**3.2 Kara-Bura test site**

We have mentioned above about our investigations on special geodetic test sites. The first region is located to the south-east of the Kara-Bura pass in the upper reaches of the Kara-Kysmak river (Chatkal region, Djalal-Abad oblast). Here the Talas-Fergana Fault passes downward from the pass and cuts Late Pleistocene recessional moraines. The fault zone is clearly marked by elongated depressions, as well as by right-lateral shifts of stream beds and watersheds between them.

We constructed a digital elevation model of the site (Fig. 9). Using the model and taking into account characteristic elements of the relief we could measure the right-lateral slip along the fault. Thus, for example, a Late Pleistocene recessional moraine (Shubin et al., 1992) was shifted for 28-34 m (В1-В2, Fig. 10). This displacement began, apparently, 5910 ± 130 years ago to what testifies the absolute age of the sample SOAN-6526 selected in detrital deposits which filled the crack, formed in the stretching zone along the fault (Fig. 11). Thus, the calculated rates of displacement in the second half of Holocene on this site of the fault were 4.70 - 4.90 mm/year.

In the body of the phased moraine of the middle of late Pleistocene (Shubin et al., 1992), the stream valley (Fig. 9 and 10) was cut. Its age, hence, is more young - apparently, the end of late Pleistocene. The thalweg of the valley was displaced to the right at 34.78m (А1-А2, Fig. 9). The watershed to the east from this stream was displaced practically on the same distance (31.07 m).

Fig. 9. Digital map of the Kara-Bura test site. Detailed mapping of the Talas-Fergana Fault zone with a use of an electronic tachometer. Pit1, Pit2- pits. Contour lines in 0.5m. Segment А1 - А2 along the fault line – displacement of the dry gully on 34,78 m. Eastern watershed of the dry gully (А – В profile) is displaced on 31,07 m. Segment В1 – В2 – right-lateral displacement of an age of Late Pleistocenemoraine on 28,34 m.

Parameters of the Strong Paleoearthquakes Along the Talas-Fergana Fault, the Kyrgyz Tien Shan 43

42 Earthquake Research and Analysis – Seismology, Seismotectonic and Earthquake Geology

4. An intake of frontal parts of recent taluses by the upslope facing scarp testifies on youth and instantaneity of its formation. This phenomenon one can observe east of the Kara-Bura pass. If not, an ancient upslope facing scarp would be covered by the colluvial material, the

We have mentioned above about our investigations on special geodetic test sites. The first region is located to the south-east of the Kara-Bura pass in the upper reaches of the Kara-Kysmak river (Chatkal region, Djalal-Abad oblast). Here the Talas-Fergana Fault passes downward from the pass and cuts Late Pleistocene recessional moraines. The fault zone is clearly marked by elongated depressions, as well as by right-lateral shifts of stream beds

We constructed a digital elevation model of the site (Fig. 9). Using the model and taking into account characteristic elements of the relief we could measure the right-lateral slip along the fault. Thus, for example, a Late Pleistocene recessional moraine (Shubin et al., 1992) was shifted for 28-34 m (В1-В2, Fig. 10). This displacement began, apparently, 5910 ± 130 years ago to what testifies the absolute age of the sample SOAN-6526 selected in detrital deposits which filled the crack, formed in the stretching zone along the fault (Fig. 11). Thus, the calculated rates of displacement in the second half of Holocene on this site of the fault were

In the body of the phased moraine of the middle of late Pleistocene (Shubin et al., 1992), the stream valley (Fig. 9 and 10) was cut. Its age, hence, is more young - apparently, the end of late Pleistocene. The thalweg of the valley was displaced to the right at 34.78m (А1-А2, Fig. 9). The watershed to the east from this stream was displaced practically on the same

Fig. 9. Digital map of the Kara-Bura test site. Detailed mapping of the Talas-Fergana Fault zone with a use of an electronic tachometer. Pit1, Pit2- pits. Contour lines in 0.5m. Segment А1 - А2 along the fault line – displacement of the dry gully on 34,78 m. Eastern watershed of

the dry gully (А – В profile) is displaced on 31,07 m. Segment В1 – В2 – right-lateral

displacement of an age of Late Pleistocenemoraine on 28,34 m.

talus would gush over the scarp and will continue its movement down the slope.

