Evidence for postglacial seismicity in lacustrine records in the western Kola Peninsula (north-western Russia)

Study area

Geological and geomorphological setting

Lake Imandra is located in the extreme north-western part of the European territory of Russia in the Murmansk region, and fills a depression separating the Kola Peninsula from the mainland. Two differently directed branches, submeridional, Bolshaya Imandra, and sublatitudinal, Ekostrovskaya and Babinskaya Imandra embedded in Precambrian crystalline rocks, are tectonically predetermined and controlled by the morphostructure of the region. The studied lakes are located on the western shore of the Babinskaya Imandra subbasin that is a graben-like depression bounded by faults from the north and south (Figure 1b). In the study area the Late Archean gneisses are overlain by moraine, fluvioglacial, glaciolacustrine, and lacustrine deposits, as well as by peat deposits (Niemela et al., 1993). Large esker ridges stretching northwestward across the area for more than 20 km and sandy deposits of lake terraces of different heights (h = 3.5–13 m) are a noticeable feature of the local environment (Figure 2).

OPEN IN VIEWER

Figure 2.

Location of the studied lake basins on the western shore of Lake Babinskaya Imandra. Surface model based on ArcticDEM data (Porter et al., 2018).

Fieldwork

Fieldwork included a combination of ground penetrating radar surveys for Lake 1 and sediment core sampling for all four lakes. All the basins selected for coring were relatively small and located in gently sloping terrain. Table 1 shows the main characteristics of the studied lakes, their localization, and size.

OPEN IN VIEWER

Table 1.

Basic characteristics of the Lakes 1–4.

Lake	m a.s.l.	Area, km2	Core	Coordinates		Depth, m	Core thickness, m
				N	E
1	133.4	0.385	2	67°31′58.8″	31°45′10.8″	3.2	2.1
2	136.4	0.736	1	67°32′22.1″	31°42′56.6″	5.8	1.7
3	132.4	0.053	3/1	67°30′25.0″	31°47′17.9″	1.0	1.5
			3/2	67°31′14.5″	31°46′18.2″	0.9	2.83
4	128.7	0.5	4/1	67°30′25.0″	31°47′17.9″	1.2	2.69
			4/2	67°30′19.4″	31°47′13.9″	1.5	4.90

GPR survey. Three GPR profiles from Lake 1 (NW-SE) were made (Figure 3). The lengths of the GPR profiles were 340–520 m. To carry out the research, we used the GPR OKO-2 with an antenna unit with a center frequency of 150 MHz (Logis-Geotech). We made the fieldwork in winter by towing GPR on ice and recording the distance with an odometer. We registered the echo-signal in a time window up to 800 ns. The step of the GPR traces along the profile is 8 cm. All measurements were implemented with 32-fold stacking, which, as a notable experimental work, increased the stability of the record (Fediuk et al., 2022).

OPEN IN VIEWER

Figure 3.

GPR cross-sections from Lake 1 were made according to survey profiles (a–c). Separate black box indicates discovered patterns for areas of washed sediment slump in the gyttja layer and their imaging on the radargram (1) and in envelope peak amplitude (2). Subhorizontal reflectors and high peak amplitudes characterized the boundaries of all sediment slumps. BL: breccia layer.

