RUSSIAN JOURNAL OF EARTH SCIENCES VOL. 8, ES1002, doi:10.2205/2005ES000185, 2006

Discussion

[54]  We believe that the difference observed in the positions of the European and Siberian poles must stem from any of the following causes:

Tectonics

[55]  It can be supposed that one of the potential causes responsible for the divergence of the Permian-Triassic poles of Siberia and Europe were the relative displacements of these continental blocks during the Mesozoic and Cenozoic periods of time.

[56]  The problem of the relative movements of the Siberian and East European platforms has been discussed repeatedly by many Russian geologists. Using the paleomagnetic data available, Khramov [1982] inferred the movement of the northern edge of the Siberian platform away from the East European platform.

[57]  Somewhat later, using the criteria of paleomagnetic reliability, Bazhenov and Mossakovskii [1986] performed a careful selection of the paleomagnetic data available for Siberia and East Europe, which allowed them to prove a notable difference in the positions of the respective Early Triassic poles. This difference was interpreted by them as the evidence proving the counter-clockwise rotation of the Siberian Precambrian continental block relative to the East European one by the value of about 10o, assuming the rotation pole to be located in the area of North Kazakhstan. The analysis of the specific pattern of the distribution of the Early Mesozoic compression and extension structural features at the periphery of the Siberian Platform carried out by Bazhenov and Mossakovskii [1986] seemed to confirm this conclusion. They noted, in particular, that the formation of a system of Triassic grabens in the west of Siberia can be explained by this hypothesis, too.

[58]  The formation history of the West Siberian grabens is still a matter of discussion, no unambiguous answer being found so far. A brief review of the work done in this field was offered by Kremenetsky et al. [2002, p. 75]. The results of interpreting the numerous studies carried out in this region suggest the West Siberian platform includes a thick (to 15 km) Meso-Cenozoic sedimentary basin resting on the Paleozoic and Proterozoic folded basement of still unknown composition. Associated with the latter are the submeridional linear mostly positive gravity anomalies, ranging between 300 km and 500 km in size and varying greatly in terms of their interpretation [Kremenetsky et al., 2002]. For instance, Aplonov [2000], who discussed this problem in many of his papers, assumed the presence of the Ob paleoocean of a submeridional strike, the rifting stage of which had begun (simultaneously with those of the other rifts) about 240-230 Ma ago, and its short-term spreading stage resulted in the 200- to 300-kilometer spreading of the rift sides and was completed about 215 Ma ago. S. V. Aplonov believes that the spreading of the hypothetical Ob paleoocean resulted in the clockwise rotation of Siberia, relative to the East European Platform, by about 12-14o around the rotation pole situated south of the 60th parallel.

[59]  It should be noted, however, that in the case of this rotation the Siberian pole must have been displaced eastward relative to the European pole, that is, the situation must have been opposite to the observed one (Figure 3).

[60]  In contrast to the view proposed by Bazhenov and Mossakovskii [1986] and Aplonov [2000], there are data which suggest the West Siberian rifts degenerated northward, which is imprinted in the lower number and poor expression of their deep-seated geophysical indications. In particular, Bogdanov et al. [1998] reported the cross-size of the Koltogor-Urengoi rift is 120-130 km in the area of the Tyumen superdeep hole (TSD-6), the size of the rift valley being about 1.5 km across. In the Arctic region the width of the rift valley is not more than 50-70 km, the depth of the trough diminishing to a few hundred meters. Farther northward the rift attenuates more rapidly and vanishes toward the Kara Sea. Similar data are available for the Khudosey Rift.

[61]  It is worth noting that he hypothesis advanced by S. V. Aplonov for the existence of the Ob paleoocean doubted by the results of drilling the Tyumen superdeep hole (TSD-6) which was drilled in the middle of the Koltogor-Urengoi rift graben inferred in the center of the supposed paleoocean. No oceanic crust has been encountered there. On the contrary, in its depth interval of 6424-7502 m (bottom hole) the hole exposed a sequence of volcanic rocks, mostly low-K tholeitic basalt ranging from P2 to T1 in age, the detailed study of which proved it to be similar to the tholeite of the trap formation of the Siberian Platform [Kremenetsky and Gladkikh, 1997]. Kazanskii et al. [1996] believe that the textures and structures of these basalts suggest that they had flowed in land conditions. Kirichkova et al. [1999] reported the finds of continental plant remains in this depth interval. The age of the West Siberian trap rocks dated by Reichow et al. [2002] using the Ar-Ar method was found to be very close to the age of the trap rocks from the Siberian Platform, which also contradicts the hypothesis offered by Aplonov [2000].

