 1987; Vernikovsky, 1996; Vernikovsky et al., 2013a; Sokolov et al., 2015). After the collision, the Proto-Arctic Ocean transformed into a bay of the Paleo-Pacific with a system of island arcs and marginal seas along the southern Siberian coast. The history of this paleo-ocean and its convergent margin was most comprehensively described in (Sokolov et al., 2015, 2021; Ganelin, 2017). The termination of seafloor spreading in the South Anyui branch of the Proto-Arctic Ocean is dated as the Oxfordian–Kimmeridgian based on the age of the youngest oceanic basalt-silica association of the intra-oceanic Kulpolney island arc.

The Volgian (Tithonian–Berriasian) was a new stage in the history of the SAFS that ended in the collision of the Chukotka microcontinent with the active margin of Siberia at the end of the Early Cretaceous (Hauterivian–Barremian). This was preceded by a period of tectonic transformation that encompassed both the paleo-ocean and its continental margins. The Proto-Arctic Ocean transformed into the residual South Anyui Basin with relict oceanic crust that started filling with turbidites.

The southern Siberian margin of the paleo-ocean was active. During the Middle Jurassic, the Kolyma–Omolon superterrane formed by amalgamation of the Khetachan, Alazeya, Oloi, and Yarakvaam island-arc terranes and the Prikolymsky, Omulevka, Omolon, and Avekova cratonal terranes (Parfenov et al., 1993; Parfenov and Kuzmin, 2001). At the margin of this superterrane with the South Anyui oceanic basin, the Oloi and Anyui-Svyatonosky volcanic belt formed (Zonenshain et al., 1990; Goryachev, 2006). The late Paleozoic–early Mesozoic convergent boundary of west-Pacific type with island arcs and back arc basins transitioned in the Late Jurassic–Early Cretaceous period into an active margin with mature continental crust and a lack of back-arc basins with oceanic crust. It differed from Andeantype margins by the accumulation of island-arc formations mainly in a marine environment.

Along this volcanic belt, an accretionary wedge formed from the side of South Anyui Basin with oceanic crust; its fragments have been identified in the SAFS (Sokolov et al., 2002, 2015, 2021). The wedge is composed of turbidites, tuff-terrigenous rocks, submarine landslide strata and a terrigenous mélange that includes oceanic crust fragments: basalt, silica, plagiogranite, and gabbro. There are also tectonic sheets of basalt and silica up to 800 m thick. Terrigenous rocks contain Volgian (Tithonian–Berriasian) and Berriasian–Valanginian fauna; the silica rocks contain Bajocian–Callovian and Oxfordian–Kimmeridgian radiolarians.

In the northern (Chukotka) margin of the Proto-Arctic Ocean, during the late Paleozoic era, there was a carbonate platform; and during the Permian–Triassic period, the passive continental margin of the Chukotka microcontinent formed by mainly terrigenous sedimentation. There are various opinions concerning the nature of the northern boundary in the Late Jurassic–Early Cretaceous time. One view (Natal’in, 1984; Zonenshain et al., 1990; Nokleberg et al., 2000; Amato et al., 2015) presents the Nutesyn island arc (Late Jurassic–Early Cretaceous) formed on the margin of the Chukotka microcontinent. This can be called the doublesided model, because the South Anyui oceanic basin had an active margin both in the south (Oloi volcanic belt) and in the north (Nutesyn arc) during this period. Therefore, the oceanic lithosphere subducted both in the southern and northern direction.

Later, an alternative, one-sided model was proposed with an active margin in the south and a passive one in the north (Shekhovtsov and Glotov, 2001; Sokolov et al., 2002, 2015, 2021). This was based on new geologic mapping data (Shekhovtsov and Glotov, 2001) which found that the volcanogenic Koranveem unit that was previously attributed to the Nutesyn island arc has an older Oxfordian–Kimmeridgian age and does not have any stratigraphic contacts with the Triassic and older deposits of the Chukotka microcontinent. Moreover, the volcanogenic-sedimentary rocks of the Nutesyn and other Lower Cretaceous formations were mistakenly attributed to the Nutesyn volcanic arc. It turned out that they overlap the base conglomerate with an unconformity and contain Aptian(?)–Albian vegetal remains and flora. These continental deposits fill the intermontane Nutesyn depression. The tectonic position and relationships of these formations are shown in Fig. 3 (Moiseev et al., 2021).