Initially, we carried out the processing of the GPR cross-sections from lakes. This is known, high-amplitude resonant reflections form when GPR survey doing from ice (Arcone et al., 2006). Therefore, the upper part of our received records was noisy up to times of 65 ns, and this required additional filtering of raw records. We carried processing of data out in the GeoScan32 software (Logis-Geotech). We used the main tools for processing GPR records: zero time correction; band-pass filter and mean subtraction (to remove DC offset and signal saturation); amplitude gain (to restore its own attenuation); and as a last stage picking of reflectors and velocity analysis (to delineation the sediment boundaries). As an addition to velocity analysis, we performed attribute analysis, as it improves signal parameters and facilitates interpretation. We calculated the instantaneous envelope of the reflected signal using the Hilbert transform, the amplitude peak of which was considered as the points of the highest signal reflection intensively. Next, we performed the recalculation from the time-section to the depth-section using velocity analysis and a multi-layer depth calculation. We identified four main separate layers, the velocities for which were assigned based on Moorman (2002) data. The first layer was ice at a velocity (V) of 13.4 cm·ns⁻¹, the second was a water column at V = 3.3 cm·ns⁻¹, the third was organic gyttja at V = 3.8 cm·ns⁻¹, and the base layer was glacio-fluvial deposits at V = 6.7 cm·ns⁻¹. As a result, we received GPR cross-sections with a fixed useful signal up to a depth of 12 m for studied lake. During the subsequent interpretation, we supplemented the sections with manual sampling columns. It should be noted that previously, GPR survey of the basin of Lake 1 have already been carried out, and this showed the existence of specific dislocations in sediments (Rodionov et al., 2018).

Coring. The four lake bottom sediments (Table 1) were sampling the floating catamaran during the summer periods of 2014–2016 (lakes 1 and 2) and in 2020 (lake 3 and 4) using a hand-operated piston corer 1 m length and 75 mm in diameter. Each core was noted to the depth from the surface in the lake, and the cores were taken with an overlap of 5–10 cm, which made it possible to obtain continuous sedimentary sequence. We pushed the equipment as far as we could, normally down to bedrock or glacial till.

The sediments were described in situ. Visually recognizable core attributes (color, texture, inclusions, mechanical composition) were studied to compose a lithological description of core and mark its stratum boundaries. The lake elevation was determined using topographical maps of 1:25,000 scale. Samples for ¹⁴C dating were picked from inside the selected horizon as well as from the overlying and underlying sediments. Thus we were able to estimate the age of an event or its lower and upper age limits.

GPR results

The basin of Lake 1 was asymmetric, its northern slope had an angle of up to 10°, and the southern angle was 2°. The maximum measured water depth was 3.4 m. For the lake, we identified two layers: mineral bed (sands) and organogenic layer (gyttja) (Figure 3). The gyttja thickness was 2.5–4 m. An additional intense reflector of unclear origin was found in the mineral bed, being possibly associated with the internal lithological boundary in the deposit.

Large hummocky bottom fragments up to 2.5 m high are revealed on the GPR profile (a) (Figure 3) of 340 m long in the southern parts of the bottom, along with the areas of sediment redeposition that indicate disturbances of the sedimentation process. We also established the position of the anomalous layer in the middle part of the gyttja, which lies conformably to the mineral bed of studied lake, but withal had a local gap at the level of 165 m and a depth of 2 m. According to well drilling data, as will be shown below, this is a breccia layer (BL) (Figures 4 and 5). Also the GPR profile (a), we identified an bottom fragment at elevations of 245–267 m with a thickness of up to 1.5, which differs in subhorizontal reflectors from the surrounding sediments with chaotic reflectors (Figure 3, a1 and a2). We interpret this fragment as a local landslide of liquefied soils.

OPEN IN VIEWER

Figure 4.

Photographs illustrating the breccia horizon in cores. Lake 1 cores show: fragment of wood with a length of 5 cm and diameter of 1.2 cm (core 1/1); mixed gyttja, sand and silt fragments (core 1/2); peat and plant residues contained in a sapropelic matrix (core 1/3).

OPEN IN VIEWER

Figure 5.

Lithology logs of the bottom sediments cored in small lake basins 1–4.

On the GPR profile (b) at marks 340–410 m along the horizontal and intense reflectors, we founded a slump composed liquefied sands that were partially intruded into the gyttja layer (Figure 3, b1 and b2).

On the GPR profile (c), starting from the distance mark of 150 m, we found a layer under the gyttja, which was characterized wavy-hummocky reflectors and hyperbolas and was limited to clear-cut interfaces at the top and bottom. We interpreted this pattern as a large subaquatic slump 450 m long and 0.5–4.1 m thick (Figure 3, c1 and c2). Probably, the upper part of the slump was redeposited, since we identified extended horizontal GPR interfaces to a depth of 0.5–0.8 m, which was not typical for the main body of the landslide with a chaotic GPR pattern. At the level of 195 m, a sandy slump 1.3 m thick adjoins the upper boundary of the landslide, washed into the gyttja and overlain by a breccia layer, which repeats the sedimentary facies from other GPR cross-section (Figure 3b).

