G.P. Avetisov

Seismologic data on the deep structure of the New Siberian Islands and adjacent sea areas

         Concepts of the deep crustal structure in the Arctic are based mainly on gravimetric and magnetic surveys, in contrast to many other regions in the USSR. It is obviously necessary to use seismic data, particularly in relation to the economic task of prospecting in the shelf seas of the Arctic Ocean. Costly deep-seismic sounding (DSS) studies in the inaccessible Arctic regions are complicated or often impossible. Particular interest is then attached to seismological observations by means of the earthquake transformed wave method (ETWM), which is widely used in various regions and provide valuable evidence on the structure of the crust and upper mantle.

Studies by the ETWM in the New Siberian Islands were performed by Sevmorgeo from 1972 to 1976. The observations were performed at eight points and covered virtually all major islands in the archipelago (fig. 1). The relatively small number of points and the separation of them undoubtedly reduce the informativeness somewhat. This is, however, partly compensated by the increased duration of the observations (up to three months) and by the corresponding substantial volume and high reliability of the information for a single point.

The seismograms were processed in the following ways: 1) determination of the apparent angles of emergence of the P waves from earthquakes e_p and the apparent velocities v*, and the dependence of these parameters on the epicenter distance at D° is used in estimating the behavior of the boundaries, 2) determination of the azimuths at the epicenters a°_obs and comparison with the actual azimuth a°_a, where the differences between the two are due to features of the section if the apparatus is reliably calibrated, 3) determination of the surface velocity v_o by a differential method using e_p and v*, with v_o corresponding to the velocity in a layer whose thickness is approximately equal to the wavelength (1), and 4) identification of earthquake transformed waves of PS type from boundaries in the crust and upper mantle.

The data were selected and the transformed waves were identified on the basis of standard criteria and methods, while the depths of the transformation boundaries were calculated from Hasegava's formula. The velocity characteristics of the section were derived from: v_o data; results of seismic studies by refracted-wave methods along a line through Bunge Land; parametric studies on Kotel'nyy Island (10); DSS results in the Sea of Laptevykh (9); and also transformed-wave studies in the sea northeast of the archipelago. These were used with laboratory determinations on specimens from geologic collections and borehole cores, along with published data on the average seismic-wave speeds in the Earth's crust and in individual layers.

Parameter K = vp/vs (vp and vs are the mean speeds of P and S waves down to the transformation boundary) were estimated from the seismic studies in the archipelago, laboratory measurements, and also the known dependence of K on the lithology, depth, and P-wave speed.

Geophysical characterization of the boundaries was based on the dynamic and kinematic features of the transformed waves, as is usual in ETWM. The dynamic features were estimated from the parameter n, which indicates the proportion of PS waves in a particular group out of the total number of earthquakes used in the processing and from the ratio Aps/App where Aps is the amplitude of the transformed wave in the horizontal components and App is the amplitude of the P wave. The kinematic features were characterized by the values of the transformed-wave delays Dtps relative 'to the push wave and from the spread in this parameter within a group dDtps).

The boundaries may be divided into three classes (3) in terms of the transformed-wave expression: I) reference boundaries forming dominant PS waves (Aps/App > 0.3, n > 60%, dDtps = 0.05 sec), II) reference boundaries forming strong PS waves (Aps/App < 0.2-0.3, n = 40-60%, 0.05 sec < dDtps < 0.1 sec, III) intermediate boundaries forming weak irregular PS waves (Aps/App < 0.2, n < 40%, dDtps > 0.1 sec). However, experience with ETWM shows that the ranges of variation in these parameters from region to region are so large that it is practically impossible to perform an unambiguous geological interpretation of the seismic boundaries. It is also often impossible to solve this problem even in DSS, in spite of the use of velocity criteria. This is due to the variety of real models for the Earth's crust and the differences in the structure and nature of the boundaries.

Only the M surface could be characterized with good reliability from the transformed-wave features. This is associated with the strongest transformed waves in the lower part of the time section. All the boundaries above M (within the crust) are denoted by K with appropriate subscripts (the numbers increase from the upper boundaries to the lower ones, while those below M are denoted by M₁. Agreement between the subscripts in the boundaries, apart from M, at the different points does not mean geologic identity. Data from other geophysical methods and from geological surveying enable one to delineate certain boundaries in the upper part of the section.