In the basin of Koranveem River, the lower contact of the Koranveem unit with terrigenous rocks of the SAFS is stratigraphic with a gradual transition from sedimentary rocks to volcanogenic ones (Shekhovtsov and Glotov, 2001; Sokolov et al., 2002, 2015, 2021). The transitional layers contain Oxfordian–Kimmeridgian fauna. The volcanic rocks have calc-alkaline and subalkaline affinities and could have formed in an island-arc setting.

In the west, near the Polyarninsk uplift, the rocks of the Oxfordian–Kimmeridgian volcanogenic complex occur structurally above the lower Carboniferous volcanogenic-carbonate deposits, Triassic turbidites and Volgian (Tithonian–Berriasian) sandstones of the Anyui–Chukotka fold system. The tectonic nature of the contact manifests as zones of silicified mylonites, intensely deformed terrigenous and greenschist rocks, and serpentinites. The Ar–Ar age of greenschist metamorphism is 158.1 ± 4.0 Ma and was probably due to subduction in the Kulpolney island arc (Sokolov et al., 2015).

Based on results of these investigations, the ensimatic Koranveem (Shekhovtsov and Glotov, 2001) or Kulpolney (Sokolov et al., 2002, 2015, 2021) island arc was defined in the Proto-Arctic Ocean. Subsequent structural and geochemical investigations confirmed the intraoceanic position (Fig. 3) and the ensimatic nature of the Kulpolney arc (Moiseev et al., 2021).

A detailed study of the lithology of Tithonian–Valanginian terrigenous deposits of the Chukotka terrane found the presence of tuff material and zircons with ages 150–140 Ma (Vatrushkina, 2020). Moreover, in several places in eastern Chukotka, there were finds of individual bodies of ignimbrite and rhyolite tuff with ages 146–140 Ma (Tikhomirov, 2018). The discovery of products of coeval magmatism led to a solution that appeared to give a common ground to the supporters of the Nutesyn and Kulpolney arcs (Vatrushkina, 2020). This model implies the existence of two island arcs of different ages. The intraoceanic Kulpolney island arc existed during the Oxfordian–Kimmeridgian ages. After its accretion to the Chukotka microcontinent, the subduction zone inverted and the new Nutesyn epicontinental arc formed on its margin. At the same time, such a model also implies a widespread development of volcanic formations that are very typical for ensialic arcs or continental margin belts. However, the singular exposures of acidic volcanic rocks, detrital zircon grains and the presence of tuff particles are not enough to accept such a model. One should also consider that the quartz diorite and plagiogranite porphyry dikes intruding the Kulpolney complex volcanic rocks age dated at 143 ± 1 and 140 ± 1 Ma (zircon, U–Pb, SHRIMP II) (Vatrushkina, 2020). The closeness of their age to that of acidic volcanic rocks of the Anyui–Chukotka fold system means they can be considered as suture formations according to terrane analysis.

Consequently, the existence of the Nutesyn ensialic arc needs further validation; therefore, we favor the one-sided model. Moreover, regardless of the chosen viewpoint on one-sided or double-sided subduction, the main destruction of oceanic crust in the South Anyui Basin both in the late Paleozoic–early Mesozoic and in the Late Jurassic–Early Cretaceous time took place along the convergent boundary between Siberia and the Proto-Arctic Ocean, along with the formation of the accretionary wedge with oceanic crust fragments (Sokolov et al., 2002, 2015).