[62]  The analysis of our mean paleomagnetic poles shows that the explanation of their noncoincidence only by the significant relative movements of the cratons discussed calls for the assumption of the significant convergence of these platforms (over the distance of about 8o of the large circle arc) in Late Paleozoic time. This convergence must have been caused by the rotation of Siberia around the Euler pole which was remote significantly from its geometric pole.

[63]  In the case of the rotation of the Siberian platform relative to "Stable" Europe, the Euler pole must have been located at the large circle arc passing across the middle of the arc connecting these poles and perpendicular to it. This pattern shows that the large circle, on which the pole of the Siberian platform rotation must rest, is located significantly far from its geometrical center, this controlling the character of this platform rotation, which could not be a simple strike-slip fault movement at the western margin of the Siberian platform, calling for the significant movement of this platform to the west.

[64]  These large-scale movements of the Siberian platform (about 700-800 km) caused the formation of large compression-type structural features in the area of the modern western margin of the platform. Yet, no geological data confirming the formation of any large compression-type structures have been found thus far. As mentioned above, the territory of West Siberia is known for the wide development of Early Mesozoic grabens, the Triassic and Early Jurassic deposits filling them being often folded [Bochkarev, 1973]. This proves some compression episode in the Mesozoic history of this area, the scale of which being incomparable with the compression that might have been produced by the above-mentioned convergence of the Siberian and East European platforms.

[65]  The only large-scale compression-type structural feature between the East European and Siberian platforms is the Ural fold-mountain belt, which shows the traces of both Mesozoic and Cenozoic tectonic activity including compression and extension. Yet, first, the scale of the compression structural features produced after the Late Hercynian orogeny corresponds to the maximum compression of a few hundred meters which is incomparable with the compression estimate of hundreds of kilometers. Secondly, the Mz-Kz tectonic activity was marked mainly by longitudinal faults [Bachmanov et al., 2001].

[66]  Thus, we reject the possibility of explaining the difference between the Permian-Triassic poles of Siberia and Europe by their relative tectonic movements.

Age

[67]  By the present time a fairly large number of data have been accumulated [Bogdanov et al., 1998], which prove that the Permian-Triassic igneous activity in the region of the Siberian Platform continued not longer than 10-15 Ma, and that most of the trap rocks were formed in the time interval of 255-253 Ma to 248-244 Ma [Zolotukhin et al., 1996]. Some researchers [Gurevich et al., 1995; Renne et al., 1995] suggest that the most active period of trap volcanism, when huge volumes of basalt lava flowing on the ground surface, might have lasted during a geologically short interval of time causing the death of living organisms and radical changes in the biocenosis at the Paleozoic and Mesozoic boundary some 250 Ma ago. This conclusion was confirmed by the recent U-Pb datings of the rocks from the upper and lower parts of trap rock complexes of the Maimecha-Kotui region reported recently by Kamo [2003].

[68]  Consequently, the time during which the study rocks had been emplaced (and hence the age of the paleomagnetic poles obtained) can be placed in the interval of 255-244 Ma and, hence, can be taken, with high probability, to be close to the Permian-Triassic boundary which has been dated recently as close to the age value of 251.4 pm 0.3 Ma [Bowring et al., 1998]. On the other hand, since the age of the basalts from the Podkamennaya Tunguska R. Valley was found using the isochronous 39Ar/40Ar method to be 238-248 Ma by Zolotukhin et al. [1996], it cannot be excluded that the trap magnetism had not been completed after the flow of the bulk of effusive rocks in the north of the Siberian Platform.