**3.2 Kara-Bura test site** 

4.70 - 4.90 mm/year.

distance (31.07 m).

and watersheds between them.

Fig. 10. Photo of the Kara-Bura polygon. View north-westward. The red line is the TFF, the white dashed line is the dry channel, the arrow Pr1 is the profile line in Fig. 9, g2QIII is a recessional moraine of the middle of Late Pleistocene. g3QIII is a recessional moraine of the end of Late Pleistocene.

Fig. 11. A schematic sketch of a trench No. 2, which was prospected through graben in a transitive zone of the stretching in the zone of the Talas-Fergana fault. 1 – is characterized by poorly developed mountain soil, 2 - is characterized by loess-like loamy soil with gravel, 3 – sand and detritus deposit/sedimentation, which filled the graben, 4 – moraine deposits/ sedimentations of the end late Pleistocene, 5 – fault planes which have formed the graben.

We examined the bore pit (Fig. 12) in the flood plain of the dry rivulet - in the SW wing of the TFF. Here the tectonic dam and impounded deposits, in connection with right-lateral displacements along the NE wing of the TFF, were formed. In the buried soil formed on a moraine of the middle of late Pleistocene (Shubin et al., 1992), we took a sample with absolute age of 6100 ± 200 years (SOAN-6523). At this particular time, apparently, there was the first earthquake which has displaced for the first time the body of the moraine and fluvial deposits. Thus, the calculated rates of displacement for the last ~ 6 thousand years comprised 4.93 – 5.89 mm/year.

Fig. 12. Impounded deposits/sedimentations in bore pit 1, examined in SW wing of the TFF.

The absolute ages of the deposits formed in the bottom parts of both pits: in bore pit 1 and trench 2, give statistically the same age ~ 6 thousand years which, apparently, is age of the first earthquake in Holocene which we managed to record. However in bore pit 1 (Fig.12) is available one more absolute age determination: in the bottom part of modern soil. Age of this sample by results of the radiocarbon analysis is 4465 ± 130 years old (SOAN-6522). This is the minimum age of the second seismic event, recorded by us: in the middle Holocene. After the 1st earthquake on the examined site of the fault the tectonic dam was formed, which has led to accumulation of impounded deposits on which the soil cover, eventually, was formed.

#### Parameters of the Strong Paleoearthquakes Along the Talas-Fergana Fault, the Kyrgyz Tien Shan 45

44 Earthquake Research and Analysis – Seismology, Seismotectonic and Earthquake Geology

We examined the bore pit (Fig. 12) in the flood plain of the dry rivulet - in the SW wing of the TFF. Here the tectonic dam and impounded deposits, in connection with right-lateral displacements along the NE wing of the TFF, were formed. In the buried soil formed on a moraine of the middle of late Pleistocene (Shubin et al., 1992), we took a sample with absolute age of 6100 ± 200 years (SOAN-6523). At this particular time, apparently, there was the first earthquake which has displaced for the first time the body of the moraine and fluvial deposits. Thus, the calculated rates of displacement for the last ~ 6 thousand years

Fig. 12. Impounded deposits/sedimentations in bore pit 1, examined in SW wing of the TFF. The absolute ages of the deposits formed in the bottom parts of both pits: in bore pit 1 and trench 2, give statistically the same age ~ 6 thousand years which, apparently, is age of the first earthquake in Holocene which we managed to record. However in bore pit 1 (Fig.12) is available one more absolute age determination: in the bottom part of modern soil. Age of this sample by results of the radiocarbon analysis is 4465 ± 130 years old (SOAN-6522). This is the minimum age of the second seismic event, recorded by us: in the middle Holocene. After the 1st earthquake on the examined site of the fault the tectonic dam was formed, which has led to accumulation of impounded deposits on which the soil cover, eventually,

comprised 4.93 – 5.89 mm/year.

was formed.

However the earthquake, which took place 4.5 thousand years ago, was not the last earthquake on given site of the TFF. We have examined two pits in hanging (SW) and foot (NE) wings of the fault scarp formed along TFF (Fig. 9, 10 and 13). In them – samples were taken in the bottom part of the layer of turf, the absolute age was 405±100 years old (SOAN-6525) for a hanging wing of the scarp and 460 ± 40 years old (SOAN-6524) for the foot wing. These two dates indicate in the occurrence of the seismic event 400-500 years ago. These data prove to be true also by the age determination SOAN-6527 (480±35 years old) of the modern soil received from the bottom part in trench 2 (Fig. 11).