Figure 4 shows a tectonic model for the Late Jurassic–Early Cretaceous interval that we chose for further discussion of the deep-seated mechanisms that determined the formation of the lateral series of paleostructures: the active margin of the Omolon microcontinent, the South Anyui oceanic basin, the Chukotka microcontinent, the Canada Basin with oceanic crust, the North America continent. The Omolon active margin is represented by the Oloi volcanic belt with a coeval accretionary wedge. In the South Anyui oceanic basin, the Kulpolney ensimatic island arc collided with the passive margin of the Chukotka microcontinent in the beginning of the Early Cretaceous epoch (143–140 Ma). The decline of the intraoceanic subduction took place after the complete destruction of the oceanic lithosphere of the South Anyui Basin in the front of the Oloi volcanic belt, which terminated in the collision of the Chukotka microcontinent with the Siberian active continental margin.

From north to south, the main tectonic units of North Alaska are the following: the Colville Basin, the Arctic Alaska superterrane, the Kobuk suture, the Angayucham and Koyukuk terranes, the Koyukuk basin (Moore et al., 1994, 1997; Plafker and Berg, 1994; Nokleberg et al., 2000; Dumoulin et al., 2004; Till, 2016).

The Colville Basin (Fig. 2) is regarded as a depression in the front of the Brooks Range orogen. It is filled with marine, shelf, and coastal marine and continental sedimentary rocks (up to 10 km thick) overlying the autochthon. These deposits are Aptian–Late Cretaceous in age, and they are younger than the synorogenic (syncollisional) deposits of the Proto-Colville basin, which is composed of allochthonous sheets of Upper Jurassic–Lower Cretaceous flysch and olistostromes of the Okpikruak Formation that was preserved in the De Long Mountains and Endicott Mountains terranes.

The Arctic Alaska superterrane (Fig. 2). The Brooks Range orogen consists of an assemblage of tectonic sheets and thrusts of northern vergence that formed in the Jurassic–Cretaceous period (Brooks orogeny). These sheets are combined in terranes (subterranes) within the Arctic Alaska superterrane (Fig. 5). The lowest in the structure is the North Slope terrane, which is a paraautochthon for the overlying terranes and serves as the basement of the Colville Basin.

Neoproterozoic metamorphic complexes function as the basement for the lower Paleozoic carbonate platform. The amphibolite facies metamorphism has an age close to 680 Ma (Dusel-Bacon et al., 1989; O’Brien and Grove, 2020) and the orthogneiss of Seward Island are 680–670 Ma in age (Amato et al., 2009).

Lower Paleozoic (Cambrian–Lower Devonian) deposits represent a carbonate platform deformed during the Ellesmere orogeny. Ellesmere fold structures can be traced from northern Greenland and Arctic Canada to Wrangel Island, Chukotka and North Alaska. The Cambrian and Ordovician fauna of the North Slope terrane has a Laurentian origin and that of the Hammond and Coldfoot terranes – a Siberian one (Dumoulin et al., 2002, 2018). During the Early Devonian epoch, large volumes of acidic magma (395, 390, 380–375 Ma) of suprasubduction origin intruded (McClelland et al., 2006; Amato et al., 2009), followed by rifting during the Middle and Late Devonian epochs (Moore et al., 1994, 1997; Lane et al., 2015).

To the north of the superterrane, the place of the presentday Canada Basin was occupied by a continental upland – Borovia or the Arctic platform, which served as the material provenance for the Ellesmere complex from the late Paleozoic to the late Mesozoic. Upper Devonian–Jurassic deposits of the Arctic Alaska superterrane accumulated in a setting of subsiding passive continental margin with more deep-water facies in the south towards the Angayucham oceanic basin. Upper Devonian continental fluvial-deltaic deposits changed in the Carboniferous period to carbonateplatform deposits and during the Permian–Jurassic to clastic rocks, silica shales, pelagic limestones. The Upper Devonian clastic rocks and younger formation contain a large population of detrital zircons aged 440–420 Ma typical for northwestern Laurentia (Beranek et al., 2010). The Late Devonian and Carboniferous biota are most similar to Canadian populations (Dumoulin et al., 2002, 2018). This data indicates that the Arctic Alaska superterrane belonged to Laurentia.

