RUSSIAN JOURNAL OF EARTH SCIENCES VOL. 10, ES4001, doi:10.2205/2007ES000218, 2008

6. Geodynamic Interpretation of Results

6.1. Clusters of Geophysical Parameters

[49]  A model of statistic cluster zoning for the Atlantic consisting of 15 stable combinations of parameters used (see Table 1) resulted from calculations using the procedure (see Section 4) and solution selection algorithm (see Section 5) based on parameters (see Section 3). The procedure being based on distance assessment in multi-dimensional space, calculations were made for standardized parameter (parameters of identical dimension with zero mean and unit variation). It is not quite clear if normalizing statistic moments should be calculated only for the study area or for the Earth as a whole. The authors preferred the latter version, because otherwise it would be difficult to quantitatively compare the results from different regions as parameter norm vary from area to area. All the data used in the present work are presented in matrixes for the entire Earth and parameter standardization was carried out for the above area. All the parameters in the study area have extreme values close to absolute minima and maxima except total seismic moment. Its mean values obtained from clusters by an order of magnitude 4-5 less maximum reported from the Pacific island arc zones not falling into the region studied. Nevertheless, this parameter has also been normalized by global value. It resulted in higher deviation of the parameter by clusters in the Atlantic (see Table 1). In such a case central values by a given parameter for clusters obtained become informative.

[50]  Figure 11 shows cluster profiles for parameter central values in dimensionless standardized coordinates according to calculated parameters. These profiles imply that each parameter value is involved in cluster combinations reflecting so nearly all the main pms range of values (equal to 1). All principal parameter values are thus incorporated in one or another stable combination i.e. cluster. The exceptions are the seismic moment, as mentioned above and Love wave tomography exhibits only positive values due to oceanic study area (negative values are attributed to continents). The above is due to the fact that values normalization was based on the known information for the entire Earth.

2007ES000218-fig12
Figure 12
[51]  The zoning obtained for distribution of parameters combination selected for geodynamical analysis is represented as a map in the end of calculation. A corresponding number of cluster is assigned to each of the one-degree cell in which it falls due to calculations. This cell gets a unique color. As the result of this procedure the map of geodynamic zonation for the Atlantic lithosphere (see Figure 12) was compiled. The above along with Table 1 represents the main result of analysis made. Later the readers will be mainly referred to this map.

[52]  Geodynamical zonation of the Atlantic lithosphere resulted in obtaining of 15 stable clusters that could be conventionally divided into 4 unequal groups according to the main ocean structure zones:

[53]  group 1 for mid-oceanic ridge (7 clusters - 2, 5, 7, 10, 11, 8, 14);

[54]  group 2 for deep sea depressions (2 clusters - 6, 12);

[55]  group 3 for continental margins (4 clusters - 4, 9, 13, 15);

[56]  group 4 for superimposed effects (2 clusters - 1, 3).

[57]  Table 2 shows calculation for areas, occupied by each cluster, accounted for change in size of degree cell at high latitudes.

[58]  74.75 mln km 2 was totally defined in the region. Cluster 6 (basin group) occupies the largest area of 16.06 mln km 2 (21.5%). These include basins with maximal water depth, sediment thickness, Bougue and Love anomalies, tomography by S-waves. Cluster 5 (ridge group) has a minimal area of 0.89 mln km 2 (1.2%). These are parts of high heat flow along the Mid-Atlantic Ridge; cluster 14 (0.84 mln km 2, 1.1%) and cluster 15 (0.57 mln km 2, 0.8%) present the region north of Iceland with high extreme values of P-waves tomography, superimposed on the ridge zone.

[59]  Comparison between the above zonation and segmentation of the Atlantic which can be made separately for each of the parameters (see Sections 3.1.-3.10.) shows that this zonation can't be carried out by means of classification using of or several parameters. Nevertheless there are parameters e.g. bottom topography or heat flow whose effect is stronger than that of the others. However, the visual analysis based only on bottom topography is not sufficient for reliable subdivision of the area analyzed: the ridge will exhibit less distinct structure, if only topography data are used, than if it is statistically compared. This may be attributed to the overloading of the topography as a geometric parameter with other parameters associated with energy release and geometry of the lithosphere inner boundaries. As a result we get a basis for geodynamic classification. Noteworthy that visual correlation leads to comparison of parameters by distinct extreme values of one or another parameter, whereas numerical correlation makes it possible to compare different background (average for separate regions) values in fact not visually discernible, but important in case of parameters to be obtained for extensive areas.

