Abstract. On the basis of recent data on the distribution of earthquakes in Arctic Canada, the earthquakes are shown to be grouped  in fairly distinct local zones, despite earlier notions. The zones are associated with contacts between areas of different history of tectonic development. Higher seismicity is due to the net action of various tectonic factors: intraplate stresses generated by processes on plate boundaries, vertical tectonic movements, and load of abnormally thick sedimentary strata. The insignificant role of glacioisostatic movements is argued.

Introduction

Certain areas of higher seismicity have fairly long been known to exist in this part of the Earth, but more or less complete information on earthquakes became available beginning from the 1960s, after local seismic networks had been created. Now uncertainty of epicenter localization generally does not exceed 50 km, and the minimal energy level of earthquakes recorded without a gap is no higher than magnitude 3.5-3.6 when averaged over the region, but in certain zones drops to 2.5-3.0.

Although epicenters in Arctic Canada are on the whole rather sparse, they are concentrated in irregularly distributed zones of various shape. Now one may state that epicentral zones are not associated with anyone of modern global seismic belts and they are undoubtedly intraplate zones.            

A number of works were devoted to the tectonic nature of Arctic Canada [Avetisov, 1993; Basham et al., 1977; Grantz et aI., 1990; Wetmiller and Forsyth, 1982: and others] but the most exhaustive one is nearly twenty years old. The goal of the present work is an attempt to discuss the given problem on the basis of the most recent seismological evidence [Avetisov and Vinnik, 1993].

Structural-Tectonic Setting

Arctic Canada is a continental bridge between rock masses of Greenland and North America, and separates the Atlantic and Arctic oceans. Because of this geographic position, it is quite soundly considered a geological object whose structure cannot fail to reflect processes of formation of adjacent water areas, which defines its key role in solving the problem stated above.

The whole course of geological evolution of this region, studied in numerous works [e.g., Balkwill, 1978; Douglas et al., 1963; Kerr, 1981; Trettin, 1989] consists of two major stages, partially overlapping in time.

The main features of the present geological structure of the Canadian Arctic Archipelago (Figure 1) were developed at the first stage, referred to as a constructive one and begun in the Archean from formation of the ancient crystalline basement. The latter is represented by metamorphic and intrusive complexes of the Archean to Proterozoic age and outcrops on Baffin Island, southeastern Ellesmere Island, eastern Devon Island, Boothia Uplift in western Somereset Island and a narrow surface bud in the land part of the Peary geanticline in northernmost Ellesmere Island. Similar ancient complexes of the crystalline basement outcrop are exposed on the Greenland coast of Baffin Bay. The rest of the archipelago is occupied by structures of the so-called Inuit continental-margin mobile belt represented by deposits practically of the whole Phanerozoic. Most developed are passive Cambrian-Devonian strata in. the relatively stable Arctic platform that were folded within the Franklin geosyncline. The relief of the underlying ancient crystalline basement exhibits uplifted zones, the largest of which are the above mentioned Peary geanticline and Boothia uplift. Geological-geophysical data [Kerr, 1977, 1981; and others] show that these structures formed as a result of multiple uplifts of lithospheric blocks along a series of vertical deep faults detected in peripheral areas of the structures. There is a noteworthy wide range of uplift strikes (from submeridional to latitude), which appears to be related to characteristic features of the inner structure of the crystalline basement. The Boothia Uplift plunges northward under strata of the Inuit belt and sedimentary Sverdrup Basin and quite probably reaches the Peary geanticline [Kerr, 1981].

 

The Sverdrup Basin lies in the central zone of the northern part of the archipelago and is of triangular  shape in the cross section, opening northward toward the ocean and closing southward. It is filled up with Early Carboniferous and younger rocks with net thickness up to 13 km and connected in the southwest with the sedimentary Banks Basin via a narrow bridge.

Formation of the above basins is genetically related to the Lower Carboniferous initial phase of the second evolutional stage of the Canadian Arctic archipelago: formation of adjacent oceanic basins and their continental branches, which in the long run controlled the modern sea and land configuration of Arctic Canada [Kerr, 1981; Trettin, 1989; and others].

The first event of this stage was active faulting (rifting) directed southeastward from the Arctic basin and concurrent sagging (extension) of the lithosphere, which resulted in development of channels and basins in the northern half of the continental structure of the Inuit belt, this structure having been formed in the process of the first, constructive stage. On the whole, this process called the Boreal rifting episode [Kerr, 1977] reached only the middle of the Queen Elizabeth Islands and it is along the modern Parry channel only that it advanced further to the east, as far as Boothia Uplift. The episode died away in the latest Cretaceous.