We also traced the breccia layer over a larger area of Lake 1, and its average depth was 3.6 ± 1.0 m (129.8 m a.s.l.).

Lithology and chronology

The 150–490 cm long successions from basins are composed of a sedimentary strata with two lithological units, the overlying gyttja and the underlying sands and silts. The gyttja is separated by a chaotic horizon (mixture of sand, gyttja, silt, and wood fragments) a 15 cm to 85 cm thick. This layer, which we called the breccia layer (BL), was found in lakes 1–4 (Figures 4 and 5).

Lake 1. The 0.4 km-wide and 3.2 km-deep lake is elongated from the northwest to the southeast for a distance of 1.2 km. The lake bottom sediment lithology was studied through the five boreholes (Figure 5). All cores show breccia layer (BL) with a thickness of 22–49 cm, which is traced at a lateral distance from northwest to southeast for 0.6 km.

The most representative disturbances were obtained from the Core 1 (from bottom to top): 533 to 514 cm – coarse-to-medium-grained sand with sporadic gravel grains; 514–506 cm – gray layered silt with sand; 506–498 cm – gray non-layered silt; 498–486 cm – black gyttja with mineral particles (sand, silt), the silt-gyttja contact area contains diffused textures formed from the sediments of different colors. The layer dated at 9250 ± 380 cal. yr BP (Table 2); 486–430 cm – brown indistinctly layered gyttja with mineral particles and vegetative detritus. Sample from the layer gave radiocarbon data of 7350 ± 270 cal. yr BP; 430–400 cm – BL: mixed gyttja and silt fragments of various shape (rounded, angular, triangular, irregular), various color (dark brown, black, olive), and size (0.2–4.6 cm), peat, plant residues, and sand contained in a sapropelic matrix. A large fragment of wood with a length of 5 cm and diameter of 1.2 cm was found (Figure 4). The radiocarbon data of 7350 ± 270 cal. yr BP are obtained from the middle of the interval and a wood fragment from the upper part of the BL was dated 6450 ± 340 cal. yr BP; 400–388 cm – dark-brown indistinctly layered gyttja with sand and plant remains; 388–320 cm – brown indistinctly layered gyttja with plant remains.

Lake 2 with an area of about 0.7 km², has an oval shape with two streams flowing into it from the west and north. The stratigraphic column consists of the following sediments (Figure 5): 725–724 cm – gray non-layered silt with sand; 724–722 cm – gray coarse-grained sand with silt and mica; 722–716 cm – chaotic and deformed massive gyttja with sand, mica and gravel. Up the section, the amount of the mineral particles decreases (<5%). Moire textures are noted in the interval; 716–710 cm – black gyttja with plant remains, with a sand and silt (<5%); 710–683 cm – gyttja is dark brown with an olive tint, with sand; 683–642 cm – gyttja is dark brown with black concretions 1–2 mm in size; 642–576 cm – gyttja brown with black concretions. Increased content of concretion is noted in the interval of 615–576 cm.

Lake 3 is a small flow-through bogged lake with an islet in the middle. It is located 1 km south-south-eastward of Lake 2 (Figure 2). The lake is shallow: its depth is 0.9–1.2 m (Table 1). The lake sediments were studied through the two boreholes (3/1 and 3/2) with the similar composition (Figure 3). Three radiocarbon dates were obtained from Core 3/1 (Table 2) that intersected the following sediments: 250–232 cm – gray silt; 232–206 cm – black non-layered gyttja; 206–170 cm – brown gyttja with vegetative detritus, the layer boundary is eroded; 170–139 cm – a mixture of sand and gyttja with inclusions of polychromatic gyttja fragments (black, olive-gray, and brown), BL; 139–137 cm – dark-brown gyttja with plant remains; 137–100 cm – brown loose gyttja with plant remains at a depth of 137–130 cm.