6.2. Description of Groups Obtained in Terms of Geodynamics

[60]  Now we turn to description and geodynamic interpretation of obtained cluster groups. For physical meaning of parameters the readers are referred to Section 3.

6.2.1. Group of mid-oceanic ridges.

2007ES000218-fig13
Figure 13
[61]  Clusters 2, 5, 7, l0, 11, 8 and 14 are assigned to this group (Figure 12, Table 1). To interpret this group together with petrological data by Dmitriev et al. [1999] geophysical and petrological parameters were correlated along the Mid-Atlantic Ridge zone (see Figure 13). Discriminant of D1 > 255.5 and D1 < 255.5 fit, respectively, to spreading and plume basalt associations in terms of the paper cited. The thickness of the basalt crust is a parameter determined from petrologic parameter Na8 depending on basalt magma production rate.

[62]  Cluster 5. This cluster is characterized by maximal heat flow, maximal seismic moment, tomography by S-waves close to minimal, and fairly high elevation. This cluster shows up (see Figure 12, Table 1) in areas where deep plumes are in superposition to the Mid-Atlantic Ridge structures (Iceland, Azores archipelago area, isolate shows north of the fault at 15o20 prime, sublatitudinal forking near Bogdanov and Sierra Leone faults, south of Ascention Islands, and the intersection with the projection of the Cameroon line, Tristan de Cunha, and the Bouve triple junction). It is in these areas, where the above parameters take values close to extreme tomography values (less than -3.5%), whereas high heat flow implies the presence of extensive zones of heated up and partially melted mantle. High density of seismic moments suggests often onset of earthquakes due to magma advance; high average elevation of the Mid-Atlantic Ridge (-2457 m) correlates along the ridge with low value of Bouguer anomaly which is proportional to the mantle depth (or crustal thickness) confirmed by high calculated values for the crustal thickness by parameter Na8 (see Figure l3) and highly prolific plume magmatism by parameter D1. This cluster has a geodynamic meaning in productive magmatism and energy release resulted in emplacement of the thick crust composed of basalts with high P-T values and depths (400-700 km) for uplift of the enriched mantle matter and formation of parental melts (50-100 km at 1400o C) [Dmitriev and Sokolov, 2003]. The lithosphere within this cluster is in the isostatic equilibrium, which was caused by low viscosity of the substrate. Thus, in the Mid-Atlantic area, geophysical and petrologic parameters for this cluster form logically agreeing and physically explicable combination.

[63]  Cluster 10. This cluster is characterized by high heat flow, high seismic moment, low tomography value by S-waves, and low gradient of the magnetic field. In general, cluster 10 has the same parameters as in cluster 5 but they are less pronounced. They differ in deeper elevation, and magnetic gradient, close to minimal. The cluster (see Figure 12, Table 1) shows itself in the same areas as cluster 5, in plan it "frames'' the distribution of extreme values physically and geodynamically characteristic parameters of cluster 10 are the same as those of cluster 5. The low magnetic field suggests, most likely, a low concentration of magnetically susceptible matter in the heated zone and fast disintegration of magnetic properties going from zones occupied by cluster 5. However, main parameters of the cluster correlate well (see Figure 13) with petrologic data.