Two major crustal blocks are identified from the seismologic and other data within the New Siberian Islands: Kotel'nyy Island, Bunge Land, Malyy Lyakhovskiy Island, Fadeyevskiy Island, New Siberian Island, and Bol'shoy Lyakhovskiy Island. The first block is much more homogeneous and can be divided into sublocks that differ fairly substantially in the type of section, which indicates features of the tectonic conditions in these. In the region of Kotel'nyy Island (table 1 and fig. 2), the crust is not very much differentiated by velocity, and its over-all thickness is 32-33 km. Here there are no boundaries that could be assigned as reference class I. Therefore, it is fairly arbitrary to divide the crust into a consolidated part and a sediment layer, or any more detailed division.

According to the geologic and geophysical data, in Kotel'niy Island there is a thick series of PZ_l-2 carbonate beds with wave speeds of 6.5-6.6 km/sec (2, 10). In this series, there is no reason to expect sharp and persistent velocity boundaries, which is confirmed by the ETWM. On the other hand, the velocities in the upper part of the section are so large as to exceed the usual values for deeper zones in the crust. Therefore, transformed waves of reverse polarity with Dtps = 1.15 ± 0.15 sec recorded at the Temp Station, although very unreliably, are identified with the bottom of the carbonate series. The intensities are low and therefore they are unreliably identified on the seismograms, which is evidently due to features of the boundary, such as a smooth transition into the underlying series. At the Cape Diring Ayan Station, the probability of recording transformed waves from the bottom of the carbonate series is even lower on account of the reduced thickness, as is indicated by the fall in v_o here to 6.2 km/sec (v_o = 6.5 km/sec at the Temp Station).

Only suggestions can be made about the lithology of the beds under the carbonate ones. In the Anzhu Islands there are no exposures of beds older than 0₃ which are represented by massive limestones. On Bennett Island there are exposures of sandstones, siltstones, and argillites of 0_1-2 age, and also intercalated argillites, siltstones, and very hard limestones of Cm_2-3 age. On Bol'shoy Lyakhovskiy Island, there are schists of Pr_2-3 age of various compositions.

One cannot assign the carbonate series to a typical sediment layer in our view because of the excessively large seismic-wave speeds. The usual velocities for sedimentary carbonate beds that have not undergone diagenesis are in the range 5-6 km/sec. Very often, lower velocity beds are assigned to the basement only on seismic data (9). On the other hand, geologic data (5, 11) indicate a slight dislocation of the 0, S, and D beds, which is not usual for basement ones. On this basis one should evidently assign these beds to the cover of an ancient platform, but one that has been subject to subsequent tectonic action. The tectonic position of the underlying beds is more undefined, since the age and lithology of them have not been established. In accordance with this, one cannot at present reliably determine the boundary between the sediment cover and the consolidated crust. It is supposed it may be associated with boundary K₃.

The absence of sharp boundaries in the consolidated crust prevents the traditional division into two layers differing in elastic properties. There is a fairly smooth change in elastic characteristics over the section from the top downward to the M surface, at which, to judge from the dynamic characteristics of the transformed waves, the velocity differentiation is sharper. Nevertheless, it should also be approximated by some transition layer.

In the regions of the Bunge Land and Cape Khvoynov stations there are rather more boundaries in the upper and lower parts of the section (table 2 and fig. 2b): apart from the M surface, there are several boundaries within the crust, whose total thickness is 34 km. However, here again there is no class I reference boundary that would indicate considerable velocity discontinuities. Nevertheless, the positions of the class II reference boundaries and particularly of K₄ indicate a clear division of the consolidated crust into two layers approximately equal in thickness divided by inhomogeneities of lower order. The boundary between the sediment cover and the consolidated crust is not a reference one. At Cape Khvoynov Station, this is presumed to be boundary K₃, while at Bunge Land Station it could equally be K₂ or K₃.