[69]  In any case, the isotopic and biostratigraphic data available [Distler and Kunilova, 1994] suggest that the accumulation of the trap rocks began the very end of the Permian and was completed at the very beginning of the Triassic. In spite of the potentially short time of the trap rock flow, it should be noted that the data obtained in this study might show the fairly good averaging of the secular variations of the magnetic field. The basis for this assumption is the fact that these data were obtained for the rocks which had been magnetized during the epochs of both normal and reversed polarity, that is, during the time of at least several dozens of thousand years.

[70]  The bulk of the European data were obtained for sedimentary rocks. In terms of their biostratigraphy these rocks compose the Late Permian (Thuringian) and Early Triassic (Indian-Olenekian) beds which were deposited immediately below and above the Permian-Triassic boundary. Menning [1995] believes that these beds accumulated in the time interval of 240-260 Ma. Only one paleomagnetic determination of those used to calculate the average European pole was obtained for igneous rocks, namely, for the Lunner dikes. These dikes were dated using the modern Ar-Ar method and found to be 243 pm 5 Ma. Proceeding from the very short time of the Siberian trap accumulation, it can be expected that they have a more narrow age range than the European objects. However, since the European ages of our data sample are distributed roughly symmetrically relative to age of the Permian-Triassic boundary, it can be expected that the average age of the European objects is close to that of the Siberian Permian-Triassic traps, and that the difference between their ages can be used to explain the difference between the Siberian and European paleomagnetic poles.

The Non-Dipole Pattern of the Geomagnetic Field

[71]  Another potential explanation of the difference between the Siberian and European paleomagnetic poles is the potential significant contribution of non-dipole components to the Earth magnetic field during the Late Paleozoic and Early Mesozoic.

[72]  To estimate the potential contributions of the quadrupolar and octupolar components to the geomagnetic field at the Paleozoic-Mesozoic boundary, we recalculated the coordinates of the European and Siberian Permian-Triassic poles (that were obtained initially proceeding from the dipole law) using the algorithm similar to the algorithm proposed by Torsvik and Van der Voo [2002], which accounted for the non-dipole character of the field (see Appendix A). New average poles were obtained for Europe (AS and VT) and for Siberia (NSP2 and VP) for each pair of the G2 and G3 values. The values of the quadrupolar ( G2 ) and octupolar ( G3 ) coefficients ranging from - 40% and 40% were recalculated, the values outside of this interval were ranked as improbable.

2005ES000185-fig04
Figure 4
[73]  The results of our calculations are presented in Figure 4, where the G2 and G3 values expressed in percent of the dipole component are plotted along the coordinate axis. The contour lines show the angular distance (gamma) between the compared Siberian and European poles, calculated in terms of the non-dipole law for the respective values of the non-dipole coefficients.

[74]  Shown in Figure 4 is only the region where the gamma angle had a value lower than the critical g cr value for the given G2 and G3 values [McFadden and McElhinny, 1990]. In fact, the G2 and G3 values corresponding to this region are the required solutions for which the differences between the Siberian and European mean paleomagnetic poles become statistically insignificant. It was also of interest to determine the G2 and G3 values responsible for the best convergence of the Siberian and European poles.

The AS and NSP2 poles (Figure 4a)
[75]  showed their best convergence (with the gamma angle between them being close or equal to 0o) in the region where the non-dipole coefficients G2 and G3 showed the values of - 10 to 10% and sim- 10%, respectively. Note that the gamma value becomes zero for G2 = 0 and G3 = - 10%. Therefore the observed difference between the AS and NSP2 poles can be eliminated easily by the assumption of a small (10%) contribution of the octupolar component to the total anomalous Permian-Triassic field.

The AS and VP poles (Figure 4b).
[76]  One can see in this figure that the region of the best agreement between these poles (with the gamma angle between them being close or equal to 1o) extends as a narrow band in the 4th quadrant of the plot in the region where the non-dipole coefficients G2 and G3 show the values of 25% to 40% and of - 20 to - 10%, respectively.