Along the northern margin of Arctic Alaska, offshore, the Jurassic and Lower Cretaceous (Berriasian–Valanginian) deposits display a complex structure on seismic lines corresponding to a shallow marine zone and slope detritus deposits progressing southward. These fine-grained progradational deposits indicate the formation of an uplift composed of the Ellesmere complex. Close to the growing Brooks Range orogen, the condensed sediments of the Endicott Mountains and North Slope were accumulating. Later, the Late Jurassic–Early Cretaceous period was a time of flysch sedimentation.

In the middle Early Cretaceous epoch, sedimentation in the north took place in a rifting setting, which ended in destruction of the continental crust and formation of the oceanic crust of the Canada Basin. Thus, a rifting mode dominated in the north, while the Brooks orogeny and formation of allochthons prevailed in the south.

The Angayucham terrane (Fig. 5) is composed of pillow-basalt, diabase, tuff, silica, limestone and sepentinite bodies (Pallister et al., 1989; Patton et al., 1994; Harris, 1995). The tectonic thickness of the terrane is about 10 km. The volcanic rocks are similar in composition to oceanic plateaus. Limestones and radiolarites contain Devonian–Jurassic microfauna.

The northward movement of allochthons during the early Brooks orogeny in the Late Jurassic and Early Cretaceous periods was related to convergence between the Angayucham oceanic terrane and the Arctic Alaska terrane continental margin. The beginning of the Brooks orogeny deformation is dated by the age of metamorphism of amphibolites (169–163 Ma) in the base of the ophiolite section of the Misheguk Mountain allochthon (Angayucham terrane), which formed due to the obduction of the young and hot oceanic crust. Peridotite and gabbro of the Misheguk allochthon form the uppermost allochthon in the structure and are thrust over pillow-basalt and diabase of the Copter Peak allochthon (Angayucham terrane) (Moore et al., 1994; Patton et al., 1994). The ophiolite allochthon overlaps the Arctic Alaska superterrane (Fig. 5).

The oceanic rocks of the Angayucham terrane, like the Arctic Alaska superterrane passive continental margin (Arctic platform), have a Devonian–Jurassic age. Such relationships imply that the opening of the Angayucham oceanic basin took place during the middle (Grantz et al., 1991) or late (Moore et al., 1994) Paleozoic era due to rifting.

The Koyukuk terrane. The Koyukuk island arc existed on the southern margin of the Angayucham basin in the Middle Jurassic–Early Cretaceous period (Box, 1985; Box and Patton, 1989; Plafker and Berg, 1994). Volcanic and plutonic rocks of the Koyukuk terrane are exposed to the south of Brooks Range and to the east of the Seward Peninsula (Fig. 2). The oldest complexes are of Middle Jurassic age and unconformably overlapped by volcanic and volcaniclastic rocks of the main phase of island-arc activity during the 145–130 Ma interval. The Angayucham terrane and the overlapping Koyukuk terrane are covered by Albian and younger deposits of the Koyukuk basin.

The terranes of the upper part of the Arctic Alaska structure (De Long Mountains, Endicott Mountains) are deformed on shallow structural levels, whereas the terranes closer to the base (Hammond and Coldfoot) have undergone ductile deformation at greater depths (Oldow et al., 1987; Till et al., 1988). Mineral associations in the terranes with ductile deformation in the lower part of Brooks Range indicate high-pressure and low-temperature metamorphism (~515 °C, 10–13 kbar). Metamorphism of the Angayucham terrane and the structurally higher terranes of Arctic Alaska took place in lower temperatures and pressures (Dusel-Bacon et al., 1989; O’Brien and Grove, 2020).

