Postglacial tectonics and paleoseismicity of the crystalline Fennoscandian Shield have been dealt with in numerous publications [Strelkov, 1973; Mörner, 1981, 1985, 2003, 2004, 2009, 2011, 2013а, 2013b, 2017; Mörner, Sjoberg, 2018; Mörner, Sun, 2013; Kuivamäki et al., 1998; Olesen et al., 2004; Lagerbäck, Sundh, 2008; Kukkonen et al., 2010; Rodkin et al., 2012; Shvarev, Rodkin, 2018; Nikolaeva et al., 2016a, 2016b, 2018, 2019, 2021, 2023; Nikolaeva, 2019, 2022; Shvarev, Rodkin, 2017, 2018; Pole­shchuk, Shvarev, 2018; Gorbatov et al., 2017, 2020; Shvarev et al., 2018, 2022; Morey et al., 2013; Evzerov et al., 2010, 2014; Baranskaya et al., 2019; Gorbatov, Sorokin, 2018; Kosevich, Romanovskaya, 2014; Shelekhova, Lavrova, 2019; Shelekhova et al., 2022, 2023; Zaretskaya et al., 2022].

Paleoseismic dislocations in Karelia were repeatedly mentioned by A.D. Lukashov [Lukashov, 1976, 1987, 1993, 1994, 1995, 2002, 2004; Lukashov, Belashev, 2002; Zhu­ravlev et al., 1988] whose studies were conducted to re­veal more intensive seismic processes that those presently known and to obtain more evidence for seismic events.

The study of paleoseismicity in Karelia is currently being continued by many researchers [Zykov, 1997; Avenarius et al., 2005; Verzilin, Bobkov, 2009; Marakhanov, Romanenko, 2014; Nikonov, 2007, 2015; Nikonov, Shvarev, 2013, 2015; Nikonov et al., 2017; Gorbatov et al., 2017; Nikolaeva, 2019; Shvarev et al., 2022].

Like the entire Fennoscandian Shield, Karelia is a low-­seismicity region with intensities (I) 2–3 MSK-64 [Luka­shov, 2004]. A higher intensity of seismic processes and their activation in the Holocene could be due to specific features of the geodynamic regime of the Fennoscandian Shield [Lukashov, 2004; Shelekhova, Lavrova, 2019]. Pa­leoearthquakes in Karelia reached 7–8 on the MSK-64 scale [Lukashov, 1995, 2004, p. 186].

At the same time, the problems of postglacial tectonics and paleoseismicity remain insufficiently studied, especially in our study area. No paleoseismodisocations have been found there so far. The problem still lies in determining the age, intensity (I), and magnitude (M) of paleoearthquakes.

It should be noted that the last glacier disappeared from the study area about 14.000 calendar years ago (cal. ya) [Demidov, Lavrova, 2001; Subetto, 2022, p. 182]. However, the blocks of dead ice, left behind by the retreating glacier, took a long time to melt. Therefore, the processes of reservoir development and vegetation colonization therein were going much slower than those of deglaciation. This is evidenced by the Vodlozero ice-dividing upland 20 km north of the area, within an elevated block of the crystalline basement, on which the dead ice melting continued till the Preboreal.

This paper presents the results of the study on local postglacial rock dislocations and shifts associated with the occurrence of strong seismic events in the Kubovo area, Pudozh District, Republic of Karelia.

The aim of the study was to describe paleoseismic events, to identify the features of dislocations, which can be interpreted as evidence for strong seismic effects, and to try to estimate the intensity and age of paleoseismic dislocations.

Preliminary studies, which undoubtedly need to be continued, have shown that among the local seismic deformations there can be distinguished seismotectonic, seismogravitational, and shaking-induced deformations. Among the seismotectonic deformations, there are normal-fault scarps, extension fractures, shifts, seismotectonic fractures with linear zones of rock displacement. Among the seismo­gravitational deformations there are seismogravitational rock falls, rock, slides, and slump blocks along the scarps. Shaking-induced deformations are represented by destruc­tion zones. Overthrust deformations form directed seismo­colluvial ejecta.