[64]  Cluster 7. This cluster is characterized by minimal heat flow, fairly high seismic moment, low tomography value by S-waves, and minimal gradient of the magnetic field. In plan, cluster 7 covers in fact the entire area along the Mid-Atlantic Ridge (see Figure 12, Table 1) not occupied by clusters 5 and 10 (except for the site north of Iceland, occupied by clusters 8 and 14). Minimal heat flow in the Mid-Atlantic Ridge area not intersecting with deep plumes cropping out at the surface is easily explicable. According to the data by Podgornykh and Khytorskhoy [1997] heat flow values on the surface reflect in general the geodynamic state of the Earth interior at the measurement site (conductive component correlating with tomography data). However, along the Mid-Atlantic Ridge heat flow varies greatly due to irregular "convective'' component whose value is caused by water washing out of heat along the strong fractures in the nearby ridge areas, and due to heat discharge by degassing of the magmatic substrate [Letyshkov et al., 1997]. A high seismic activity occurs along the entire Mid-Atlantic Ridge as a result of processes responsible for emplacement of the young oceanic crust due to accretion of products of magmatic activity. However, in the main part of the Mid-Atlantic Ridge seismic activity is lower than in zones presented in clusters 5 and 10. Seismic events within cluster 7 area are not so often and of higher magnitude [Dmitriev et al., 1999], but in general manifestation the seismic moment is higher. Low value of tomography by S-waves is characteristic of the entire Mid-Atlantic Ridge due to the mantle uprise and formation of melts, but within cluster 7 it is lower than in case of "plume'' clusters. Based on the data by Dmitriev and Sokolov [2003], this area is marked by low production rate of basalt magmatism of spreading association caused by the adiabatic rise of the depleted mantle from a depth of below 200 km and its weak partial melting at depths of 15-30 km at a temperature of about I200o C. According to the classification of Wilson [1989] these basalts are assigned to N-MORB. Weak, in this case, minimal value of the magnetic field gradient is a characteristic feature of areas adjacent to the Mid-Atlantic Ridge (see Figure 10). This can be attributed to the fact that the field reduced to an altitude of 100 km reflects an integral characteristics of the layer commensurable to an altitude of field sources. In this case, a high frequency spatial magnetic field will be strongly smoothed. However, the main factor responsible for low value of this parameter we ascribe to a low total concentration of the matter in the heated zone with high partial melting in the upper mantle.

[65]  Geodynamically, cluster 7 is similar to clusters 5 and 10: environment with high energy capacity along with the rise of the mantle matter, basalt magmatism of different productivity rate, and high seismicity, specific variation of Bouguer anomaly (see Figures 4, 13), and tomography by S-waves. They differ fundamentally in discrete conditions of magma formation and P-T values. In this case change in space from one type to another may take place along the Mid-Atlantic Ridge within the range of 70-100 km. Figure 13 shows a close correlation of zones along the Mid-Atlantic Ridge with D1 > 255.5 (spreading basalt association) and geophysical parameters reflecting respectively the low crustal thickness, low elevation value, and high Bouguer anomaly.

[66]  Clusters 2 and 11. These clusters present flanks of the Mid-Atlantic Ridge and differ slightly in heat flow and seismic moment. They show a transitional zone from purely ridge clusters 5, 7 and 10 to the ocean floor clusters inheriting some features of the Mid-Atlantic Ridge clusters. In plan, this clusters (see Figure 12, Table 1) occupy a band of about 500 km on either side of the Mid-Atlantic Ridge, within the equatorial Atlantic they "strangle'' the Mid-Atlantic Ridge represented by clusters with high heat flow and form "a cold belt of the equatorial Atlantic'' [Bonatti et al., 1993]. The latter in all other parameters analyzed is similar to an intermediate type with respect to that of basins. All the above clusters of the group are, in general terms, isostatically compensated. Clusters 2 and 11 in the southern Atlantic are deeply incised into the area of basins and join clusters of the superimposed events (Martin Vaz, Rio Grande, Walvis Ridge) as will be discussed later. Cluster 2 having higher heat flow, than that of cluster 11, is more close to clusters of these superimposed events. Cluster 2 also has higher (though low in absolute expression) seismic moment, than cluster 11, implying that it encloses zones of intraplate earthquakes. Noteworthy, that zones of cluster 2 in general are oriented northwestwards, this agrees with published data on the intraplate earthquakes and north-western faults [Mazarovich and Sokolov, 2002]. Of interest also, that these zones branch off the Mid-Atlantic Ridge on sites of plume clusters 5 and 10. Thus, these clusters having resulted from the action of forces, whose superposition gives rise to a system of north-western spurs, start to show themselves from weakened and energetically active plume zones on the Mid-Atlantic Ridge and involve flanks and partly basins in the field of their activity. Of interest also is projection of zones of these clusters into the north-west striking Labrador Basin.