Continuation of the second stage is related to Late Cretaceous-Miocene tectonic activity that propagated at this time out of the North Atlantic basin [Kerr, 1977; Pogrebitskiy, 1976]. It was associated with formation of rift structures of the Labrador Sea, Davis Strait and Baffin Bay (Eureka rifting), as well as straits and channels controlled by faulting in the southeastern part of the archipelago. As is evidenced by geophysical (seismic concluded) data [Grantz et al., 1990; Shih et al., 1988; and others], Baffin Bay is now a large sedimentary basin, with the thickness of its deposits amounting to 6 or more kilometers. Its deep-water part is underlain by the oceanic crust about 10 km thick. Geophysical data synthesis fixes the boundary between different types of the crust at the 1000-m isobath.

Rifting was connected genetically with development of a belt of compressional structures (Eureka orogen) at its front; the belt extends from the northwestern Greenland coast southwestward, through the Nares Strait and Ellesmere Island to Ellef-Ringnes Island, and then turns southward and reaches Lancaster Sound. This belt formed due to the superimposed action of two tectonic phenomena: horizontal motions of diverging lithospheric blocks and gaps in rift fault propagation when it encountered ancient transverse structures such as, for instance, the Franklin geosyncline fold belt or Boothia uplift [Balkwill, 1978; Kerr, 1981; and others]. In its conclusive phase, Eureka rifting overcome resistance of transverse structures and faults propagated in two directions: westward and northward, along the modern straits of Lancaster and Nares, respectively; thus they extended into the Arctic basin and the triangle-shaped subplate of the Queen Elizabeth Islands was formed. This event in fact completed formation of the region in its present appearance.

Hypocenters and Focal Mechanisms of Earthquakes

In line with everything said above, it is reasonable to analyze the epicentral field separately in areas differing substantially in histories of their tectonic development. These are Baffin Bay and Baffin Island, formed in the process of the Eureka rifting, the Queen Elizabeth Islands and their surroundings included in the zone of Boreal rifting occurrences, and the comparatively stable southwestern part of the Canadian archipelago.

Baffin Bay and Baffin Island are characterized by fairly high seismic activity (Figure 2). The strongest Arctic earthquake with magnitude 7.3 and its aftershocks with magnitudes to 6.5 were recorded here on Baffin Bay in 1933. Strong earthquakes (M = 6.0) oc­curred in 1945, 1947 and 1955. On the basis of magnetic data [Grantz et al., 1990; Jackson et al., 1977; Kristofferson and Talwani, 1977] plate-tectonic reconstructions in the Arctic relate the region under study with the former boundary between the Greenland and North American plates, so particular emphasis has always been placed here on the study of local seismicity [Grantz et al., 1990; Hashizume, 1973; Qamar, 1974; Reid and Falconer, 1982; and others]. It is in this region, though difficult to access and carry out observations, that field seismological studies were conducted using 6 portable (including 3 sea-bottom) stations [Reid and Falconer, 1982]. Notwithstanding the short duration of these observations (from a week to a month) and the considerable distances between observational points (100 to 300 km), the authors managed to reach fairly high accuracy in localizing 9 epicenters in well-known land and water areas of epicenter concentration and thus confirmed validity of the stationary seismic network data.

As is seen from Figure 2 which reflects all available information on earthquakes until 1990 inclusive, a main feature of the earthquake distribution on Baffin Bay is a fairly wide (up to 200 km) zone of epicenters, their density being clearly variable. The zone is apparently offset from the bay axis toward the eastern coast of Baffin Island. Only a few epicenters lie within the deep-water area of the bay and even this probably caused is by spread due to localization uncertainties.

The beginning of this zone can be marked by a few epicenters of fairly strong earthquakes near the bay axis at the latitude 68-69°N, south of which no epicenters are detected. With some interruptions, this zone extends northwestward toward the Baffin Island coast, becomes distinct at 71° N, and continues as far as the traverse of northern Lancaster Sound, generally following the zone of bottom depths from 500 to 1000 m. Further epicenters evidently branch off as three chains of different orientation and distinctness: The northern one runs along the northern coastal zone of the Lancaster Sound, dying out at the longitude of central Devon Island; the northwestern one extends into Jones Sound north of Devon Island; and the meridional one extends as far as 100-150 km toward Nares Strait and quickly dies out. The virtually complete aseismicity of the northern and southern continuations of Baffin Bay (Nares and Davis straits, respectively), noted long ago [Basham et al., 1977; Reid and Falconer, 1982; Wetmiller and Forsyth, 1982; and others] is completely substantiated. Thus the available data clearly indicate that the seismically active zone near the axis of Baffin Bay bears a somewhat local character, being closed and unassociated with any modern global seismicity belt.