Fig. 13. The late Holocenic seismoshoulder/seismoledge along the TFF (it is shown by a red faltering line), which is breaking the flood plan of the unnamed rivulet/say. The SW wing is raised. In both wings pits on depth of 15 cm were selected samples in the bottom part of the turf/peat layer for definition of their absolute age.

Summarizing data on the Kara-Bura test site, we received (calculated) the following rates of horizontal tectonic displacement since the middle Holocene: 4.70-5.90 mm/year. During this time in the named area along the Talas-Fergana fault minimum three strong earthquakes took place: about 6000 years, 4500 years and 400-500 years ago. The data of absolute age, determined on the Kara-Bura test site regarding the latter earthquake, coincide with the archeo-seismologic data on the destroyed caravansary located in the middle part of the Kara-Bura river valley (Korjenkov et al., 2009).

#### **3.3 Investigations in the Kara-Kasmak – Sary-Bulak interfluve**

In the interfluve of right inflows of the Karakuldzha-Chatkalskaya river (Dzhashilsai, Dzhosho and others) (Fig. 14A) there are no clear evidences of vertical differential displacements of the TFF walls. However in the south of the region the vertical component can be observed. Thus, above the mouth of the Chiimtash river (right inflow of the Karakuldzha river) and a pass of the same name one can observe a typical upslope facing scarp up to 8 km long half-filled with clastic material that results in only 1.0-1.5 m height difference between the hanging and foot walls of the fault. Nevertheless, it is seen from walls of gullies cutting the seismogenic structure that the depth of the upslope facing scarp is about 4 m (Chediya, 1986). V.S. Burtman et al. (1987) reported a depth of the seismic depression (or the height of the seismic scarp in the shape of swell) up to 5 m (point 23 in Fig. 15). Besides, it is possible to observe a 10 m thick Middle Pleistocene moraine adjoining to the downslope vertically thrusted wall of the fault (the left slope of the Karakuldzha river valley with elevation mark 3296 m), i.e. in this case there is a clearly marked 15-m thrusting of the north-eastern wall in the Late Pleistocene – Holocene. If to prolong mentally toward the fault the Early Pleistocene surface of the fault walls (Fig. 14 B) the stated above will be proved (Chediya, 1986).

V.S. Burtman et al. (1987) noted that the vertical fault plane is clearly marked due to the ragged relief. By the field station 20 (Fig. 15) in a section of a slope of the spring valley one can observe changing of a tilt of the fault plane upward. At the depth 20 m the nearly vertical (80) fault plane becomes more gentle (up to 45); near the surface it becomes steeper again (65). Such phenomenon is probably conditioned by plastic flow of Quaternary alluvial sediments downward the valley slope. The flow was more intense at the depth 5-20 m that resulted in preservation of the steeper fault plane near the surface and tectonic swell near the fault line (Burtman et al., 1987).

At the Karakuldzha-Narynskaya river head (Fig.14 C) the amplitude of the north-eastern wall thrusting can be inferred from the shifted Early Pleistocene surface as 250-300 m (Chediya, 1986). As for horizontal movements along the TFF within the region, V.S. Burtman et al. (1987) found 26 ruptured forms of the modern relief within the site of 16-km long, showing horizontal displacement along the fault line (Table 2). The majority of the forms are small channels. At two points (18 and 24 in Table 2 and Fig. 15) one can observed shifted slopes of channels and at other two points (2 and 20) watersheds of mountain ridges are shifted. Statistically horizontal displacements up to 25-35 m are prevailing (histogram in Fig. 16), but in some cases (marked by a star in Table 2) displacements up to 35-45 m have been measured. In some places the main line of the fault is divided into some branches which either die out (point 11 in Fig. 15), or join the fault line again limiting the tectonic lens (point 23). Moreover, fault-satellites pass parallel to the main fault line at the distance of 70- 100 m, along which horizontal displacement of the relief forms from 3 up to 10 m is observed (points 15 and 18). Displacement along the fault-satellites together with plastic deformation of the fault walls compensates changing amplitude of the slip along the main fault plane (Burtman et al., 1987).