The reference boundary in the upper parts of both sections is undoubtedly associated with the top of the carbonate beds whose contact with the overlying Mesozoic-Cenozoic rocks is the sharpest seismic boundary in the sediment cover in the region (10). The carbonate series in Kotel'nyy Island descends to the east, but this is not smooth but rather as a series of fault lines dividing the beds into a series of uplifted and depressed blocks, as established by seismic and gravimetric studies on Bunge Land (10). This is confirmed by the seismologic data, particularly by the increase in e_p by comparison with the values obtained on Kotel'nyy Island and the reduction in v_o (5.4 km/sec) at the Bunge Land Station.

In that case, v_o reflects the average velocity in the layer including the overlying terrigenous beds and the upper part of the carbonate series.

The Bunge Land Station lies directly on the north coast of Sannikov Strait, and the sections obtained by ETWM in the Arctic lie several kilometers to the south because of the positions of the seismic rays. On this basis, one assumes that there" is a fault line somewhere along the zone of transition from the land to the strait, since gravimetric data indicate that the depth of the top of the carbonate beds at the southern end of Bunge Land is not more than 0.8-1 km. This fault evidently extends to the Gulf of Hedenstrom, along the western coast of which gravimetric data indicate a sharp descent in the carbonate series.

Two values of v_o were obtained at Cape Khvoynov Station: west of the station, the value was as in Bunge Land, while to the east, i.e., in the sea, it was much less (3.3 km/sec). Here evidently the carbonate series descends even deeper, and therefore v_o characterizes the overlying low-velocity beds.

The non-inversion boundaries K₃ (Cape Khvoynov) and K₂ (Bunge Land) lie at small depths, which indicates that the thickness of the carbonate series is here less than that in Kotel 'nyy Island.

This descent of the Paleozoic carbonates under the Mesozoic-Cenozoic cover in Bunge Land and Malyy Lyakhovskiy Island indicates that the latter areas constitute a depressed part of the Kotel'nyy Island structure. However, there are clear differences in the crustal sections for the two parts, which indicate distinct tectonic conditions. I therefore have reservations about combining Kotel'nyy Island, Bunge Land, and Malyy Lyakhovskiy Island into one tectonic zone, such as the Kotel'nika central massif (11).

The boundaries at Cape Nerpichiy are very similar to almost (identical with) those at Cape Khvoynov and particularly Bunge Land (table 3, fig. 2), but there is considerable deterioration in the velocity differentiation with depth.

While K₁ and K₂ are identified very reliably and can be assigned as class. II reference boundaries, K₃ is somewhat illdefined, and K₄ and particularly M(?) produce very irregular transformed waves, and they undoubtedly must be classified as intermediate ones. The correlation of the M surface in that case is very ambiguous, and the sole basis for it is the depth. Another feature of the data from Cape Nerpichiy Station is the anomalous behavior of the seismic waves from earthquakes to the south, southwest, and southeast: a deviation of a°_obs from a°_a is 80-90°. The available data indicate that there is a deep fault line or an entire fault system with a strike of 160-340°· southwest of the station.

This deep fault line evidently explains why Fadeyevskiy Island has been subjected to tectonic action that is largely or completely absent in Bunge Land, although it has a deep structure similar to that in the territory to the west. This is also indicated by the refracted-wave data, which indicate that the simple and undisturbed seismic section in Bunge Land becomes much more complicated as Fadeyevskiy Island is approached (10). A consequence of the marked difference in tectonic conditions arise from the bounding systems of deep faults, within and near which there is considerable elimination of the inhomogeneities because of the crushing and displacement. Outside this region one expects an undistorted velocity distribution. The position of the fault zone relative to the station is evidently such that the information on the deep boundaries arises from anomalous parts, while that on the subsurface ones arises from normal ones.

The change in tectonic conditions in the new Siberian Islands is largely related to the Mesozoic tectogenesis. The above evidence indicates that the zone of this extends further to the north in the region of Blagoveshchensk Strait and Fadeyevskiy Island. Some other researchers (6) have drawn the same conclusion from gravitational and aeromagnetic data. The low value of the surface velocity recorded at Cape Nerpichiy (3.5 km/sec) indicates that the high-speed carbonate beds do not influence the value, i.e., the top of them lies at a depth greater than the thickness of the influencing layer. In my view, there is no basis for suggesting that the carbonate beds vanish to the east of the New Siberian Islands. The top of these apparently relates to horizon Ф, which is regularly traced at depths of 3-3.5 km in the East Siberian Sea (Avetisov, 1978). The carbonate beds are also associated with a structural stage characterized by velocities of 3.0-4.5 km/sec (7) as found in the Lomonosov ridge and· in the zone where it adjoins the shelf. Finally, the carbonate beds are exposed in Vrangel' Island, where they constitute the main part of the middle structural stage, which is assignable as a quasiplatform cover over a Baykalian basement (8).