Apart from the zone mentioned above, one may state that a zone of epicenters with magnitude up to 4.5 exists along the eastern part of the bay; the density of its epicenters is irregular and on the whole smaller than that of the first zone. Nearly all earthquakes of this eastern zone are located in areas of water with depths of 100-200 m at the most or in the deglaciated coastal zone of Greenland. However, scattered but fairly intensive epicenters are reliably recorded far off the coast line, in the zone of thick ice cover.

By and large, available numerical determinations of hypocentral depths as well as qualitative estimates based on waveforms point to shallow seismicity in the bay. For instance, the values 24 and 10 km were ob­tained from pP waves of strong earthquakes of 1976 and 1983, respectively [Grant at al., 1990]. A wider range with maximum values up to 50 km has been obtained for near earthquakes [Reid and Falconer, 1982].

Prior to 1960, only one very strong earthquake with magnitude 5.5-6.0 was recorded on Baffin Island, in 1935. A wider seismic network and field studies now have established higher seismicity in a considerable part of the northeastern coast of the island, with epicenters clearly concentrated in the Buchan Bay and Scott Inlet north of the latitude 70° and in the Home Bay south of it. The very strong 1963 earthquake with magnitude 6.0 and several events with magnitude 5.0 were recorded here beginning from 1960. Hypocentral depths when determined do not exceed 6-9 km. Most of the epicenters do not penetrate inland farther than fjord closure although the number of epicenters within the central part of the island is rather considerable. Scattered though fairly strong events are recorded outside the zone considered. In particular, a few of them are located in the northwestern most part, on Brother Peninsula. Thus the seismically active zone of Baffin Island also bears a clearly limited, local character and is not associated with global seismic belts.

The first arrival fault plane solutions have been obtained for two earthquakes in Baffin Bay and three earthquakes on the Baffin Island (Figure 3a). All authors obtained an upthrow mechanism for both earthquakes (1933 and 1976) in Baffin Bay, although marked discrepancies are obvious in orientation determinations of the compression axis, particularly for the second of them. These discrepancies are associated with different values used for the hypocentral depth. However the northwest orientation of the axis should be adopted according to a more reliable solution for the 1933 earthquake [Stein et al., 1979] and solution for the 1976 earthquake [Avetisov, 1993] where the hypocentral depth was set equal to the real value 23 km.

Normal-fault or strike-slip mechanisms were obtained for all earthquakes on Baffin Island, with the subhorizontal extension axis being fairly stable and transverse to the shoreline.

By and large, the Queen Elizabeth Islands and adjacent northern and southern areas are also characterized by an appreciable level of seismic activity. It is presently safe to peak of fairly distinct localization of seismically active zones here (Figure 2).

First of all one should note a quite distinct high-density epicenter zone extending virtually all along the Arctic coast of the Queen Elizabeth Islands, from Prince Patrick Island in the southwest to southwestern Ellesmere Island in the northeast. On the basis of then available evidence, Basham et al. [1977] established the western part of the zone only. Numerous local events were detected at station Mold-Bay here, although coordinates of epicenters were not determined. For instance, a cloud of 2000 earthquakes occurring in 30-day period from March to April, 1965 was recorded on the southwestern coast of Prince Patrick Island [Grantz et al., 1990; and others]. One may suggest that this zone would be still more narrow if the accuracy of epicenter localization were improved, with most of the epicenters being fixed by 3-4 stations. However even this accuracy leaves no doubt that, on the whole, epicenters are confined to land areas, follow the shoreline in a number of places, and virtually do not occur in water areas with depths more than 200 m.

Another sublinear, though less distinct zone can be traced from the eastern half of Melville Island and the water area northeast of it to Bathurst Island and the northwestern end of Devon Island (Grinell Peninsula) and further nearly strictly southward, through the Somerset Island and Boothia Peninsula to the continent. On the north this zone and the above-mentioned coast line of epicenters appear to merge.