[67]  Based on other parameters, clusters 2 and 11 are presented as follows. Bouguer anomalies acquire values close to maximal of those in basins. Tomography by Love waves shows stable high values characteristic of the oceanic zones. Tomography by P-waves in fact does not differ over most of the region except its northern part. The sedimentary cover within margins (like that on the Mid-Atlantic Ridge) has a minimal thickness of about first hundreds of meters. Tomography by S-waves has low negative values characteristic of marginal zone. Elevation is close to the weighted average for the ocean. Magnetic field gradient is close to minimum.

[68]  Cluster 8. This cluster is marked by the lowest for the Mid-Atlantic Ridge Bouguer anomaly, heat flow above the average value, low tomography by Love waves implying greater thickness of the crust in the region. Besides, this cluster presents high seismic moment, minimal tomography by S-waves, and high elevation value. The cluster (see Figure 12, Table 1) takes up Iceland area. According to the above set of specific parameters this region should be assigned to cluster of type 5 or 10, but it was defined as a separate type. The thing is, that this region is superimposed by a zone of strong negative values, based on tomography by P-waves characteristic of collision areas of the Earth (see Figure 8). Besides, this cluster has high gradient of the magnetic field. The comparison between the main parameters for the region of this cluster and petrological data for the Mid-Atlantic Ridge (see Figure l3) does not doubt the plume nature of cluster 8. Moreover, a deep-seated nature of the heated zone in the Iceland plume area is well known and was widely discussed. The fundamental interpretation of the region is mainly out of question. Superposition of such a outstanding plume event as that in Iceland area, manifested as a specific extremum according to tomography by S-waves (see Figure 7) with a similar extremum according to tomography by P-waves (see Figure 8) seems quite unique for the Earth. This superposition allows to arrive at two different conclusions. First, due to outstanding nature of the Iceland plume event, this zone is the only place on the Earth's surface showing correlation of strong minima both by P- and S-wave tomography and having similar thermal nature. Second, these extreme still differ in nature. It means that the Eurasian collision zone, singled out by minimum of P-wave tomography, incorporates the Iceland area owing to its extensive front. The following facts favor for the second interpretation. The Azores plume, similar in degree of contrast in tomography, has no analogy to the above one, and two types of tomography show no correlation. The crustal thickness in Iceland area inferred from the DSS data reaches 30 km [Udintsev, 1990]; this igneous formation mass is fairly large to affect greatly the tomographic image by the presence of forces responsible for collision of the type existing in Eurasia. At present, the authors cannot give the ultimate answer to this question. A high magnetic field gradient in cluster 8 may be explained in the following way. The integral effect of the field at an altitude of 100 km does not result in complete averaging of the high frequency component because the magnetic anomaly pattern in Iceland area exhibits the presence of extensive zones with a homogeneous field stripped of linear magnetic anomalies [Udintsev, 1990]. Besides, an extreme deep type of basalt magmatism (TOP-Fe) was recorded in Iceland area [Dmitriev et al., 1999], which probably contributes to concentration of magnetically active matter, though the other factors within this cluster should decrease the value of the magnetic field.

[69]  Cluster 14. This cluster is marked by low Bouguer anomaly, high heat flow, maximal values of tomography by P-waves, large seismic moment, great average thickness of the sedimentary cover, positive value of tomography by S-waves, and high gradient of the magnetic field. In plan (see Figure 12, Table 1) this cluster occupies the northern Atlantic north of 71o N and nowhere else. Its main feature is the maximal value of tomography by P-waves occurring north of the minimal value of the same parameters in cluster 8 along with the combination of minima and maxima of P-waves in northern Europe (from the North sea to Scandinavia, see Figure 8). Besides, tomography by S-waves shows for this region a uniform chain of minima, that is in keeping with the Mid-Atlantic Ridge, but here there are positive values characteristic of cold oceanic and continental areas. Even within the cold equatorial segment of the Mid-Atlantic Ridge, tomography by S-waves has not been disturbed in such a way. Bouguer anomaly values are low there approaching to continental ones and extend for the entire width of the northern Atlantic, from Greenland to Scandinavia not showing a distinct minimum of the Mid-Atlantic Ridge (see Figure 4). A high heat flow and marked increase of seismic moment there, nevertheless point to an active phase of rifting. A great thickness of sediments suggests a proximity and activity of the source areas. A high magnetic field gradient has a pattern similar to that of the adjacent land. Such a set of main parameters in cluster 14 area implies that it is unique. The geodynamic interpretation of the region seems to be fairly difficult, and a possible solution may be an assumption of the original continental nature of this block, going now through the early phase of rifting.