For determination of the slip time Burtman et al. (1987) used organic material collected in pits. 5 samples were collected for radiocarbon age determination by Burtman et al. (1987) in pits dug in a drainless sagging (a, Fig. 15 and 16) and depressions (b-d, Fig. 15 and 16) with springs draining along the fault plane. The samples were collected from clays rich in organic material and overlaid by thick soil. Radiocarbon ages (determinations GIN-4300-4304) determined from the total of accumulated organic material is always younger than the time of the accumulation beginning. Therefore, the oldest age (2020±50 years) is starting the

displacements of the TFF walls. However in the south of the region the vertical component can be observed. Thus, above the mouth of the Chiimtash river (right inflow of the Karakuldzha river) and a pass of the same name one can observe a typical upslope facing scarp up to 8 km long half-filled with clastic material that results in only 1.0-1.5 m height difference between the hanging and foot walls of the fault. Nevertheless, it is seen from walls of gullies cutting the seismogenic structure that the depth of the upslope facing scarp is about 4 m (Chediya, 1986). V.S. Burtman et al. (1987) reported a depth of the seismic depression (or the height of the seismic scarp in the shape of swell) up to 5 m (point 23 in Fig. 15). Besides, it is possible to observe a 10 m thick Middle Pleistocene moraine adjoining to the downslope vertically thrusted wall of the fault (the left slope of the Karakuldzha river valley with elevation mark 3296 m), i.e. in this case there is a clearly marked 15-m thrusting of the north-eastern wall in the Late Pleistocene – Holocene. If to prolong mentally toward the fault the Early Pleistocene surface of the fault walls (Fig. 14

V.S. Burtman et al. (1987) noted that the vertical fault plane is clearly marked due to the ragged relief. By the field station 20 (Fig. 15) in a section of a slope of the spring valley one can observe changing of a tilt of the fault plane upward. At the depth 20 m the nearly vertical (80) fault plane becomes more gentle (up to 45); near the surface it becomes steeper again (65). Such phenomenon is probably conditioned by plastic flow of Quaternary alluvial sediments downward the valley slope. The flow was more intense at the depth 5-20 m that resulted in preservation of the steeper fault plane near the surface and

At the Karakuldzha-Narynskaya river head (Fig.14 C) the amplitude of the north-eastern wall thrusting can be inferred from the shifted Early Pleistocene surface as 250-300 m (Chediya, 1986). As for horizontal movements along the TFF within the region, V.S. Burtman et al. (1987) found 26 ruptured forms of the modern relief within the site of 16-km long, showing horizontal displacement along the fault line (Table 2). The majority of the forms are small channels. At two points (18 and 24 in Table 2 and Fig. 15) one can observed shifted slopes of channels and at other two points (2 and 20) watersheds of mountain ridges are shifted. Statistically horizontal displacements up to 25-35 m are prevailing (histogram in Fig. 16), but in some cases (marked by a star in Table 2) displacements up to 35-45 m have been measured. In some places the main line of the fault is divided into some branches which either die out (point 11 in Fig. 15), or join the fault line again limiting the tectonic lens (point 23). Moreover, fault-satellites pass parallel to the main fault line at the distance of 70- 100 m, along which horizontal displacement of the relief forms from 3 up to 10 m is observed (points 15 and 18). Displacement along the fault-satellites together with plastic deformation of the fault walls compensates changing amplitude of the slip along the main

For determination of the slip time Burtman et al. (1987) used organic material collected in pits. 5 samples were collected for radiocarbon age determination by Burtman et al. (1987) in pits dug in a drainless sagging (a, Fig. 15 and 16) and depressions (b-d, Fig. 15 and 16) with springs draining along the fault plane. The samples were collected from clays rich in organic material and overlaid by thick soil. Radiocarbon ages (determinations GIN-4300-4304) determined from the total of accumulated organic material is always younger than the time of the accumulation beginning. Therefore, the oldest age (2020±50 years) is starting the

B) the stated above will be proved (Chediya, 1986).

tectonic swell near the fault line (Burtman et al., 1987).

fault plane (Burtman et al., 1987).