Epicenters of the zone are much more scattered and irregularly distributed. The most prominent concentration of epicenters is observed in the shallow part of Byam-Martin Sound north and northeast of Melville Island. Some ten earthquakes per year with magnitude above 3-3.5 occurred here until 1972, which was believed to be typical of these areas of the Canadian archipelago. An abrupt activity outburst began in late 1972 when 8 strong earthquakes with magnitudes up to 5.6 occurred during November and December, with the total number of events (foreshocks and aftershocks included) amounting to 52. A rather high activity level persisted in the first quarter of 1973 when 38 earthquakes were recorded, including one with magnitude 4.9; then it began to drop and in 1975 reached the former level (existing prior to 1972). The latter lasted until now.

A fairly distinct local cloud of epicenters in this zone is observed at its intersection with the shallow zone of Barrow Sound (continuing Lancaster Sound) where 23 earthquakes were recorded in an area 100 x 50 km. Two of them, occurring in 1974 and 1987, had magnitudes of 1.9 and 5.2, respectively.

A high-seismicity area exists in the Arctic Ocean, northwest of the archipelago. In spite of large epicenter distances, 57 earthquakes were recorded here, of  which 13 had magnitudes 4.0 to 4.9 and two above 5.0. The cloud of epicenters in its densest part is of sublinear shape and NE strike, lies in a water area with bottom depths ranging from 200 to 1000 m, and traces the zone of abyssal-shelf transition. Epicenters are more sparse in its northern part, which might be related to decrease accuracy of localization. Noteworthy is an obvious concentration of epicenters in the region bounded by 79-80°N and 107-108°W, where 14 earthquakes were recorded in a very limited area; six of them had magnitudes above 4.0 and two were above 5.0. It is interesting that here; just as in Byam-Martin Sound, a pronounced outburst of seismic activity (7 of 14 events, including one with magnitude 5.0) occurred in the second half of 1972.

Three focal mechanisms [Dziewonski et al., 1981] were derived by the method of centroid moment tensor for this part of the archipelago. Also six mechanisms were obtained from first arrivals for four earthquakes in the Byam-Martin Sound [Grantz et al., 1990], one in Lancaster Sound and one on the western coast of Boothia Peninsula [Avetisov, 1993]. Solutions by both methods are available for the latter two earthquakes (Figures 3a and 3b).

A main feature of nearly all solutions is the predominant role of strike-slip motions, with upthrow component present too. Noteworthy is the general uniformity of the azimuths of stress axes: NNW extension and ENE compression. Taking into account the considerable separation of observational points, one may speak of a regional character of the stress field.

The southwestern part of the archipelago, including Victoria and Banks islands and adjacent sounds, is distinguished by very low seismicity, as is evident from mapping of epicenters. This is particularly conspicuous against the fairly high seismic activity of surrounding areas, especially eastern and northern ones. Throughout the whole history of observations, no more than 10 earthquakes with magnitudes 3.0-3.2 or less were recorded in this area, about 600,000 square kilometers m size.

Tectonic Nature of Earthquakes

Characteristics of the spatial distribution of earthquake sources and parameters of their focal mechanisms obviously depend both upon stress-generating tectonic processes and, to a large extent, elastic properties of real geological materials affected by them. This leads to the conclusion that earthquakes occur most frequently in preexisting weakened zones of the lithosphere, and a forthcoming structural-tectonic setting inherits much from an older one. This conclusion has both theoretical and applied significance, (for seismic risk assessment specifically).

Comparison of the map of epicenters with a geological diagram (Figures 1 and 2) leads to two indisputable, in our opinion, conclusions.

Firstly the comparison obviously points to confinement of seismically active zones to contacts of blocks differing substantially in their geological development. There is no doubt that such contacts, along with disjunctive dislocations, are weakened lithospheric zones. We should note that epicenters clearly trace the northwestern near-coastal boundary of the sedimentary Sverdrup Basin virtually along its entire extent. Epicenters vanish only at its northeastern end. The second seismically active zone undoubtedly correlates with the linear structure of the crystalline basement, the Boothia Uplift, extending far to the north and, as suggested by geological evidence, with its continuation beneath depositional strata of the Franklin geosyncline and Sverdrup Basin. The zone is submeridional and extends away from the continent. One may say that the peculiar features of the distribution of earthquakes afforded independent information on the relief of the buried part of the basement. Two local clouds of epicenters of the zone located in Barrow and Byam-Martin sounds characteristically coincide with its intersections with southern boundaries of the Franklin geosyncline and Sverdrup Basin, respectively. It is also interesting that an oceanic area of markedly higher seismicity lies north of and in line with this linear zone of epicenters. Since adjacent parts of the continental slope are aseismic, the higher seismicity of this very area might be explained by the suggestion that the buried structure of the Boothia Uplift intersects here the continental slope, which is a contact zone between blocks of continental and oceanic crust.