6.2.2. Group of deep ocean basins.
[70]  Clusters 6 and 12 were assigned to this group (see Figure 12, Table 1). The interpretation of clusters is given below.

[71]  Cluster 6. This cluster is characterized by low Bouguer anomaly, average for the Earth heat flow, isostatic equilibrium, maximal value by Love waves, sediment thickness of about 700 m, positive tomography by S-waves, maximally deep topography, and low magnetic field gradient. In plan (see Figure 12) the cluster takes up areas of ocean basins as reflected in deep and almost flat bottom topography, but cluster occupies less square compared to generally accepted because it does not include basin areas occupied by tails of the Mid-Atlantic Ridge flank clusters, as well as projection of the continental rise, assigned to a different class of clusters. Minimal values of Bouguer anomaly are indicative of cold and dense state of the lithosphere. Besides, along with the minimal seismic moment and average heat flow point to geodynamic "state of rest'' in which these areas mainly reside. Maximal and positive values respectively by S-waves and Love waves show that this area is typically oceanic and not affected by deep energy transferred across the plume system. It lies very deep (averaging -5012 m) and is overlain by pelagic sediments averaging 685 m in thickness. Disturbances of "the state of rest'', caused by vertical movements, are isostatic in nature; north-west strike slips; delivery of energy of plumes, and other superimposed events, are formalized in other types of clusters.

[72]  Cluster 12. This cluster is a modified cluster 6 presenting the transition from ocean basins to superimposed (see 6.2.4.) or plume areas in the absence of clusters 2 and 11 in plan (see Figure 12) essentially represented as a pair symmetrical to the Mid-Atlantic Ridge (see also Figure 10) suggesting an impulse appearance of this part of the lithosphere in the Mid-Atlantic Ridge area and its further division due to spreading. It occurs in symmetrical formations such as the Agulhas plateau: north of South Georgia islands - Walvis Ridge (northern part) - Rio Grande Rise, areas south of the Blake plateau and south of Guinea plateau, parts of Newfoundland and south of Josephine bank, in Labrador basin area and southern spurs of Rockall plateau, Western flanks of Greenland and Norwegian basin. The main features there are a high magnetic field gradient reaching that of the continental field and Iceland area (see Figure 10). In addition to the above pairs, occurring in the study area, there are also pairs covering mainly the continent. It means that the lithosphere may owe its origin to magmatism similar to that of Iceland, i.e. deep ferric varieties of basalts with emplacement of covering sediments exhibiting extensive distribution and similarly magnetically oriented, therefore they haven't been averaged by the field record from satellite. In this case, products of deep plume magmatism were laying both on the young oceanic lithosphere and on continental lithosphere subjected to rifting. However, magmatism of this type has not "healed'' the entire rifting zone. The ascent of deep mantle plumes to the surface was spartially a fairly rare event, nowadays basalts of plume association present an episodic event, whereas spreading association basalts are common all over the Mid-Atlantic Ridge.

[73]  The other parameters of cluster 12 are confined to its primary magmatic nature. There were high elevation caused by high rate of magmatic activity, high, as compared to that of basins, seismic activity, because of the proximity of cluster 12 areas of active superimposed events; not very high values by Love waves, low value of Bouguer anomaly, and a greater thickness of sediments because cluster 12 areas are located around the periphery of the ocean. Geodynamically, this cluster presents relict effect of paleoplumes on the Atlantic lithosphere.

6.2.3. Continental margin group.
[74]  This group includes clusters 4, 9, 13, and 15 (see Figure 12, Table 1). The interpretation of these clusters is the least conjectural.