Fig. 14. Geological-geomorphological cross-sections of the Talas-Fergana Fault zone (modified from Chediya, 1986). T-F – Talas-Fergana Fault, L-N – the line of Nikolaev.

Fig. 15. Talas-Fergana Fault in the Karakuldzha-Narynskaya and Karakuldzha-Chatkalskaya river basins (modified from Burtman et.al., 1987). 1 – line of the modern fault, 2 – tectonic swell, 3 – near-fault valleys, 4 – streams.

Fig. 16. Position of radiometric samples in sections of near-fault saggings and the histogram of recent movements along the fault (modified from Burtman et.al., 1987): 1 – soil and peat rich in organic material; 2 – clay, loam, clay sand; 3 – basement rocks.

swamping along the line of the Talas-Fergana Fault. V.S. Burtman et al. (1987) also examined a sample (GIN-4299) collected in the upper part of the peat bog at the point 11 (Fig.15). The age of the sample is 25050 years that does not contradict to the age given above.

The investigation carried out by V.S. Burtman et al. (1987) allows conclusion that a displacement with a horizontal amplitude 30-40 m and small vertical component occurred along the Talas-Fergana Fault at historic time. This displacement took place after an earthquake occurred about 2000 years ago.

Passing downward along the Karakuldzha-Narynskaya river up to the place where the fault divides the Atoinock uplift and Ketmentyube depression one can observe turning of the hanging north-eastern wall of the fault into the foot wall (Fig. 14 D). On the right watershed of the Ustasai river the amplitude of vertical displacement is equal to previous one in the upper reaches of the Karakuldzha-Narynskaya river but with reverse orientation. Near the Naryn river valley the amplitude reaches more than 1000 m (Chediya, 1986).

V.G. Trifonov et al. (1990) found Holocene gullies and small watersheds on the right bank of the Dhzanaryksay river shifted for 6, 8-9 and 14-17 m, respectively. The upper reaches of the river with traces of trough structure and a lateral moraine aging to the end of the Middle Pleistocene are shifted for 1.5 km relatively nowadays cut off lower parts of the valley. In other words the shifting of the Dzhanaryksai river valley started in the beginning of the Late

Chatkalskaya river basins (modified from Burtman et.al., 1987). 1 – line of the modern fault,

Fig. 16. Position of radiometric samples in sections of near-fault saggings and the histogram of recent movements along the fault (modified from Burtman et.al., 1987): 1 – soil and peat

swamping along the line of the Talas-Fergana Fault. V.S. Burtman et al. (1987) also examined a sample (GIN-4299) collected in the upper part of the peat bog at the point 11 (Fig.15). The age

The investigation carried out by V.S. Burtman et al. (1987) allows conclusion that a displacement with a horizontal amplitude 30-40 m and small vertical component occurred along the Talas-Fergana Fault at historic time. This displacement took place after an

Passing downward along the Karakuldzha-Narynskaya river up to the place where the fault divides the Atoinock uplift and Ketmentyube depression one can observe turning of the hanging north-eastern wall of the fault into the foot wall (Fig. 14 D). On the right watershed of the Ustasai river the amplitude of vertical displacement is equal to previous one in the upper reaches of the Karakuldzha-Narynskaya river but with reverse orientation. Near the

V.G. Trifonov et al. (1990) found Holocene gullies and small watersheds on the right bank of the Dhzanaryksay river shifted for 6, 8-9 and 14-17 m, respectively. The upper reaches of the river with traces of trough structure and a lateral moraine aging to the end of the Middle Pleistocene are shifted for 1.5 km relatively nowadays cut off lower parts of the valley. In other words the shifting of the Dzhanaryksai river valley started in the beginning of the Late

rich in organic material; 2 – clay, loam, clay sand; 3 – basement rocks.

earthquake occurred about 2000 years ago.

of the sample is 25050 years that does not contradict to the age given above.

Naryn river valley the amplitude reaches more than 1000 m (Chediya, 1986).

Fig. 15. Talas-Fergana Fault in the Karakuldzha-Narynskaya and Karakuldzha-

2 – tectonic swell, 3 – near-fault valleys, 4 – streams.