Therefore, the upper transformation boundary near Cape Nerpichiy Station (boundary K_l) should be identified with the top of the carbonates, which here is assigned not to the sediment cover but to the basement. The depth of this (4 km) corresponds to that obtained to the east of Cape Khvoynov Station. It may be that these are points on a terrace in the relief of the carbonate-bed surface that runs along the coast of Malyy Lyakhovskiy Island through Sannikov Strait to the east coast of Hedenstrom Gulf. Another step out has been indicated above and is traced evidently along the southern coasts of Kotel'nyy Island and Bunge Land and the west coast of Hedenstrom Gulf.

In the section at the New Siberian Station, one largely finds the same number of boundaries at the same depths as for Cape Nerpichiy, Cape Khvoynov, and Bunge Land. However, the character of the boundaries is very different (table 4, fig. 2): all the boundaries give rise to very strong regular transformed waves and can be assigned as class I reference boundaries.

The intensity of the PS waves de­pends mainly on the velocity discontinuity at the boundary, the absorbing properties of the medium for P and S waves, and the structure of the boundary (zone). Theoretical calculations (12) indicate that a high transformed-wave intensity such as is recorded at the New Siberian Station is possible for a boundary of the first kind (velocity discontinuity) and real v₁/v₂ at the boundary of 0.7-0.9 if there is a ratio of two between the absorption coefficients for the P and S waves a_p/a_s  In the case of a boundary  of type III (sequence of alternating layers with velocity inversion), such an intensity is obtained even without allowance for absorption. Under real conditions it is difficult to determine unambiguously -the reason for the sharp rise in the transformed-wave intensity.

DSS with the use of subcritical reflections in some cases has revealed boundaries of type III in the lower parts of the crust in which the overall thickness h_S of the set of beds is not more than 1 km. At the frequencies of 0.8-1.2 Hz used in ETWM, h_S should be of the order of 2-2.5 km in order to produce strong PS waves. One cannot rule out the existence of such transition zones in the consolidated crust. In particular, such a model is acceptable for K₄ under the New Siberian Station, as this gives particularly strong transformed waves. The boundaries in the sediment layer and the base of this are more realistically represented by a type I model in my view. Theoretical calculations indicate that transformed waves from boundaries of type III are very unstable, and the intensities may vary sharply under the influence of many factors, of which the most important for prolonged observations at one point are the angle of incidence and the frequency composition. In our case, a high PS wave intensity was recorded in virtually all cases.

So far, no definite connection has been established between the zones of elevated PS wave intensity and the structure of the region. Very often, the high-intensity transformed waves are encountered in sections with thick terrigenous beds at the surface. For example, PS waves with Aps/App > 1 are recorded in the North German depression (12), which for a long time was under platformal conditions. At the same time, Aps/App of 1.2-1.5 have been recorded at the Avacha Station in Kamchatka (4), i.e., in a region with highly active neotectonics.

There are various viewpoints on the geologic history of the New Siberian Islands (5, 6, 11). The seismologic section near the New Siberian Station is very different from all the others, which undoubtedly indicates specific features of the tectonic conditions there. The transformed-wave data indicate good stratification in the upper part of the section and conforming level lie in the reference reflecting horizons in the western part of the East Siberian Sea directly adjoining the New Siberian Islands and in the zone of function between the shelf and the underwater Lomonosov ridge (7). Therefore, it is best to relate the stability, order, and dynamic expression of the PS waves at the New Siberian Station to an analogous character in the crustal section throughout its thickness.

The above basis for considering that the carbonate beds extend to the east is confirmed by the identification of a clear K₂ boundary at a depth of 3.6 ± 0.5 km at the New Siberian Station. A somewhat smaller depth here agrees with the increase in Vo (4.6 km/sec). The quality of the K₃ boundary is much better here than at other points, which probably indicates some reduction in the velocities in the overlying beds, and particularly in the carbonates. This agrees well with the values for the stratal velocities obtained to northeast of the archipelago (7).