[75]  Cluster 4. In plan this cluster (see Figure 12) is located on the continental shelf very close to the land or to the source area. The distinctive features of the cluster are: anomaly Bouguer of 56 mgal implying its continental nature, because its value of 175 mgal is the most acceptable value for of separation of the continent from the ocean. Love wave value is below zero and along with positive values by S-waves they argue for continental nature. Heat flow is close to the average. Isostatic anomaly is slightly higher because it reflects loading of sediments on the crustal and mantle substrate not fully compensated in the course of downwarping. The thickness of sediments averages 2041 m at an average water depth of 561 m. The magnetic field gradient is high implying the effect of the continental basement rocks. The geodynamic meaning of the cluster is the operation of the process responsible for isostatic geometric smoothing of the crustal block underlain by viscous substrate resulted from increase of sediment load.

[76]  Cluster 13. This cluster essentially inherits all the features of cluster 4 except value for elevation and sediments. Sediment thickness averages at maximum 6491 m at an average elevation of -1317 m. It places (see Figure 12) cluster 13 on the shelf edge and into the upper part of the continental slope. Whereas compensation downwarping in the cluster area gave rise to a great submergence of the substrate, the intensity of isostatic process being the same as that of cluster 4.

[77]  Cluster 9. This cluster inherits features of cluster 13 and is located offshore from cluster 4 and cluster 13, respectively. Main parameter values are as follows there. Bouguer anomaly is high as compared to that of clusters 4 and 13 (26l mgal), it means that sediments rest there on the peripheral oceanic crust which cold and more dense, than the continental crust. These factors do not result in down-warping reflected in isostatic anomaly. The magnetic field gradient also has features similar to those of the ocean basins. The average thickness of sediments equals 3273 m at an average elevation of -3747 m. So, this cluster in plan (see Figure 12) occupies areas of the continental rise and adjacent part of ocean basins filled in sediments.

[78]  Cluster 15. This cluster is similar in geodynamics with that of clusters 4 and 13; but based on the following reasons, is set aside as a separate type. In plan (see Figure 12) it occurs only in the northern Atlantic along the periphery of also a unique cluster 14, whose peculiar features affect those of the continental margin cluster showing a great thickness of sediments averaging 6040 m. Bouguer anomaly in cluster 15 is typical of continents and equals 57 mgal. Isostatic anomalies within the cluster area (see Figure 12) are marked by difference in direction resulting in neutral average value for the cluster as a whole. Strong positive anomaly inferred P-wave tomography extends to cluster 15 area as well. Elevation averages 542 m. The magnetic field gradient was found to have high value characteristic of the northern clusters in the Iceland plume area. In terms of geodynamics this cluster is meant to combine features characteristic of the continental margin clusters with those of the anomalous northern block as a substratum for intense sedimentation.

6.2.4. Group of superimposed events.
[79]  The group includes clusters 1 and 3 (see Figure 12, Table 1). Their interpretation seems to be the most nontrivial.

[80]  Cluster 1. This cluster is marked by low, as compared to those of basins, Bouguer anomaly, maximal isostatic anomaly for the Atlantic, and high seismic moment. The remaining parameters have similar average values. In plan (see Figure 12) this cluster presents areas superposed on the basin and flank zones of the Mid-Atlantic Ridge. Locally, these areas form separate extensive groups (Walvis Ridge, Rio Grande, Cape Verde Islands, Cameroon line), but essentially always they occur as. pairs pseudosymmetrical about the Mid-Atlantic Ridge. They might well be tracks of the plume magmatic events that took place in the Mid-Atlantic Ridge area and occurring in the course of spreading on the opposite sides from the divergent zone of the ocean. Strictly speaking, these formations may be called microscopic "bull's eyes'' or be considered as traces hotspots [Courtillot et al., 2002] that functioned over a period of time under the Mid-Atlantic Ridge system, hence their asymmetry on the opposite sides of the ocean. Zones occupied by cluster 1 show up as volcanic edifices, some of them are still active. The edifices have high rate of magmatic activity that gives rise to excessively massive formations above the compensation surface within the ocean floor. This provides an explanation of extreme positive value of isostatic anomaly in the volcanic edifice area, as well as low Bouguer value as compared to background value for basins. High seismic moment may be attributed to the intense recent volcanic activity. The geodynamic meaning of this cluster is interpreted in a way similar to that of clusters 5 and 10 (and to a certain extent cluster 12) as resulted from deep plume energy release. However, in this case it is not aligned with the Mid-Atlantic Ridge system whose heat flow is slightly higher than an average for the Earth and differs in magma composition. Based on the data by Sobolev and Nikogosyan [1994], Sobolev [1997], parental melts of oceanic intraplate igneous products are resulted from the mantle uprise from a depth of about 1000 km and its melting at depths of 100-130 km, at a temperature of l400-l650o C.