[77]  It should be noted that although the poles become statistically undistinguishable with the minor displacement of the coefficients from zero, the angle between them becomes close to the minimum value only under the condition of the fairly high contribution of both the octupolar and quadrupolar components. Following Van der Voo and Torsvik [2001], it should be noted that the substantial contribution of the quadrupolar component to the geomagnetic field would lead to the notable displacement of the paleoequator position, determined by the paleomagnetic method, from the position based on the study of various paleoclimatic indicators. What actually happens is that this effect is not observed, this fact being confirmed by the results of the study carried out by Kent and Olsen [2000] for the purpose of studying the paleolatitudinal position of the Late Triassic sedimentary basin extending along the eastern margin of the North American continent.

The VT and VP poles (Figure 4c).
[78]  These poles show their maximum convergence (with the gamma value close to 0o) for the relatively small G2 and G3 values equal to 10-12%.

The VT and NSP2 poles (Figure 4d).
[79]  Like in the case of the AS and VP poles, the point of the best convergence of these poles (with the gamma value close to 0o) is displaced significantly into the region of the high values of the quadrupolar coefficient (with the G2 value being - 30%), whose substantial contribution to the geomagnetic field seems to be very doubtful. The value of the G3 coefficient is about - 5%.

2005ES000185-fig05
Figure 5
[80]  The inadequate choice of a geomagnetic field model must cause a greater scatter of the paleomagnetic poles obtained for different objects in the same area. Seemingly, the maximum grouping of these poles could be used as a criterion for choosing some optimum model. It is obvious, however, that the great effects of some other factors on the close grouping of the poles, such as, the local tectonics, the errors of determining the dips and strikes, the inadequate averaging of secular variations, to name but a few, preclude the use of the grouping of regional poles for the solution of this problem. This conclusion is illustrated by the series of curves presented in Figure 5, where the maximum crowding is achieved for different regions and different pole combinations in the case extremely improbable values of the G2 and G3 coefficients.

Inclination Shallowing of the European Data

[81]  Do the above statements prove that the non-dipole components played a significant role in the geomagnetic field of the Permian-Triassic boundary? In spite of the fact that our results generally agree with this hypothesis, this conclusion cannot be made definitely, because the disagreement between the European and Siberian poles might have been caused by some other reason. This reason might have been the potential inclination shallowing of the European paleomagnetic directions, since them have been obtained (except one of them) using sedimentary rocks, in which inclination shallowing is often observed.

[82]  Since the coefficient of inclination shallowing must be evaluated for each particular case separately, and we did not have any results of such studies carried out for the European data available, we could estimate some general averaged value ( f ) of inclination shallowing, for which the compared average poles of the Siberian Platform and "Stable" Europe would not show any statistical difference.

[83]  We used the following relationship for the mean European pole recalculation:

2005ES000185-fig06
Figure 6

tan(I observed) = ftimes tan(I field)

(where I observed is the mean inclination obtained from the mean European pole (which was calculated, using the dipole law) for the average European site, and I field is the inclination of the geomagnetic field during the rocks magnetization), proved empirically and known as the King's Rule [Barton and McFadden, 1996; King, 1955]. As a result we found the variation of the gamma angle between the pairs of the poles as a function of the ratio of the inclination shallowing factor f (see Figure 6), where plotted along the horizontal axis are the values of the f parameter (over the interval of 0 to 1), those plotted along the vertical axis being the values of the gamma angle between the Siberian Pole and the recalculated European one.

[84]  Figure 6a clearly shows that the divergence of the AS and NSP2 poles is minimal for the inclination shallowing factor f = 0.62 (with the gamma angle being higher than 2o), the poles being not different statistically for the f values ranging from 0.46 to 0.91.

[85]  Figure 6b shows that the divergence of the AS and VP poles might have been caused by the magnetic inclinations shallowing in the European values with an average f value ranging from 0.47 to 0.86. The best convergence of the poles was found for f = 0.62.

[86]  Figures 6c and 6d show that the divergence of the AS-NSP2 and VT-NSP2 pole pairs can be explained by the inclination shallowing in the European data, the poles showing no statistical differences over the large interval of the f values ranging from 0.45 to 0.95. However, the minimal angle between them (higher than 1.5o and 4o, respectively) differs from zero. This can be explained by the fact that the Siberian NSP2 pole is remote from the large circle connecting the European pole and the center of Europe.