The 2023 field studies in the Pudozh District (Kubovo village) revealed three local paleoseismodisocations in the junction zone of major neotectonic structures, the Fenno­scandian Shield and the Russian Plate (Fig. 1).

Moreover, they are located within the so-called Polkanov marginal flexure, which is the main element of the tectonic framework of the Fennoscandian Shield junction zone with the Russian Plate and one of the most active tectonic structures in Karelia. The geoflexure controls the location of the Late Cenozoic ice sheet and defines the boundary of the mega-arch uplift of Fennoscandia of modern and all previous epochs of glacioisostasy.The modern epoch is associated with catastrophic phenomena, such as destructive earthquakes and flood zones along the coasts of Holland and Denmark [Sviridenko, 2008]. The position of the axial zone of the Polkanov large radial flexure is shown in Fig. 1, a (after [Zykov, Poleshchuk, 2016; Kolodyazhny et al., 2020].

Geologically, the study area is a part of the Late Archean (2680–2670 Ma) Kubovo porphyric granite massif with fis­sure intrusions (Fig. 1, c). The massif extends NNE 10–60° for 20–25 km. The western contact of the massif is not exposed. Its eastern contact with host tonalites and migmatite-plagiogranites was studied in a series of bedrock exposures to the east of the study area and to the south of Vodla village. Porphyric granites display a persistent appearance, a massive texture and a pink to greyish-pink color with dark-smoky quartz zones. The granites are medium-grained (3–5 mm), with porphyraceous K-feldspar grains measuring 1.0–1.5 cm [Kostin, 1989].

The Late Archean crystalline basement is almost everywhere overlain by a loose layer of the Quaternary glacial deposits. The sedimentary cover is dominated by the Last Glaciation moraine consisting of heavy brownish to dark-grey sandy loam, loam and sometimes clay. The average moraine thickness varies from 25–30 to 100 m. Crystal­line rocks are exposed locally in the Vodla River valley and less frequently visible on the flat glacial plain, where the there is a thinner (5–15 m) layer of moraine. The loose sediments at Kubovo site, are irregularly distributed, filling depressions (irregularities) in the peneplanated crystalline basement. Rhythmically layered clayey-silty deposits occur along the Vodla River. The river valley is filled with alluvium, which consists in the rapids of boulders and pebbles with a small amount of well-washed inequigranular sand infill. The flood plain terraces are partly made up of 1–2 m thick sand and silt strata. The revealed topographic changes are confined to bedrock exposures along the fault scarps. The traces of paleoseismic dislocations were studied at local sites marked by stars numbered 1, 2, 3 in Fig. 1, b.

Paleoseismogeological methods were used to determine pleistoseist zones. The term "paleoseismic deformations" [Nikonov, 1995] was used to designate the combination of seismic effects on the relief, including seismotectonic, seismogravitational and gravitational–seismotectonic disloca­tions [Solonenko, 1972, 1973], as well as shaking-induced, seismohydrodynamic, seismodynamic, and reverse faulting (thrust) deformations [Nikonov, 1995]. The list of seismic deformations of the Kubovo massif includes seismic ruptures with the linear fracture zones in rocks (seismotectonic type), seismic collapses (seismogravitational type), areal fracture zones (shaking-induced deformations) and earth­quake-related colluvial wedges (thrust deformations).

An estimation of the true spatial parameters of paleo­seismodisocations was performed using a program for sat­ellite image processing and spatial analysis, which includes a free and open source geographic information system Quantum GIS (QGIS 3.42).

Fig. 2 shows the location, boundaries, and areas of the study objects.

The latest lineaments were described by analyzing remote data (space and aerial photographs, as well as large-scale topographic maps). Geological-structural studies included on-site observation, assessment of bedrock de­formations, and analysis of massif fracturing. Nowadays, the estimation of quantitative parameters of faults, as well as age determination for dislocations, are performed during trenching study [McCalpin, 2009]. However, this method cannot be fully utilized to study dislocations at these sites, because their cover is either softly thin or absent.