This fairly sharp change in elastic characteristics in my view is primarily associated with reduced dislocation, which goes with the above conformity in the boundaries to indicate that tectonic conditions in this region were quieter than those in the west. In the upper part of the section, boundary K_l is seen at azimuths of 160-210· from the station, i.e., on passing into the sea, which corresponds to a boundary in the beds lying above the carbonates. It is quite likely that this is related to the top of the Mesozoic terrigenous beds.

A major feature of the lower part of the section is the sharp division of the consolidated crust into two layers. We have seen above that this division occurs only at the Cape Khvoynov and Bunge Land Stations and to a smaller extent at Cape Nerpichiy. On the whole, the deep structure in the region of the New Siberian Station is reliably approximated by the classic three-layer model for the continental crust. The thicknesses of the sediment, granite, and basalt layers are correspondingly 10, 10 and 13 km. According to a listing of the layer parameters for the crust in major tectonic components of the USSR, this relation between the layer thick­nesses corresponds to plates in ancient and young platforms.

The purely seismologic characteristics of the section in the second crustal block, which coincides geographically with Bol'shoy Lyakhovskiy Island, is similar to that for Kotel'niy Island. This part is distinguished mainly on geologic and gravimetric data, and also with allowance for the seismically active suture in the crust along Eterikan Strait. The over-all crustal thickness is here slightly increased (34 km) by comparison with Kotel'niy Island, and surface M is probably represented by a transition layer (table 5 and fig. 2).

However, the internal structures of the crust differ considerably. At Kotel'niy Island, the thicknesses of the consolidated and unconsolidated parts are approximately equal, whereas at Bol'shoy Lyakhovskiy Island the consolidated part is assigned as all beds up to the Permian inclusive as it lies within miogeosyncline and eugeosyncline zones of Mesozoic folding. This means that there are differences in composition and age of at least the upper parts of the consolidated crusts. Boundaries K₁ and K₂ within the crust appear not to  correspond to zones of sharp change in the elastic properties to judge from the transformed-wave dynamics. This applies especially to K₂.· Therefore, under these stations the section is represented by a medium with a fairly weak gradient characterized by a general fairly smooth increase in velocity with depth.

Against the background of this rise there are undoubtedly possible local steps in velocity caused by changes in rock composition and by ones in rock state, but these cannot be observed at the present stage. For example, the reduced value of v_o (5.3 instead of 5.9 km/sec) in the upper part of the section at Dymnoye Station indicates that the thickness of the dislocated Permian beds is somewhat greater (up to 1 km) than that at the Kigilyakh Station, although transformed waves from the bottom of these are not recorded. At Kigilyakh Station, there is very unreliable detection of a boundary at a depth of 48 km below the M boundary. At Dymnoye Station, no stratification of the upper mantle is recorded by seismic-wave speed.

I ana1yzed the v* = f(D°) curves for pairs of observation points at the Tiksi and Ostrovnaya Stations and the e_p = f(D°) curves for each of these stations in order to evaluate the character of the major interfaces in the central and eastern parts of the Sea of Laptevykh. There is a meridionally elongated ridge uplift in the boundary, whose axis runs to the east of the Tiksi Station. This is confirmed by the DSS and ETWM data in the southern part of the Sea of Laptevykh (9) in the region of Tiksi, and in North Yakutia (3), according to which the minimum depths occur for the base of the crust and the reference boundary within the crust in the Sea of Laptevykh: correspondingly 30-32 km and 11-12 km as against 41-42 km and 17-18 km in the North of Yakutia and 32-34 km and 19-20 km in the New Siberian Islands.

To sum up, the deep structure of the region around the New Siberian Islands has generally the same features.as the entire shelf of Eurasia: relative1y thick crust (25-35 km) split up into blocks and of a heterogeneous-continenta1 type. The present stage of investigation is clearly one of reconnaissance, which provides general concepts on the deep structure of the region and which poses some new problems, the resolution of which requires continued and extended observations.