The main stage of deformation and growth of the foldand-thrust structure of the superterrane during the Late Jurassic–Early Cretaceous (Valanginian) is considered as the cause for the accumulation of the Okpikruak Formation that contains fragments and olistostromes of ophiolite allochthons (Moore et al., 1994, 1997). Deformations of the late Brooks orogeny were recorded in the exhumation and cooling ages of the terranes at ca. 120–90 Ma (white mica, K–Ar and Ar–Ar methods (Plafker and Berg, 1994; Moore et al., 1997; Dusel-Bacon et al., 1998)). Erosion of the growing uplift caused the accumulation of huge masses of sediments in the Aptian(?), Albian–Cenomanian interval in the Colville and Koyukuk basins. The total shortening of the Brooks Range orogen is estimated at 250–700 km.

A tectonic model for the formation of the main structures of North Alaska is shown on Fig. 6. During the Middle Jurassic epoch, the older ensimatic Koyukuk island-arc complex formed in the south of the Angayucham Ocean (Box, 1985; Moore et al., 1994, 1997). Underwater highs and guyots with intraplate basalts were located in the northern part of this oceanic basin (Fig. 6A).

The oceanic crust was completely consumed in the subduction zone at around 145–140 Ma, which was followed by collision of the island arc with the passive margin of the continent at ca. 130–125 Ma (Fig. 6B). Subduction of the passive margin was accompanied by the formation of a foldand-thrust structure in the upper crust (thin-skinned tectonics) in the northern Brooks Range and by underplating in the southern part (Plafker and Berg, 1994). The Neoproterozoic basement and its lower Paleozoic sedimentary cover were subducted to depths of approximately 35–45 km. The sheeted structure of the basalt-silica rocks implies the existence of an accretionary wedge (Pallister et al., 1989). The end of deformation on Brooks Range is limited by the Valanginian (140–133 Ma) and the Hauterivian (130 Ma) stages as these are the youngest strata not affected by it.

The general similarity of tectonic models for the arctic margins of Chukotka and North Alaska is obvious. (1) The arctic margins of Chukotka and Alaska have a Neoproterozoic basement and an intricately deformed sedimentary cover, which carry indicators of the Ellesmere orogeny. (2) The South Anyui and Angayucham oceanic basins have a common geologic history from their opening in the late Paleozoic time up to their closing in the end of the Early Cretaceous, which is evidence for considering them as branches of the single Proto-Arctic Ocean whose northern margin was passive and the southern one – active. (3) Destruction of the oceanic and then continental lithosphere took place southward. (4) The southern vergence of subduction zones in the South Anyui and Angayucham oceanic basins caused an extension on the margin of the North American continent. (5) Collision of the structures of the passive and active margins of both basins took place simultaneously at the end of the Early Cretaceous and ended in the Hauterivian–Barremian. (6) The collision caused the formation of fold-andthrust structures of northern vergence in the Chukotka folded area and in the Brooks Range orogen.

These basic points of the tectonic models for Chukotka and Alaska must be taken into account when creating a geodynamic model for the formation of the Amerasian Basin.

The main problem of published kinematic reconstructions of the Amerasian Basin, including the rotation hypothesis or the back-arc spreading model, is that they do not answer the question of what physical mechanism caused the kinematics of the drifting tectonic plates. The breakup of Arctic Alaska and Chukotka from the Canadian arctic continental margin and subsequent drifting away is one example. The great variety of published kinematic models of the late Mesozoic development of the Amerasian Basin lithosphere indicates a serious crisis in the understanding of this process because none of these models can be selected as the most realistic and physically substantiated from a purely kinematic approach. Using this approach, each author, based on available geological and geophysical materials, tries to visualize the tectonic development of the region by means of kinematic schemes of drift of lithospheric blocks accompanied by areas of extension, compression or transforms, while at the same time being distracted from analyzing the causes of the drift. In the best cases, the authors limit themselves to mentioning general geodynamic concepts like mantle plumes or back-arc extension. To overcome this crisis one must transcend the purely kinematic approach and present a quantitative geodynamic model that would include flows in the upper mantle and that would be capable to explain the kinematics of the drift of lithospheric blocks and their deformation. In recent years, Russian researchers proposed such a geodynamic model for the development of the Arctic lithosphere for the Late Cretaceous–Cenozoic in the context of scientific validation of the Russian Federation’s claim in the UN for broadening the continental shelf of the Arctic Ocean (Lobkovsky et al., 2011, 2013; Laverov et al., 2013).