Another method is sampling lake bottom sediments in the vicinity of paleoseismic dislocations. The disturbed sedi­ments are revealed and dated by radiocarbon analysis. The assessment of general paleogeographic conditions of the study area can be made using micropaleontological methods (pollen and spore and diatom analyses). Unfortunately, there is no suitable body of water near the paleoseismic dislocation. However, we took samples from a small lake (a blue circle in Fig. 1, b) located 16 km northwest of the study area.

The radiocarbon age date (Table 1), obtained from the contact zone of mineragenetic and organic deposits in the lake, is indicative of their accumulation in the first half of the Preboreal.

The deposits were dated in the Laboratory of Geomor­phological and Paleogeographic Studies of Polar regions and the World Ocean, SPbU. using standard methods [Ar­slanov, 1987].

Earthquake intensity (I) was estimated by the paleoseismogeological method for local bedrock disturbances in Fennoscandia [Rodkin et al., 2012] in comparison with the existing paleoseismic dislocation intensity scales [Med­vedev et al., 1975; Solonenko, 1977; Michetti et al., 2004; McCalpin, 2009; Nikolaeva, 2019].

In the context of structural geology, the study area is located in the contact zone of major neotectonic structures: the Fennoscandian Shield and the Russian Plate. It was considered active until the Proterozoic [Lukashov, 1995]. As noted in [Rodkin et al., 2012], the considerably dissected relief of Karelia is predetermined by recent block movements, in particular, the rejuvenation of active faults or of their parts in the Late Glacial and Holocene periods. As it turned out, the territory within which the studied paleoseismodislocation is located also experienced tectonic activation presumably at that time.

A map of fractured rocks of Karelia (Fig. 3) shows that the northwestern (300–340°) and northeastern (20–40–60°) lineament trends are predominant therein [Lukashov, 2004].

The study area is located in the South Karelian structural zone, which has a sublatitudinal strike as a whole [Lukashov, 1987, 1995; Shelekhova, Lavrova, 2019, p. 191]. The structural element of this zone is a deep tectonic depression with signs of neotectonic movements, buried entirely beneath the Quaternary deposits. The bottom of the depression lies 80 m below the sea level. Its Quaternary cover is 150 m thick [Ekman, 1987].

However, the structural relief pattern of the paleoseismodislocation, both in Karelia and in this structural zone, shows mainly two directions: northwest and transverse, northeast. The results of interpretation of aerial and satellite images, as well as of ground survey, show that the predominant strikes of active lineaments – scarps, large hollows, ditches and gorges – are 290°; 330–340–345°; N–S: 0–180° – submeridional; 60–70°. At the same time, the above-mentioned northwest and northeast directions are characteristic of the scarps. Gorges and hollows are more often elongated submeridonally.

The analysis of microfractured massifs in sites 1, 2 (see Fig. 2, b, c) allowed us to identify three main fracture systems with the strike azimuths: 1) 60–70°, 2) 320–340–345°, 3) 290°. These directions are in good agreement with the strikes of neolineaments (Fig. 3), and their associated seismic dislocations confirm their seismic activation in the Holocene.

The strikes of active faults, consistent with the fracture strikes of the scarps and gorges, are also responsible for their seismic activity in the Holocene. The postglacial origin of this disastrous event is indicated by the smoothed primary surface of the crystalline basement showing distinct patterns of seismic activity, e.g. the abundance of dissected, displaced, scattered and shifted blocks and seismogravitational falls.

As already noted, there were found three areas of occurrence paleoseismic events.