Lake 4 is located 1.5 km south-eastward of Lake 3. A stream flows into the lake from the west and another stream flows from the northern part of the lake. The lake sediments were studied through the two boreholes (4/1 and 4/2). The column 4/1 (Table 2) consists of the layers: 429–423 cm – non-layered silt with sand; 423–416 cm – layered gyttja with silt and sand. The gyttja dated at 12700 ± 200 cal yr BP (Table 2); 416–339 cm – brown layered gyttja with black flat laminae and Fe-Mn round concretions of 1–7 mm size, there are amphibolite pebbles of 1.5 cm × 1.5 cm × 0.8 cm size found at a depth of 345–348 cm. This layer was accumulated 6080 ± 160 cal. yr BP; 339–332 cm – BL?: a mixture of gyttja, silt, sand, and plant remains, the layer is enriched with the mica plates; 332–160 cm – brown gyttja. The radiocarbon data of 6600 ± 150 cal. yr BP are obtained from the lower part of the interval, (332–326 cm).

OPEN IN VIEWER

Table 2.

Radiocarbon dates and calibrated ages based on sediment bulk samples for the sediment succession from Lake 2 to 4.

Lake	Depth, m	Sample ID	Laboratory 14C age (yr BP)	Calibrated age 68% range, (cal. yr BP)	Mean calibrated age 14C date (cal. yr BP ± 1σ)	Material dated
1	4.00–4.05	LU–7365	5620 ± 300	6790–6010	6450 ± 340	Wood
	4.20–4.30	LU–7671	6710 ± 120	7680–7470	7580 ± 110	Gyttja + peat + silt
	4.30–4.40	LU–7364	6490 ± 270	7670–7070	7350 ± 270	Gyttja
	4.86–4.98	LU–7363	8300 ± 290	9550–8770	9250 ± 380	Gyttja
2	6.76–6.83	LU–10792	7050 ± 80	7790–7970	7870 ± 80	Gyttja
	7.04–7.11	LU–10791	9500 ± 90	10,590–11,070	10,840 ± 170	Gyttja
	7.15–7.22	LU–10790	10,300 ± 140	12,470–11,830	12,100 ± 280	Gyttja + silt
3	1.38–1.32	LU–10089	7220 ± 140	8190–7870	8050 ± 140	Gyttja
	1.58–1.64	LU–10088	7980 ± 290	9270–8520	8890 ± 340	Gyttja
	2.32–2.26	LU–10087	9770 ± 120	11,390–10,810	11150 ± 210	Gyttja
4	3.32–3.26	LU–10086	5790 ± 130	6740–6440	6600 ± 150	Gyttja
	3.39–3.45	LU–10085	5310 ± 150	6280–5930	6080 ± 160	Gyttja
	4.23–4.16	LU–10084	10,750 ± 160	12,900–12,490	12,700 ± 200	Gyttja + silt

Calibration was performed using OxCal v.4.4.2 and calibration curve IntCal20 (Ramsey, 2008; Reimer et al., 2020). Modeled ages are obtained using the P Sequence model (Reimer et al., 2020). Radiocarbon dates from Lake 1 after (Nikolaeva et al., 2017).

According to the lithological and stratigraphical analysis the accumulation of sediments in the studied lakes began the end of the Younger Dryas and beginning of the Preboreal (12,700 ± 200–11,150 ±210 cal. BP) (Table 2). Since that time the western shore territory of Lake Babinskaya Imandra was already cleared of ice (see Chapter 2). The lakes began accumulating sands and silts of a periglacial lake and in the Preboreal and Boreal they accumulated organogenic sediments of a modern fresh-water lake, that is, gyttja.