This model is based on the idea of the existence of a horizontally extended convection cell in the upper mantle under the Arctic and Northeast Asia that would have a length of several thousand kilometers (Fig. 7). The subducting Pacific plate lithosphere is involved in this convection, moving along the base of the upper mantle towards the continent from the subduction zone, forming the lower segment of the cell. The closing upper segment of the cell forms a reverse flow in the mantle under the lithosphere towards the subduction zone. This is actually the driving force that determines the surface kinematics of the crustal blocks and the deformation of the lithosphere.

This model is based on seismic tomography data for the upper mantle in the transition area from the Pacific Ocean to East Asia and the Arctic (Zhao, 2004, 2009). Recently it has received important support from results of detailed geochemical studies of the composition of rocks samples from Gakkel Ridge. These turned out to carry traces and markers of rocks from the crust that was consumed in the subduction zone that is located at a distance of several thousand km, which is in good agreement with the upper mantle convection model (Richter et al., 2020). Notably, a recent study (Lobkovskii and Ramazanov, 2021) obtained strict quantitative validation for our model based on analytical and numerical solutions of equations of hydrodynamics of viscous fluid. These equations were used on a mathematical problem describing a developing convection in a long horizontal layer of the upper mantle adjacent to a subduction zone. Because our new geodynamic model is universal, it can be used when analyzing older Jurassic–Cretaceous development stages for the East Arctic, when the Canada Basin was forming and the South Anyui and Angayucham paleo-basins were closing.

From the basic geodynamic model (Lobkovsky et al., 2011, 2013; Laverov et al., 2013), it would be natural to assume that the mechanism of tearing of the Alaska and Chukotka blocks from the Canadian Arctic margin was determined by basal drag of the Amerasian lithosphere by the horizontal flow in the mantle directed from this margin towards the Pacific. This horizontal flow under the lithosphere is part of a closed convection cell in the upper mantle that connects the downward mantle flow in the large-scale Pacific subduction zoned (involving the Uda–Murgal, Gravina–Nutzotin, Wrangelia island arcs, etc.) (Fig. 8) to the upward flow under the Canadian margin. This cell combines the upper and lower horizontal flows in the mantle into a single unified hydrodynamic system. The two flows move in opposite directions. The lower one, including the subducting lithosphere moves along the upper-lower mantle boundary from the Pacific subduction zone towards the Canadian margin. The upper flow beneath the lithosphere moves from the margin of the continent towards the Pacific Ocean, causing deformation and drift of the lithosphere by basal drag, causing crustal blocks to break off the continental margin (Figs. 7, 8).

Because of the arched shape of the Pacific subduction zone, the reverse flow beneath the lithosphere from the Canadian margin towards the Pacific Ocean has a two-tier divergent structure. The flow lines close to the eastern Alaskan part of the breaking away band of blocks are directed to the southeast, while the flow lines in the central Chukotka part of the breaking off margin segment are directed to the south. This leads to a fan-like divergence of Alaska and Chukotka during drift that is accompanied by the formation of the Canada Basin in the back area and by the closing of the adjacent Angayucham and South Anyui paleo-oceans (Figs. 8, 9, 10).

It should be noted that the model considered here for the breaking away of lithospheric blocks from the continental margin and the opening of the Canada Basin is fundamentally different from the known back-arc mechanism of opening of marginal basins. This is because the subduction of the Pacific Ocean lithosphere involved in convection in the upper mantle, especially its southern part located in front of the Chukotka block, is at a distance of several thousand kilometers from the continental margin. The proposed mechanism for lithospheric break-off can be called a subduction-convection mechanism, placing emphasis on the development of convection in the upper mantle that is connected to the subduction process.