The main, most representative paleoseismic dislocation (see Fig. 1, b, Fig. 2, b, claster 1) is located north of the Berezovy Island at an absolute altitude of about 60 m (61°57'42.39" N; 37°13'04" E). The fault scarp (strike azi­muth NE 60°; Н=20 m), extending for more than 50 m, is in the contact zone of heterochronous granitoids: the 2680–­2670 Ma Kubovo massif and the Archean undissected tonalites, trondhjemites and granodiorites 3000–2750 Ma old (see Fig. 1, c). On the top of the scarp, there are piles of crushed rocks. These are separate blocks (20, 60 and 75 cm) moved from their original location and 5–25 cm wide open fractures, not filled by moraine deposits, which indicates their postglacial origin. Splits shaped like gaping joint-fissures disturbe glacial polishing and striation. The fractures occur in different configurations, sometimes non-­linear or tortuous, and have width 5–70 cm or more (in swells) and depth more than 4 m. Separate blocks, detached from the rock mass and displaced from 20 to 50 cm along the subhorizontal to slighly inclined surface, are present throughout the site. The displaced blocks vary in shape and size, but they all display fresh faces and edges. It looks that some of the blocks were detached from the basement rocks (Fig. 4, 5).

The measured directions of the block displacements al­lowed us to identify the main trends in displacement observed in the northwestern (320–340°) and northeastern (60°) directions. The scarp slope is steep and bears fresh traces of block detachment. The rockfall blocks were formed at the scarp base. Huge rocks, displaced and separated from the main basement, are exposed in the scarp (see Fig. 4; Fig. 5).

Besides, abundant niches, wells and cavities more than 50 cm in width, as well as extension fractures, occur on the steps and on the top of the scarp (see Fig. 4; Fig. 5). Rock fractures and extension faults, showing various trends (NE 70°; NW 320–330–345-350–355; N–S; SE 70°), were measured in different segments of paleoseismic dislocation. The morphology of the fractures, as well as their inconsistency with the main strike of the rocks (NNE 10–60°), the direction of ice movement and the freshness of the scarp walls suggest that the fractures were formed after the glacial retreat. According to the obtained dating (Table 1), this could have occurred in the Holocene no earlier than 11350±230 cal. ya.

The signs of paleoseismic activity on the right bank of the Vodla River were found at an absolute altitude of about 50 m: 61°56'06.08" N; 37°08'17.26" E (marked by a star No. 2 in Fig. 1, b; cluster 2 in Fig. 2, b; Fig. 6).

The ruptures therein are trending SE 110°. The polished crystalline rock exposures with glacial scars were originally reworked by a glacier, which confirms the postglacial origin of the ruptures. The extensional fault, up to 1.5 m in width, forming a niche, is especially noteworthy (Fig. 6, a), The extensional fault edges are sharp and rectangular, with no traces of glacial reworking, which clearly indicates its postglacial formation. The niche is overthrust by a big block, 3×4 m in size (Fig. 6, a).

The block surface is polished subdued (reworked by the glacier), but its basement has sharp straight edges. 20–30 cm wide ruptures in crystalline rocks occur behind the "cave" (61°55'55.79" N; 37°08'04.73" E) (Fig. 6, b, c).

Also noteworthy are the ruptures in the crystalline rocks near the Padun Waterfall (61°56'04.98" N; 37°08'17.22" E; see Fig. 1, в (3), Fig. 2, d (claster 3); Fig. 7, а–d). The apparent length of the ruptures is approximately 100 m; in some places, they are 40 cm wide.

It is believed that "earthquake dislocations include such phenomena bedrock disturbances with a considerable displacement of its constituent parts, fracturing, rock breaking, block movement along slightly inclined or subhorizontal surfaces, block detachments and block turnings, as well as rockfalls" [Nikolaeva, 2019, p. 446].

Paleoseismic dislocations are indicated by the presence of open fractures. Such fractures have been revealed at all the sites studied. At the same time, the observed displacement of the blocks falls largely within the fracture opening, which does not relate their formation to frost weathering (see Fig. 4, с, Fig. 6, в, с; Fig. 7, в–f).

At paleoseismodislocation 1, the dynamic effect on the basement rocks is indicated by the position of a detached rock block on the fractured top surface at the very edge of the fault (Fig. 8, a); detachment of big blocks from the rock base (Fig. 8, b); phenomena of detaching blocks from the the rock base and their displacement along the slope (see Fig. 5, b, c), formation of the scarp with a height of more than 10 m (Fig. 8, c); detachment and removal of the block from the rock base (Fig. 8, d).