[87]  To sum up, the observed difference between the Siberian and European poles can be explained by the inclination shallowing in the European sedimentary rocks. It should be noted that the statistically significant difference between the Siberian and European poles can be removed assuming some small inclination shallowing, namely, f =0.9-0.95, associated potentially with some packing of the studied sedimentary rocks. It should be noted that experiments proved the possibility of even some greater inclination shallowing in sedimentary rocks, the f value of which may be as high as 0.4 [McFadden and McElhinny, 1990].

[88]  Interesting observation follows from the comparison of the average Siberian poles with the pole having the coordinates Plat=53.0 and Plong=152.9, which was obtained by averaging the data available for the Lunner dikes and the Esterel volcanic rocks (see Table 3). Although the latter were discarded from the data samples used because their age (261 Ma) is formally beyond the age range used in this study (240-260 Ma), this does not preclude the possibility that the pole obtained by the averaging of their pole with the pole of the Lunner dikes may turn out to be close to the true European paleomagnetic pole of the Permian-Triassic boundary. Our comparison shows that the mean pole obtained for the European igneous rocks discussed resides in the close vicinity of the Siberian poles: VP (with the angular distance of 3.3o) and NSP2 (with the angular distance of 4.2o), inside their confidence circles. This circumstance can be treated as another indication of the potential inclination shallowing in some of the European paleomagnetic directions.

2005ES000185-fig07
Figure 7
[89]  In practical viewpoint, the important parameter is the space error in determining the paleolatitudes without using the non-dipole components of the magnetic field or ignoring the potential effect of inclination shallowing. Figure 7 shows the latitude dependence of errors in determining paleolatitudes when the GAD hypothesis is used to the components of the geomagnetic field differing from the dipole field (Figure 7, blue curve), and also by neglecting the effect of inclination shallowing phenomen (Figure 7, red curve).

[90]  Seemingly, the choice of the hypothesis which is more suitable for explaining the difference between the European and Siberian Permian-Triassic poles can be made using the recent paleomagnetic determination obtained using the Semeitau rocks of the same age (Kazakhstan) [Lyons et al., 2002]: Plat=56o, Plong=139o, N=15, K=24.6, and A95=7.9o. This pole, obtained for igneous rocks, differs significantly from the European AS pole ( g/g cr = 12.1o/11.8o) and is not different from the Siberian VP pole ( g/g cr = 6.7o/10.8o) and from the NSP2 pole ( g/g cr = 4.5o/11.1o). This situation must have existed in the case of inclination shallowing in the European data. Therefore, the use of the Kazakhstan pole does not help to chose any of these explanations as the most probable one.

The Instability of Solving the Problem Because of the Small and Inadequate Initial Data Sample

[91]  The number of the data used for averaging the paleomagnetic poles of the Siberian Platform and "Stable" Europe may appear to be insufficient for getting any stable, statistically correct result. In order to verify the effect of this factor, we compared the average paleomagnetic poles obtained by different authors using different criteria for data collecting. Apart from the average poles obtained in this study, we also used the average poles used by Gialanella et al. [1997], Iosifidi et al. [2005], and Lyons et al. [2002].

[92]  The analysis of these data shows that in spite of the different coordinates of the average Permian-Triassic poles of Siberia and Europe, their relative positions remain to be stable: the average pole of Europe is invariably displaced to the southeast relative to the average pole of the Siberian platform, being located at a greater distance from Europe than the Siberian pole (see Table 4).

[93]  To sum up, the observed divergence of the poles is of the systematic type, rather than being a consequence of some inadequate data sample.


RJES

Citation: Veselovskiy, R. V., and V. E. Pavlov (2006), New paleomagnetic data for the Permian-Triassic Trap rocks of Siberia and the problem of a non-dipole geomagnetic field at the Paleozoic-Mesozoic boundary, Russ. J. Earth Sci., 8, ES1002, doi:10.2205/2005ES000185.

Copyright 2006 by the Russian Journal of Earth Sciences

Powered by TeXWeb (Win32, v.2.0).