In this context, the seismically active zones in Baffin Bay and adjacent areas are no exception, the zones tracing boundaries and contacts of the Canadian and Greenland crystalline shields as well as the junction of continental and oceanic crustal blocks. Weakened zones originating back at the spreading stage of basin development appear to play an appreciable role in the oceanic part of the bay. In particular, we agree with the suggestion of Sobczak and Halpenny [1990] that the very strong. earthquake of 1933 and its vigorous aftershocks occurred in the zone of an ancient transform fault. Chains of epicenters in Lancaster and Jones sounds appear to be as 30ciated with similar weakened zones.

The second conclusion lies in the fact that by no means all of established weakened lithospheric zones are 5eismic while nearly all epicenters concentrate around that or other of the zones. For instance, the eastern junction of Sverdrup Basin and adjacent zone of the Franklin geosyncline, contact zones of the Banks Basin, the Canadian crystalline shield west of the Boothia Uplift  and its narrow projection in the Victoria Island area (Minto Uplift), and others are aseismic. Sharp discontinuities and gaps in the distribution of epicenters often occur, without any apparent reasons, within seismic areas, i.e., within boundaries of a seemingly uniform lithospheric block: at the eastern coast of Baffin Island (70°N) between two intense epicenter clouds as well as in northern and southern parts of the coast, in the central Boothia Uplift, and others. This second conclusion raises the question of the presence of additional factors that affect seismicity level variations. These are undoubtedly still unknown details of the deep structure of the lithosphere as well as characteristic features of stress source (s).

It is generally agreed that the higher seismicity of Arctic Canada is connected with three major factors: glacial-isostatic motions, sagging of the lithosphere due to loading of thick sedimentary strata and release of stresses generated on boundaries of lithospheric plates. With plate boundaries thousands of kilometers away, the effect of the last-mentioned factor should be uniform on the whole and appear as a kind of background. The lithosphere response to removal of a glacial load is a regional factor since the last glaciation involved virtually the whole region under study, although its role is highly variable with position and depends upon ice thickness and time of deglaciation. An finally the third factor undoubtedly bears a local character and its role can be significant only in Baffin Bay and the Arctic Ocean. Various stress configurations superimposed on a heterogeneous and inhomogeneous structure, often poorly studied, just give rise to the observed intricate pattern of seismicity in the given region, which greatly aggravates the evaluation of the contribution of each of the factors in a specific area and precludes discussion of any regional correlations.

If we recognize current notions of the tectonic nature of Arctic Canada earthquakes as being, generally valid, the available geological-geophysical and glaciological data allow us to state some additional considerations concerning the role of the above factors separately.

We cannot agree with Basham et al. [1977], Stein et al. [1979] and Wetmiller and Forsyth  [1982], who state that seismicity all throughout Northern Canada correlates well with areas of maximal differentiation of glacioisostatic movements.

The presence of glacioisostatic movements in Northern Canada, as in Fennoscandia, incidentally, was established back at the end of last century and, on the basis of this historical fact, the higher seismicity of the both regions fixed instrumentally somewhat later was related by all investigators to just this mechanism. However, no special attention was paid to the evident fact that, in each of the regions, higher seismicity zones are distributed quite differently with respect to areas of various degree of glaciation and hence of various intensity and differentiation of glacioisostatic movements. It is in Fennoscandia that the zone of maximal glaciation and maximal glacioisostatic movements in the central Baltic Shield (northern Gulf of Bothnia) coincides with the maximum seismicity zone, while both parameters diminish toward the shield periphery, with some fluctuations being due to the block structure of the lithosphere. In Arctic Canada, on the other hand, zones of maximum glaciation and related movements in the Foxe Basin and northeastern Ellesmere Island (Figure 2) as well as the most pronounced of such zones in Hudson Bay [Andrews, 1970; Fulton, 1989; Wetmiller and Forsyth, 1982] are virtually aseismic, and higher seismicity is associated rather with the periphery of the glacial shield. With the tectonic position of the regions being similar, this circumstance alone leads to the suggestion that, in one or both of these regions, glacioisostasy does not play a decisive role in formation of higher seismicity.