The results of studies of seismodeformations in seismically active zones indicate that the well-pronounced seis­modeformations occur during earthquakes with a magnitude (M) of 6.0–6.5 or higher (after [Lukashov, 2004]). According to experts [Nikolaeva, 2019], depending on the fracturing degree of the rock massif, such seismic dislocations could be formed as a result of seismic events of I=VII–IX. According to the classification presented in INQUA Scale, the bedrock fractures tens of meters in length and tens of centimeters to a few meters in width are caused by earthquakes with a shaking intensity of at least IX–X [Michetti et al., 2004].

Fracturing and rockfalls on steep slopes may be caused by seismic shaking with an intensity of at least VI [Med­vedev et al., 1975]. Fractures (see Fig. 7, b–f) and rockfalls (see Fig. 5) are clear signs of a sufficiently high intensity of seismic events (≥VIII).

The validity of these estimates requires assessment and further investigation.

We assume that the dislocations discussed herein were caused by a one-time event, which took place after the glacial retreat and the removal of glacial load. This is sup­ported by evidence found for tectonic disturbances: fractures not filled with glacial material, fracturing of the crystalline basement, rockfalls with sharp and fresh edges of the blocks, disruption wall, block displacements etc.

The radiocarbon dating of the organogenic nearby-lake sediments indicates their accumulation in the first half of the Preboreal 11350±230 cal. ya. This suggests that the earthquake could occur no earlier than the retreat of the glacier from this area (13200–14200 cal. ya). During the ac­cumulation of organogenic sediments in the studied lake, the edge of the glacier was already located in western Karelia, where the Salpausselkä ice-marginal formations (ca. 10200 BP) had been deposited [Ekman, Iljin, 1995] at a distance of more than 1500 km from the study area. In spite of the fact that the preliminary study of the lake sedi­ment core sequences showed no visible disturbances in the stratigraphy, further micropaleontological (pollen and spore and diatom analyses) and other studies of marshy areas near local paleoseismodislocations may either support or disprove our assumption. Unfortunately, we have so far not been able to detect any soil filling numerous fractures and use it for dating. Most of the fractures are either unfilled or overgrown with mosses and lichens. Per­haps, a more thorough study will yield traces of fracture filling with loose sediments, and it will be possible to obtain more accurate dates of fracture opening.

А.D. Lukashov [Lukashov, 2004] believed that the earliest paleoseismic dislocations in Karelia occurred 9800–9500 BP. The paleoearthquake and the formation of the "Kubovo" paleoseismic dislocation in Karelia in Karelia could occur earlier than previously assumed.

Thus, the study of local topographic disturbances and their distinctive features, as well as the analysis of the geological-structural plan and the lineament network, show the signs of the late glacial and postglacial tectonic and seismic activation in the study area. Young faults at paleo­seismic dislocation sites occur in the oldest-known Late Ar­chean crystalline basement rocks (dated at 2680–2670 Ma) of the Fennoscandian Shield-Russian Plate contact zone. Based on the data obtained, it can be concluded that after the glacial retreat and the removal of glacial load in the postglacial time (no earlier than 12000 cal. ya) the study area experienced an earthquake with a seismic intensity of at least VIII on the MSK-64 scale. The age date obtained for the organic deposit in the nearby water body suggests that it could be the earliest Holocene paleoseismic event in Karelia and that the paleoseismic dislocation could be no older than 11350±230 cal. yr.

Further studies in this region are needed to more accu­rately estimate the intensity and age of paleoearthquakes and to correct a map of paleoseismic processes in Karelia. More detailed studies will reveal other types of tectonic disturbances besides those described in this article. Despite a weak seismic activity in Karelia, there is no doubt that such studies are necessary to obtain reliable data on a possible manifestation of seismic activity in the region.