There are no disturbances or deformations at the top and bottom of the cores (Figure 5), which indicates a relatively stable depositional environment. In the Middle Holocene there was a disruption of normal sedimentation, as evidenced by the breccia horizon in lake cores (Figures 4 and 5). The BL thickness and lithology vary depending on geometry of each lake and the local environment. The breccia texture is expressed more explicitly in cores from lakes 1 and 3. These are angular isolated fragments of polychromatic gyttja of a size up to 2.5 cm × 3.5 cm, fragments of silt, gravel, broken wood, and abundant vegetative detritus, often with a peat admixture.

The radiocarbon dates indicate the synchroneity of the BL accumulation (6400–6100 cal. yr BP) in lakes 1 and 4 (Table 2). A gyttja sample from non-damaged layer underlying the BL in Lake 1 was dated 7350 ± 270 cal. yr BP, and a wood fragment from the upper part of the BL was dated 6450 ± 340 cal. yr BP. The age of the BL formation in Lake 4 is in 6600 ± 150–6080 ±160 cal. yr BP range. We expected to obtain the similar age from Lake 3. However, the samples indicated an interval between 8050 ± 140 and 8890 ± 340 cal. yr BP which differs from the ages obtained from Lakes 1 and 4 (Table 2). This disagreement can be associated with either blurring of the BL boundaries that probably led to the miscalculation, or another more ancient event.

Detailed sedimentological core analysis correlates well with the GPR data: Two sedimentary strata (mineral and organogenic) are well distinguished, and there is an anomalous breccia horizon (BL) enclosed in gyttja (see Figures 3–5). In addition, GPR data on the lacustrine sediments revealed vertical displacements of glacio-fluvial deposits and various landslides. The intrusion of eroded sands into gyttja allows us to consider them as criteria for additional external (seismotectonic?) impacts that occurred during the accumulation of gyttja, that is, in the Holocene. Thus, both lithological and geophysical data definitely indicate a clear disturbance in the sedimentation and mass movement of lacustrine deposits.

Conceptual model of mass movement in bottom sediments of lakes

A mixture of fragments of sand, silt, peat and wood embedded in the organogenic matrix in lake cores testify to catastrophic mass movement and local landslides. The following conceptual model of their formation can be proposed (Figure 6).

OPEN IN VIEWER

Figure 6.

The scenario chart illustrating the trigger mechanism that induces lake disturbance formation. The inset shows a–b profile.

Accumulation of BL layers in the studied lakes could be associated with the catastrophic water breakthrough of the Lake 2 across the esker ridge. It should be noted that all the studied lakes are located along the esker ridges 5 km long that is oriented NW to SE (Figure 2). According to the paleogeographical reconstructions (Armand and Samsonova, 1969; Korsakova et al., 2020; Nikolaeva et al., 2015) the Lake 2 was part of a large paleo-reservoir that occupied a bigger area ca. 6000–6500 years ago than at present. Lake elevations disconnected from the Imandra basin during the period of glacioisostatic uplift of the territory was 4–5 m higher in the Middle Holocene than in present times. Lake 2 was located approximately at the level of 141–143 m a.s.l., and esker ridge at the level of 145–145 m a.s.l. (see Figure 6, inset). As a result of the breaching of esker ridge, that bounded the paleolake two from the east, water was drained from this basin. Part of the sediments was eroded, including the underwater slopes of the basin and the lake margins. Lacustrine sediments were carried by fan-shaped water streams downward the slope eroding the local environment. The direction of these “mud flows” was determined by the topography of the area: they moved from NW to SE a narrow zone between the two esker ridges. “Mud flows” sediments accumulated in lakes 1, 3 and 4 as BL layers. The source of transported material was lacustrine gyttja, mineral sediments from the shores (mainly sands and pebbles), and vegetation. In support of this model, we have noticed that the BL lithology changes with distance from the breaching point (Figure 6). Lakes 1 and 2 contain the BL of 68 cm thick which encompasses polychromatic gyttja with sand interlayers. On the contrary the cross-section of Lake 4, which is the farthest from the breaching point, represents the BL containing only a mixture of gyttja, silt, and sand without breccias and debris.

The observed mass movements in lake sediments are interpreted as potential consequences of the earthquake shaking in the Imandra neotectonic depression.