REFERENCES

1. Avetisov, G.P., 1974, PUSH-WAVE EMERGENCE ANGLES INDICATED BY OBSERVATIONS AT STATIONS IN FRANZ JOSEF LAND. In GEOPHYSICAL SURVEY METHODS IN THE ARCTIC: Leningrad, no. 9, p. 96-101.

2. Avetisov, G.P., 1975, PLUTONIC STRUCTURE OF THE EARTH'S CRUST IN THE REGION OF BOL'SHOY LYAKHOVSKIY ISLAND AND THE WESTERN PART OF KOTEL'NYV ISLAND AS IN­DICATED BY SEISMOLOGICAL DATA. In GEOLOGY AND MINERAL RESOURCES OF THE NEW SIBERIAN ISLANDS AND VRANGEL' ISLAND: Leningrad, p. 43-47.

3. Avetisov, G.P., and Piskarev, A.L., 1979, PLUTONIC STRUCTURE OF WEST YAKUTIA INDICATED BY EARTHQUAKE BULK WAVES: Sovetskaya Geologiya, no. 4, p. 114-119.

4. Bu1in, N.K., 1977, PLUTONIC STRUCTURE OF KAMCHATKA AND THE KURIL ISLANDS INDICATED BY SEISMIC DATA: Sovetskaya Geologiya, no. 5, p. 140-147.

5. Vo1'nov, D.A., 1975, GEOLOGIC HISTORY OF THE REGION OF THE NEW SIBERIAN ISLANDS. In GEOLOGY AND MINERAL RESOURCES OF THE NEW SIBERIAN ISLANDS AND VRANGEL' ISLAND: Leningrad, p. 61-71.

6. Genin, B.L., Lipkov, L.Z., and Piskarev, A.L., 1977, BASEMENT STRUCTURE OF THE EAST SIBERIAN ' SHELF WITH REFERENCE TO THE NEW SIBERIAN ARCHIPELAGO. In TECTONICS OF THE ARCTIC AND THE FOLDED BASEMENTS OF SHELF SEDIMENTATION BASINS: Leningrad, p. 86-97.

7. Demenitskaya, R.M., and Kiselev, Yu.G., 1968, STRUCTURE, MORPHOLOGY, AND SEDIMENT COVER OF THE CENTRAL PART OF THE LOMONOSOV RIDGE AS INDICATED BY SEISMIC STUDIES. In GEOPHYSICAL SURVEY METHODS IN THE ARCTIC: vyp. 5, Leningrad, p. 33-46.

8. Kameneva G.I., 1975, STRUCTURE OF THE CENTRAL PART OF VRANGEL' ISLAND. In GEOLOGY AND MINERAL RESOURCES OF THE NEW SIBERIAN ISLANDS AND VRANGEL' ISLAND: Leningrad, p. 72~77.

9. Kogan. A.L., 1974, SEISMIC STUDIES BY DSS-REFRACTED-WAVE METHODS FROM SEA ICE ON THE SHELF IN THE ARCTIC SEAS (EXPERIENCE IN THE LAPTEVYKII SEA). In GEOPHYSICAL SURVEY METHODS IN THE ARCTIC: vyp. 9, Leningrad, p. 33-38.

10. Piskarev, A.L., Avetisov, G.P., Genin, B.L., and Larin, S.M . 1975, STRUCTURE OF BUNGE LAND AND OF THE ZONES OF JUNCTION WITH THE KOTEL'NYY AND FADDEYEV ISLANDS. In GEOPHYSICAL SURVEY METHODS IN THE ARCTIC: vyp. 10, 35-40.

11. Vinogradov, V.A., Gaponenko, G.I., Gramberg, I.S., and Shimarayev, V.N., 1976, STRUCTURAL AND FORMATION COMPLEXES IN THE ARCTIC SHELF OF EAST SIBERIA: Sovetskaya Geologiya, no. 9, p. 23­33.

12. Pomerantseva, I.V., and Mozzhenko, A.N., 1977, SEISMIC STUDIES WITH THE ZEMLYA APPARATUS: Izd-vo Nedra, Moscow.

Translated from Glubinnaya struktura Novosibirskikh ostrovov i prilegayushchikh akvatoriy po seysmologicheskim dannym, Sovetskaya Geologiya, 1982, no. 11, p. 113-122. The author is with the Sevmorgeo Organization.