As was shown by data of numerous studies [Andrews, 1970; Fulton, 1989; Grachev and Dolukhanov, 1970; Morner, 1978; and others], maximum uplift rates are reached on the average as late as 1000 years after complete deglaciation, and then decrease exponentially, with an abrupt drop in the rates occurring as early as 1000-1500 already years after the maximum. At least the one third of the full uplift is accomplished during the 1000 years. The full isostatic adjustment is commonly estimated to occur not later than in 10,000, and Morner [1978] reasonably believes that the glacioisostatic factor finally ceased to affect uplift of the Baltic Shield as soon as after 900 years. In any case, taking into account the fact of general deglaciation of Fennoscandia completed 9000-9500 years ago at the latest, one may agree with the currently adopted opinion that glacioisostatic movements in this region at best do not play an appreciable role in generating seismogenic stresses. In the case of Arctic Canada, arguments concerning the parameters considered appear to be less convincing since general deglaciation here occurred later, namely 6000-6500 years ago. However, these numbers are valid for centers of maximal glaciation, and the seismic northern part of the Canadian Arctic archipelago, with the exclusion of the Ellesmere and Axel Heiberg islands, was not at all subject to the last glaciation and underwent no vertical movements in this period [Fulton, 1989]. Apart from presently seismic Baffin Island and aseismic Ellesmire Island, all the other, areas of the archipelago, both seismic and aseismic, were free of ice a by 9000-9500 years ago. Thus, evidently the lack of correlation between the distribution of seismicity and deglaciation time allows us to claim that, in Arctic Canada also, glaciois isostatic movements on the whole do not play an appreciable  in formation of the present seismogenic stress field. It appears that this conclusion should not be extend to the western coast of Greenland, which only now is being enlisted in the deglaciation process.

As to assessment of the role of glacioisostasy at present, of interest is the map of present rates of uplift [Andrews, 1970] according to which zones of highest rates (0.8-1.3 centimeters per year at most) are not at all in the same position they occupied 6000-8000 years ago, when glacioisostatic movements prevailed. First of all these zones include the seismic northeastern mountainous coast of Baffin Island, where coastal beaches are established to subside and, less obviously, the seismic northern part of the Canadian Archipelago; outside the region considered here, one such zone is the seismic southeastern coast of Labrador.

Analysis of focal mechanisms provides additional information that confirms on the whole our conclusion on the role of glacioisostatic movements in buildup of the present seismicity field. Using data for 29 strong earthquakes throughout the Earth, Talwani and Rajendran [1991] distinguished two main types of intraplate earthquakes differing in focal mechanisms: type "A" represented by strike-slip movements, with directions of stress axes being stable for large regions and caused by tectonic motions on plate boundaries, and type "B" represented by focal mechanisms that result from superposition of local disturbing stresses on a background field of type "A." As is mentioned above, it is strike-slip mechanisms that have been established for the majority of the region under consideration, with their parameters being fairly stable and compression axis nearly orthogonal to the North American plate boundary in the Eurasian subbasin of the Arctic Ocean (Figure 3). This fact indicates that the background stress field here is not much distorted by local perturbing objects and, although raising the general level of seismicity, these disturbances are not intensive enough to change the type of focal mechanism. In our opinion, this is also corroborated by relatively low intensity of earthquakes, the strongest of which has a magnitude of 5.5 or smaller, as well as by a comparatively steady gravity field [Sobczak and Halpenny, 1990] in areas of "A"-type earthquakes, which is undoubtedly indicative of a predominant role of horizontal movements compared to vertical ones. Logically, one should suggest that, without other disturbing factors, the background stress itself is insufficient for earthquake occurrence, which explains the existence of aseismic areas.

Expected exceptions are earthquakes of Baffin Island and Baffin Bay, where fault plane solutions are indicative of uplift of the shield edge and subsidence of the bay bottom, which agrees well with the data of Andrews [1970] and a map of isostatic anomalies [Sobczak and Halpenny, 1990]. This all suggests that land uplift apparently is due to epeirogenic movements of deep origin rather than to glacioisostatic ones. The same mechanism underlies the bay bottom subsidence, which is well fixed by a positive isostatic anomaly on the gravity intensity map. The existence of the mentioned movements itself is favorable for earthquake occurrence and, in the long run, a decisive factor is quantitative relationships between value, of accumulated stresses and strength characteristics of the lithosphere. Just as regional tectonic movements superimposed on a background intraplate stress field disturb it, factors of lower order are able to attenuate or enhance the effect of these movements, which ultimately can and does give rise to very strong earthquakes with magnitudes of 6.0 and higher. Within the boundaries of Baffin Bay, such factors, apart from still unknown deep-seated inhomogeneities mentioned above, can be represented by areas of abnormally thick sedimentary cover, which admit locally enhanced amplitudes and differentiation of lithosphere movements that increase the probability of earthquake occurrence. This factor also is likely to play an essential role in the seismic ally active Arctic Ocean zone adjacent to the Canadian Arctic archipelago. We believe that the presence of an abrupt gravity step there points to a predominant role of vertical movements.