One of the important deductions from this model is the assumed diffuse character of the deformation area in the lithosphere to the west of the southward drifting Chukotka block during the closing of the South Anyui Ocean (Lobkovskiy et al., 2021). Such a configuration of diffuse deformation of the Amerasian lithosphere is due to its basal drag with the upper mantle currents beneath the lithosphere in a wide area. As already noted, purely kinematic schemes (Kazmin et al., 2014; Nikishin et al., 2021) out of touch with the physical mechanism of drift and deformation of the Amerasian lithosphere by its basal drag with the upper mantle reverse flow are not capable of giving a noncontradictory explanation of the general evolution of the region.

The model we propose for the development of diffuse deformation in the western part of the Amerasian lithosphere accompanying the opening of the Canada Basin and synchronous closing of the South Anyui Ocean solves the issue of space naturally. Unlike the rotation model, it does not require the formation of significant areas of oceanic crust during the opening of the Canada Basin. In this model, the diffuse deformation in part of the Amerasian lithosphere caused by the ductile current of the reverse flow in the upper mantle cell under the lithosphere leads to the closing of the South Anyui Ocean and the accretion of all tectonic units of the Amerasian Basin continental margin to the Siberian continental margin along its entire length. This leads to the formation of the South Anyui suture from the New Siberian Islands to the eastern border of Chukotka (Figs. 8, 9). At the same time, the rate of deformation of the Amerasian lithosphere in our model must decrease to the west corresponding to the decrease rate of the area of the South Anyui Ocean that was coneshaped with the vortex close to the New Siberian Islands. Note that the lithospheric deformations occurring on a large area of the Amerasian Basin were accompanied during some periods by strong magmatism. This mainly concerns large igneous provinces that were formed in the Early (130 Ma) and Late (80 Ma) Cretaceous in the western part of the Amerasian plate (the HALIP events) (Grantz et al., 1998; Buchan and Ernst, 2006; Dobretsov et al., 2013).

Let us now consider the questions related to the closing of the Angayucham and South Anyui oceans and to the collision processes between the Alaska and Chukotka blocks and ensimatic island arcs, microcontinents (the Yukon-Tanana and Rubi terranes) and the Siberian margin (Kolyma–Omolon superterrane). As noted above, during the closing of the Angayucham Ocean, the passive margin of the Alaska block converged with the ensimatic Koyukuk island arc in the Early Cretaceous (Fig. 6). At the same time, the subduction that caused the closing of the Angayucham Ocean was dipping to the southeast. In the framework of our general geodynamic model, the large-scale subduction of the Pacific Ocean lithosphere with a northern dip reaches the base of the upper mantle. There it transitions into a horizontal layer flowing along the base of the upper mantle towards the Canadian margin and forms the stable lower tier of the subduction-convection system. The process of closing of the Angayucham paleo-ocean (Figs. 6, 8, 10) accompanied by the collision of the Alaska block with the Koyukuk island arc developed because of the convergent southeast-dipping subduction that took place in the upper tier of the system due to reverse flow in the upper mantle.

The subducting plate of the upper tier reaches a depth with dominating lower flow, bends and joins this current. Thus, the two-tier subduction system is formed (Fig. 10). It fittingly explains the phenomenon of opposite vergence in the “outer” Pacific Ocean subduction zone and the “inner” subduction zone located in the Angayucham Ocean. This ensures the closing of this paleo-ocean and subsequent collision processes. The closing of the South Anyui Ocean followed a similar two-tier subduction system, followed by the collision of the Chukotka block with the Siberian margin in the Late Jurassic–Early Cretaceous period (Figs. 4, 10), which is represented by the Kolyma–Omolon crustal block.