[81]  Cluster 3. This cluster (see Figure 12) represents mainly a sublatitudinal structure superimposed on ocean basins and the Mid-Atlantic Ridge Zone, it differs in two features from the cluster area on which it is superposed. It has extreme maximal values of Bouguer anomaly and extreme minimal value of isostatic anomaly, such a combination of the above parameters has been recorded only in the fore-arc zones, e.g. of the Pacific Ocean, when extensive thrusting in the island arc area, when ahead of their front there are formed zones with isostatic mass deficit and high Bouguer anomaly. Noteworthy, that based on data of Silantiev [2003] igneous rocks, having no equivalents in recent magmatism of the northern Mid-Atlantic Ridge, match these anomalous sublatitudinal zones at 15o N and 25o N. According to this author, products of magmatic activity in these zones might well be formed in the presence of the subcontinental mantle substrate or due to active mixing of melting products and the lithospheric matter as exemplified by the subduction zones. Such a coincidence of geophysical and petrologic characters suggesting not a quite ordinary environment for the Atlantic is hardly to be considered casual. The authors of the present paper think possible to make a tactful statement concerning the change in the Altantic of the horizontal movement vector from sublatitudinal to sublongitudinal during the recent epoch. The GPS and VLBI data reported from the adjacent continents make the above statement more reliable.

6.3. Comparison of the Results Obtained With Main Notions About the Atlantic Ocean Geodynamics

[82]  As was mentioned in the introduction, the accumulation of geological-geophysical data on the structure of the Atlantic ocean floor made possible constant recognition of facts not easily explicable in terms of classical geodynamic model for the ocean. The model showing the interaction between the lithosphere formed in the divergent zone and main forces, used in the classical model, which are responsible for the present day tectonics of the ocean. The present paper is not aimed at the discussion of the alternative mechanisms, therefore we have proposed a version to systematize geodynamic environments of the Atlantic with account of new data available. It was done unbiased by means of digital methods and have proposed a geodynamic interpretation of the obtained clusters following the principle of "maximum likelihood'' by data interpretation. However, alternative interpretations are not ruled out. First, it should be noted that the authors have obtained a set of clusters reflecting geodynamic features of the ocean structural elements easily explicable in terms of the plate tectonic model. Their description is given in Section 6.2. and presents a small part of the results of the investigation. The authors also obtained a number of clusters that can be explained only by involving geodynamic mechanisms and by means of considerable modification of the existing mechanism. But this goes beyond the scope of this paper. We shall just emphasize the nontrivial results of cluster analysis.

2007ES000218-fig14
Figure 14
[83]  In our opinion, of most interest is the inhomogeneity of geodynamic conditions along the strike of the ridge, which in diversity of clusters recognized in the Mid-Atlantic Ridge zone is not equal to that reflected in any parameters used. The diversity of clusters is shown by alternation of "hot'' and "cold'' blocks, strongly differing in heat flow, total seismic moment, and Bouguer anomaly evidencing high rate of magmatic activity. It concerns also the thickness of the crust, formed in the rift zone, S-wave tomography characterizing a degree of the mantle partial melting, along with the occurrence of isolated plumes deep in the mantle (up to 700 km). The established zones are compatible well with the published data on variation of P-T conditions during the formation of MORB and also on a depth of magma ascent. Variations in conditions of melt formation result in the presence of two main associations of the basalts. They are [Dmitriev et al., 2006] plume-deep and highly productive - and spreading - less deep and low productive associations. Figure 14 shows their distribution in space. Dmitriev et al. [l997, l999] took into account discrete conditions of melt formation and that they were closely spaced (70-100 km) and the authors cannot state that it was a uniform process operating along the Mid-Atlantic Ridge marked by the ascent of heated up matter subjected to partial melting but having different P-T conditions along the divergent zone. Figure 14 also shows the ubiquitous distribution of the spreading association basalts, whereas those of the plume association are superimposed on spreading ones and were reported from the Mid-Atlantic Ridge only in the presence of deep isolated plume feeders, in other words there is not one but two independent superimposed processes, hence their geodynamic effect should be considered separately, and their superposition is to be calculated in areas of plume associations. However, in this case we need a new separate geodynamic model based on two mechanisms, but this goes beyond the scope of this study. This conclusion will be an assumption about the relation between the basalt associations and related geodynamic mechanisms. It could have remained unchanged during the entire period of the Atlantic opening as follows from the properties of cluster 12 (see 6.2.2.) accounted for the presence of paleoplumes evidence.