In general, we should note that examination of results of seismological investigations in Arctic Canada suggests that seismologists underestimate vertical tectonic (nonglacioisostatic) movements as a factor favoring earthquake occurrence, although the geological­geophysical literature affords numerous indications of their existence throughout the geological history of the region. Such evidence includes the above mentioned map of present tectonic movements [Andrews, 1970] as well as data on multiple vertical movements of the Boothia Uplift and Piry geoanticline along a series of boundary faults and on three phases of jump-wise subsidence of the Sverdrup Basin [Sweeney, 1977]. On the basis of the obvious mismatch (as high as 100-110 m) in the position of equally aged shorelines, a conclusion has been drawn on the existence of differentiated vertical tectonic movements between the Prince of Wales Island and Boothia Peninsula and Somerset Island [Fulton, 1989].

We have compared yearly plots of energy released by earthquakes in various zones of Arctic Canada both among themselves and with similar data on mid-ocean ridge zones in the Norway-Greenland basin and Eurasian subbasin of the Arctic Ocean. Results of the comparison have confirmed that none of the above mentioned seismogenic factors alone can give rise to high-seismicity zones. Comparison with outer zones was made for earthquakes with magnitudes of 5.0 and higher because it is these earthquakes that are representative for the whole period of instrumental observations (since 1909). The comparison did not reveal any noticeable correlation between seismicity of above mentioned interplate zones and intra plate seismicity of Arctic Canada both as a whole and [or its separate areas. The cross correlation coefficient does not exceed 0.2-0.3 at various time shifts. It is noteworthy that we have obtained the same result for Beaufort Sea earthquakes.

Seismically active zones within Arctic Canada were compared in the period 1960-1990 for which earthquakes with magnitudes of 4.0 and higher proved to be representative. Here too, practically no time correlation was found, even for such close zones as, for example, Baffin Island and Baffin Bay. The only exception is the above-mentioned coincidence of seismic activity outbursts of 1972 in the Byam-Martin Bay and northwest of the archipelago. Moreover, we should note the evident general lowering of seismic activity in Arctic Canada that began sometime after 1975 and persisted at least until 1990.

 

Conclusion

The present-day seismological evidence indicates the existence of seismically active zones localized fairly well in Arctic Canada and adjacent areas. All of them trace contacts between blocks of different tectonic development and thus corroborate the fact of earthquake association with weakened lithospheric zones. On the other hand, a number of similar contact zones are aseismic.

In general, higher seismicity is due to the net action of several tectonic factors, with regional ones represented by background intraplate stresses generated by tectonic processes on boundaries of lithospheric plates. Other factors include vertical tectonic (nonglacioisostatic) movements and, in water areas, local effects of higher thicknesses of sedimentary cover. Analysis of glaciological and geological-geophysical data reveals the insignificance of the contribution of glacioisostatic movements to tectonics and seismicity of Arctic Canada and adjacent areas, with the possible exception of the western coast of Greenland.

Areas of predominant background stresses are characterized by strike-slip movements, a fairly steady gravity field and a relatively low seismicity level. Only single and weak earthquakes are possible in zones where only background stresses are active. Areas of the combined effect of several factors are characterized by focal mechanisms of different types, a highly variable gravity field, and increased intensity of earthquakes. The strongest earthquakes occur in weakened lithospheric zones at cophased (resonance-type) superposition of all seismogenic factors.

None of the above factors taken alone , background stresses included) can give rise to high seismic activity. This is confirmed indirectly by the lack of correlation between temporal characteristics of seismicity in seismically active intraplate zones of Arctic Canada and interplate ones of the Norway-Greenland basin and Eurasian subbasin of the Arctic Ocean.

Acknowledgments. In conclusion, the author would like to express his thanks to collaborators of the All-Union Research Institute of Oceangeology: RAS Corresponding Member Y. Y. Pogrebitskiy, Y. Y. Musatov, V. V. Yerba, and B. G. Lopatin  for  information provided and necessary consultations a during preparation of this work.

References

Andrews, J. T., Present and postglacial rates of uplift for glaciated northern and eastern North America derived from postglacial uplift curves, Can. J. Earth Sci., 7, 2, 703-715,1970.