Criteria for seismogenic disturbances

The question of how seismogenic disturbances in sediments can be recognized on the basis of specific criteria has been discussed extensively in the literature (Moretti and van Loon, 2014; Owen and Moretti, 2011; Sammartini et al., 2021). The following main criteria to recognize seismogenic disturbances summarizing some studies on this same topic were proposed: (1) type of disturbances in lacustrine; (2) sediments simultaneous formation in more than three lakes; (3) lithological distribution features within the boundaries of one lake to avoid an artifact must be taken into account; (4) lack of indications for any other causal mechanisms; (5) stratigraphic position in between undisturbed layers; (6) location of sedimentary basins within the areas of tectonic or seismic activity; (7) substantial lateral extent. We will consider the breccia layers of lacustrine sediments in the following sub-sections in the context of the criteria above mentioned.

Substantial lateral extent is the best criterion for the seismic origin of the disturbances in lacustrine basins (Owen and Moretti, 2011). Figure 2 shows that the BL horizons are revealed not in a single lake, but in several lakes spaced more than 5 km apart from each other. This completely excludes the artifact. The anomalous layers (BL) are underlane and overlane by non-deformed sediments as shown in Figure 5. As a result, the studied lake deformations satisfy the 1–5 criteria. In addition, detailed sedimentological core analysis correlates well with the GPR data (Figure 3). The intrusion of eroded sands into gyttja allows us to consider them as criteria seismotectonic impacts that occurred during the accumulation of gyttja in the Holocene.

The studied reservoirs are located directly within the sources of large paleoearthquakes confined to active fault zones (Nikolaeva et al., 2018). Therefore, their bottom sediments are expected to confirm major earthquake events. Traces of past strong earthquakes (rock slides, split of the upper parts of watersheds, fractures) are revealed on the western flank of the Lake Imandra depression. The fault activity was manifested as reverse-fault–shear processes along the uplifted vertical western wall of the fault (Figure 1, Chuna fault), which occurred in several stages throughout the Late Glacial and postglacial periods: ~13.5 ka, 10.3–7.1 ka, ~2.5 ka. Another normal fault of the northeastern extension, more than 10 km long, was identified on the western flank of the Babinskaja Imandra, SW of the Upoloksha railway station (Figure 7). Segments of the sublatitudinal fault that bound Lake Babinskaya Imandra from the south were also reactivated. The horizons reflecting the Late Pleistocene and Holocene seismic events are recorded in the loose sediments of the Imandra Lake terraces (Nikolaeva, 2022). These events led to the occurrence of secondary deformations (phenomena of liquefaction, cracks, and dykes) in the sandy-silty sediments. The amplitudes of known fault displacements in loose sediments range from 0.2–0.5 to 1.8–2 m. Weak modern earthquakes also indicate the seismic activity of the territory (Godzikovskaya et al., 2010). As a result, the studied deformations in lakes satisfy several criteria known both from the literature data and from field observations.

OPEN IN VIEWER

Figure 7.

NE trending seismic dislocations in gneiss-granite on the western flank of the Babinskaja Imandra.

The earthquake, which presumably was a trigger for the formation of the deformations, was generated by the activation of ancient fault zone of the NE strike or by the activation of sublatitudinal faults bounded of Babinskaya Imandra depression. The earthquake epicenter could be located either in the tectonic zone under the bottom of Lake Imandra or in the Khibiny center, one of the most seismically active nodes of the Kola region (Godzikovskaya et al., 2010).