[84]  The cluster classification of the Atlantic geodynamic lithosphere resulted also in establishing of zone oblique to major structural elements (flank clusters, see 6.2.1.). These zones have been repeatedly mentioned as inferred from detailed geophysical survey in the Atlantic [Mazarovich and Sokolov, 2002]. Their presence have found confirmation in macroscopic description of the lithospheric geodynamic parameters implying fundamental cause-and-effect relationship between parameters of different scale. The north-western orientation correlates with that of many small structural elements inferred from multi-beam survey of the ocean floor and from seismic profiling of sediment deformations. Geodynamically, it means that the actual tectonosphere has a system of forces tangential to the earth's surface and not orthogonal to the Mid-Atlantic Ridge, responsible for this phenomenon. The working geodynamic model cannot provide an explanation for this phenomenon.

[85]  A combination of cluster 3 parameters (see 6.2.4.) characteristic of the fore-arc zones is considered to be nontrivial. The sublatitudinal orientation of this cluster zones really suggests the presence of the sublongitudinal movement component that is not consistent with the working geodynamic hypothesis. An agreement with exceptional geochemical data ( [Silantiev, 2003] is hardly to be accidental. Besides, in this cluster area near the fault at 15o20 prime N and east of the Mid-Atlantic Ridge a so-called convergence zone of passive parts of transform faults was recorded. This also requires the involvement of the sublongitudinal component of the lithospheric mass movement. There are two subduction zones (Puerto Rico trench and northern margin of the Scotia Sea) in the western Atlantic. The movement along them is also northward and this makes the question, whether its thrusting or subduction, of secondary importance. This cluster forms a latitudinal zone between 22o N and 28o N in the study area along the Canary-Bahama geotraverse [Mashchenkov and Pogrebitsky, l998], where there were reported, especially east of the Mid-Atlantic Ridge, deformations and oblique reflectors of sublongitudinal dip in the consolidated crust. These reflectors provide different interpretation in terms of mineralogy to tectonics. The relationship between the crustal structures and sediments allows to conclude that the age of these dislocations may vary from the recent in the Mid-Atlantic Ridge area to Mesozoic in the Canary Basin. Still it should be noted the agreement between the zones established by means of cluster analysis and characteristic features of the fore-arc zones, on the one hand, and, on the other hand, deformation of the consolidated crust along the Canary-Bahamas geotraverse exhibiting thrust features.

[86]  Anisotropy of sediments in the Atlantic basins [Mazarovich and Sokolov, 2004] represented by foldings, discernible only on sublongitudinal seismic profiles, also implies the presence of the sublongitudinal component of movement. GPS and VLBI data on the adjacent continents support to such pattern of movement. The above-listed set of factors made us possible to speak with confidence about the phenomena whose presence seems imperative in terms of classical plate tectonics theory. The authors are not going to develop in this paper the ideas about a mechanism responsible for their origin, but it should be pointed out, that the classical theory may be applied if it is granted that the crucial changes in all movement components began during the present epoch.


RJES

Citation: Sokolov, S. Yu., N. S. Sokolov, and L. V. Dmitriev (2008), Geodynamic zonation of the Atlantic Ocean lithosphere: Application of cluster analysis procedure and zoning inferred from geophysical data, Russ. J. Earth Sci., 10, ES4001, doi:10.2205/2007ES000218.

Copyright 2008 by the Russian Journal of Earth Sciences

Powered by TeXWeb (Win32, v.2.0).