Avetisov, G. P., Geodynamics of seismically active zones in the Arctic region, Otechestvennaya Geologiya, 10, 52-62, 1993.

Avetisov, G. P., and A. A. Yinnik, The bank of Arctic seismological data, Abstract, in Problems of Development of Geoengineering, Information Technology and Geoecology, 5th Interdepartment Conference on Recent Progress in Marine Geology, Izd-vo YNIIOkeanologii, St Petersburg, pp. 71-72,1993.

Balkwill, H. R., Evolution of Sverdrup basin, Arctic Canada, Am. Assoc. Petrol. Geol. Bull., 62, 1004-1028,1978.

Basham, P. W., D. A. Forsyth, and R. J. Wetmiller, The seismicity of Northern Canada, Canad. J. Earth Sci., 14, 1646-1667, 1977.

Douglas, R. J. W., et al., Geology and petroleum potentialities of Northern Canada, Geol. Surv. Can., 63-31, 1963.

Dziewonski, A. M., T.-A. Chou, and J. H. Woodhouse, Determination of earthquake source parameters from waveform data of global and regional seismicity, J. Geophys. Res., 86, 2825-2852 1981.

Fulton, R. J. (Ed.), Quaternary Geology of Canada and Greenland, 839 pp., Geology of Canada, 1,1989.

Grachev, A. F., and P. M. Dolukhanov, Postglacial uplift of the crust in Canada and Fennoscandia as derived from radiocarbon datings, Vilnus, Baltica, 4, 297-312, 1970.

Grantz, A., L. Johnson, and J. F. Sweeney (Eds.), The Arctic Ocean Region, The Geology of North America, vol. 1,644 pp., 1990.

Hashizume, M., Two earthquakes on Baffin Island and their tectonic implications, J. Geophys. Res., 78, 26, 6069-6081,1973.

Jackson, H. R., C. E. Keen, and D. L. Barrett, Geophysical studies on the eastern continental margin of Baffin bay and in Lancaster Sound, Can. J. Earth Sci., 14,9, 1991-2001,1977.

Kerr, J. Wm., Cornwallis fold belt and the mechanism of basement uplift, Canad. J. Earth Sci., 14,6, 1374-1401, 1977.

Ken, J. Wm., Evolution of Canadian Arctic islands: A transition between the Atlantic and Arctic oceans, in The Ocean Basins and Margins, edited by A. E. M. Narin, M. Jr. Churkin, and F. G. Stehli, vol. 5,105-199, 1981.

Kristofferson, Y., and M. Talwani, Extinct triple junction south of Greenland and the Tertiary motion of Greenland relative to North America, Geol. Soc. Am. Bull., 88, 1037-1049,1977.

Morner, N. -A., Earth movements in Sweden, 20,000 BP to 20,000 AP, Geologiska Foreningens i Stockholm Forhandlingar, lOO, Part 3, 279-285, 1978.

Pogrebitskiy, Y. Y., The geodynamical system of the Arctic Ocean and its structural evolution, Sovetskaya Geolgiya, 12,5-22, 1976.

Qamar, A., Seismicity of the Baffin Bay region, Bull. Seismol. Soc. Am., 64, 87-98,1974.

Reid, 1., and R. K. H. Falconer, A seismicity study in northern Baffin Bay, Can. J. Earth Sci., 19, 7, 1518-1531, 1982.

Shih, K. G., et al., Crustal thickness, seismicity and stress orientations of the continental margin of Eastern Canada, Map 1710A, scale 1:5,000,000, Geological Survey of Canada, 1988.

Sobczak, L. W., and J. F. Halpenny, Isostatic and enhanced isostatic gravity anomaly maps of the Arctic, Geol. Surt:. Can., Paper 89-16,1990.

Stein, S., et aI., Earthquakes along the passive margin of eastern Canada, Geophys. Res. Lett., 6, 537-540, 1979.

Sweeney, J. F., Subsidence of the Sverdrup basin, Canadian Arctic islands, Bull. Geol. Soc. Am. 88, 1,41-48,1977.

Trettin, H. P., The Arctic islands, in The geology of North America: An overview, 349-370, V. A., 1989.

Talwani, P., and K. Rajendran, Some seismological and geometric features of intraplate earthquakes, Tectonophysics, 186, 1/2, 19-41, 1991.

Wetmiller, R. J., and D. A. Forsyth, Review of seismicity and other geophysical data near Nares strait, Meddelelser om Gronland, Geoscience, 8,261-274,1982.

  

(Received May 11, 1994.)