Age

The radiocarbon dates indicate the synchroneity of the BL accumulation in lakes 1 and 4 (Table 2), which indicates the age of the earthquake around 6400–6100 cal. yr BP. We expected to obtain the similar age from Lake 3. However, the samples indicated an interval between 8050 ± 140 and 8890 ± 340 cal. yr BP which differs from the ages obtained from Lakes 1 and 4 (Table 2). This disagreement can be associated with two aspects: (1) the boundaries of the breccia layer in the Lake 3 are very indefinite, “blurred,” and mixed, that probably led to the miscalculation due to confusing lithology, (2) or another more ancient event. Event 6400–6100 cal. yr BP is more reliable. The reasons for this conclusion are as follows. There is geological evidence of earthquakes with an age of 6000–7000 cal. yr BP in this area. For example, in the Kola region the sediments of Kovdor Lake 50 km to the west of studied area are eroded and have inclusions of “foreign” material, such as pebbles, gravel, and wood remains formed due to seismic shakes (Nikonov, 2007). The event occurred approximately ca. 6500 cal. yr BP. Deformations and lithological unconformities were discovered in the bottom sediments of Lake Tikozero on the southern shore of Lake Ekostrovskaya Imandra 10 km away from the studied lakes within the time span of 6500 and 3500 cal. yr BP (Korsakova et al., 2022). In fact, we interpret that the BL as a consequence of an earthquake happened around 6400–6100 cal. yr BP. However, it should be noted that we cannot completely exclude another event, more ancient.

Intensities and magnitudes

Noteworthy, the present day archive of seismites of the Eastern Fennoscandia is far from completion. Currently we do not have a fairly representative database on the same-aged paleoseismic deformations and cannot estimate the propagation area of seismic shakes. Thus, the so called “Imandra event” should rather be considered as a local manifestation of seismicity. We can make only a rough estimation of intensity and extent area of seismic shakings.

Liquefaction and deformation in loose sediments usually happen when M_w ⩾ 5 (Ambraseys, 1988; Obermeier et al., 2005). Nevertheless, some seismically induced deformations are known to occur in the result of the earthquakes with the shaking intensity of VI–VII and M_w = 3 (in case the earthquake is not a deep one) (Keefer, 1984; McCalpin, 2009). Regarding the small depth of hypocenters of modern earthquakes in the Eastern Fennoscandia (5–20 km) as well as the intensity on the INQUA scale (Michetti et al., 2007) we may assume that the minimum magnitude of the local Lake Imandra Event was about M_w = 4.5–5, and the shaking intensity was about V–VI.

The generalization of paleoseismic data on Finland, Sweden, Norway, and Russia shows that the high-magnitude peak and frequency of the earthquakes took place during deglaciation and the Early Holocene (Kukkonen et al., 2010; Lagerbäck and Sundh, 2008; Stewart et al., 2000). According to Ojala et al. (2019) faults with 1–2 cm shift in sediments of Lake Immeljarvi in Finland are interpreted as the marks of an earthquake of 6700 cal. yr BP, and a landslide in Lake Laukkujarvi happened ca. 7000 cal. yr BP. Based on deformations of clastic-biogenic varves in Lake Nurmijärvi, a relatively high-magnitude seismic event (M_w ≈ 5.1–6.9) occurred in southern Finland 7400 cal. yr BP, which is about 3500 years after the deglaciation of the area (Ojala et al., 2018). Scientists in the Eastern Norway (Mangerud et al., 2018) made a conclusion that the diamicton (landslide) fragments in Homme region were produced by an earthquake of the age of 6400 cal. yr BP. A number of events have been documented in Sweden. Among them are a regional earthquake with М_w = 6–7 happened in Billingen region ca. 5700 cal. yr BP, a local earthquake with М_w > 8 in Hudiksval region ca. 6100 cal. yr. BP, and regional earthquake in Umea area ca. 6680 cal. yr. BP (Mörner, 2005).

Thus, our data and the studies in recent decades have shown that the eastern Fennoscandian shield was struck by strong earthquakes not only in the Late Glacial and Early Holocene periods, but also at a later time. At the time of deglaciation up to Middle Holocene time, Fennoscandia constituted a high to super-high seismic region in magnitudes as well as in frequency (Mörner, 2005). The seismotectonic manifestations in the Middle Holocene were likely to have been less strong than during the deglaciation. Fault movement and seismicity were caused by a combination of tectonic and glacially induced isostatic stresses, with the magnitude of the glacioisostatic effect decreasing.