When analyzing the proposed geodynamic model, one should note the geometric relationships between the directions of subduction zones showing the directions of flows under the lithosphere, and the orientation of the Canadian arctic margin, as well as the forming margin of eastern Siberia. As shown in Fig. 8, the main Pacific Ocean subduction zone that determines the flow direction in the upper mantle convection cell is oriented at an angle to the Canadian margin. Therefore, the reverse flow under the lithosphere will have both an orthogonal and a strike-slip component relative to the margin. In addition to rifting and the formation of the Canada Basin, this will lead to the formation of a large-scale sinistral strike-slip along the Canadian margin and, at the same time, to a similar dextral strike-slip transform zone (Sokolov et al., 2015) along the closing linear area of the South Anyui Ocean. The existence of such mega-strike-slip zones on continental margins in the Arctic is fully supported by geologic data (Lobkovsky et al., 2013; Sokolov et al., 2002; 2015; Shipilov, 2016; McClelland et al., 2021).

Above we noted that our basic geodynamic model is universal. Let us state again that in its initial version (Fig. 7) it was used to explain extension processes that took place in the Eurasian and Amerasian basins of the Arctic in the Late Cretaceous–Cenozoic interval (Lobkovsky et al., 2011, 2013; Laverov et al., 2013). In subsequent publications (Lobkovskiy et al., 2021), including the present article, a generalized version of this model is used to analyze an older period of development of the Amerasian Basin during the Late Jurassic–Early Cretaceous.

In our opinion, this model can be also used for other subductional-collisional structures taking into account regional particularities, as well as other time intervals. Specifically, it is quite suitable for describing the late Cenozoic tectonic and volcanic processes in the area of the lithosphere between the eastern margin of Australia and New Zealand (Mather et al., 2020).

The formation of the arctic margins of Chukotka and North Alaska followed a similar scenario both structurally and in terms of time: the collisional stage in the Late Jurassic–Early Cretaceous and fold-and-thrust structures of northern vergence that formed during the Chukotka (Brooks) orogeny in the end of the Early Cretaceous.

The arctic margin of Chukotka formed during the closing of the South Anyui oceanic basin and subsequent collision of the Chukotka microcontinent with the active continental margin of Siberia. Destruction of oceanic lithosphere took place southward in subduction zones of the Kulpolney ensimatic arc (Oxfordian–Kimmeridgian) and the Oloi volcanic belt (Late Jurassic–Early Cretaceous) on the margin of the Kolyma–Omolon superterrane.

The formation of the arctic margin of Alaska took place during the closing of the Angayucham Ocean during a south-vergence subduction of oceanic crust under the ensimatic Koyukuk arc and subsequent obduction of oceanic and island-arc complexes onto the passive continental margin of the Arctic platform (Arctic Alaska superterrane).

The South Anyui and Angayucham oceanic basins have a common geologic history from the beginning of their formation in the late Paleozoic to their closing in the end of the Early Cretaceous. This is grounds for considering them as branches of the same Proto-Arctic Ocean, whose northern margin was a passive one and the southern – an active one.

To explain the tectonic mechanisms of formation of the arctic margins of Chukotka and Alaska, we propose the subduction-convection geodynamic model for the opening of the Amerasian Basin in the Late Triassic–Early Cretaceous.

The oceanic lithosphere plunging in the Pacific Ocean subduction zone drifted horizontally along the base of the upper mantle towards the continent, forming the lower segment of a convection cell. The closing upper segment of the cell formed a reverse flow in the mantle under the lithosphere towards the subduction zone. These processes served as the driving force that determined the surface kinematics of lithospheric blocks. Due to basal drag of the Amerasian lithosphere by the horizontal upper mantle flow towards the Pacific, the Alaska and Chukotka blocks broke off together from the Canadian arctic margin. This was accompanied by diffuse deformation that caused the formation of structures with thinned continental crust in the western part of the Amerasian Basin and oceanic crust in the Canada Basin.

We are thankful to reviewers M.K. Kos’ko and A.I. Khanchuk for attentively reading the article and making remarks and recommendations that enabled us to improve on the manuscript.

This work was supported by grants from the Russian Science Foundation: projects 19-17-00091 (development of an understanding on the tectonics of the western framing of the Amerasian basin and the Mendeleev Rise) and 20-17-00197 (development of tectonic models for Chukotka and Alaska). It also received funding from Basic research projects 0135-2019-0078 and 0128-2021-0004.