Pleistocene, therefore the average slip rate is about 1.5 cm/year. On the right bank of the Dzhanaryksai river a terrace aging to the beginning of the Late Pleistocene as well as the river inflows, formed at the same time, are shifted to the right for 550-650 m. Terraces and valleys formed in the end of the Late Pleistocene are shifted for 150 m.


Table 2. Horizontal displacement of the relief forms along the fault in the Karakuldzha-Narynskaya and Karakuldzha-Chatkalskaya river valleys (modified from Burtman et.al., 1987)

To the north-west the fault is divided into several branches. Thus, there are two parallel branches in the Ustasai river valley. The north-eastern one is younger. One can observed rightlateral displacement of Late Holocene channels along the branch for 10 m. As for the southwestern branch one can observe galleys cutting the Late Pleistocene terrace and shifted for 150 m to the right. The last displacement probably characterizes the whole Holocene slip; in this case the slip rates is about 1.5 cm/year. Thus, average rate slips of Holocene and Holocene-Late Pleistocene displacements along the segment of the fault in the north-west of the Toktogul reservoir display the same tendency as those in the Late Holocene: they increase in 1.5-2 times comparing with slip rates on the south-eastern portions of the fault.

We pointed above, that many researchers reported about displacement of channels of the gullies on first tens meters. Thus V.S. Burtman et al. (1987) cited evidences of strike-slip movements in Karakuldzha-Chatkalskaya and Karakuldzha-Narynskaya river basins (Fig. 15, Table 2).

V.S. Burtman et al. (1987) point that the same width and morphology of displaced parts of the dry valley above and below the strike-slip fault testifies on a fact that the displacement occurred fast. This circumstance has allowed V.S. Burtman et al. to tie such displacements with earthquakes. Citing on a fact that every valley has its only one ancient continuation, they came to a conclusion that observed displacements are the result of one strong earthquake.

For us it's seemed impossible that one-act horizontal displacement along the strike-slip fault can reach a value of 60 m during one event. World experience of recent strong and catastrophic earthquakes gives us examples of 10 m, maximum 15 m displacements during one event (Strom and Nikonov, 1997). It is not clear also a value of possible creep displacements, their contribution into total value of observed displacement of valley's thalvegs. Most probably a number of the strong earthquakes, occurred along described segment of the fault, have led to formation of a shift of gullies on few tens meters.

The same authors (Burtman et al., 1987) cite data on radiocarbon dating of the samples collected by them in pits excavated in the fault zone (Fig. 15 and 16).

In Fig 16 one can observed that at investigated segment of the fault the dated soils give us at least 4 events led to accumulation of loose slope material in a near-fault depression where soil was formed later. Minimum ages of these events:


Stations ## 22 and 23 are most close to places where samples were collected. There 17 m and 20 m displacements correspondently were measured. Most probably for these total (cumulative) values of displacement there were responsible 3 strong earthquakes, that is 6-7 m during one event in average. These values are in agreement with the world data analysied by A.L. Strom and A.A. Nikonov (1997).

#### **3.4 Sary-Bulak test site**

The second test site was located in the Sary-Bulak river basin of the Toktogul Region of the Jalal-Abad oblast (Fig. 1 and 2). The mapped area was in the left bank of the upper part of the Sary-Bulak river (the right tributary of the Uzun-Akhmat river). Here we observed the right-lateral displacement of many river beds of small water and dry gullies, as well as watersheds between them.

We performed mapping with electronic tachymeter of the area in the upper part of the valley of a large dry stream along the TFF zone (Fig. 17 and 18). Here systematic displacement to the east (to the right) took place at 65,40 m (section А-Б in Fig. 17); 113,06 m (section А-В in Fig. 17) and 336,49 meters (section А-Г in Fig. 17) of the bottom parts of the dry valley. Age of the displaced dry valley, apparently, is the beginning of late Pleistocene, since it cuts the mid-Quarterly alluvial surface.

Bore pits, dug on the left slope of the «legless» valleys (Fig. 19), give a similar picture of the near-surface deposits developed there. At the top there is the soil developed on the loesslike loam. This layer of loam with scattered detritus fragments, probably, is «a colluvial wedge», which collapsed downwards from the nearby slope during the earthquake. Soil formation processes started on loamy deposits after that collapse about 5 thousand years ago (samples SOAN-6529 and SOAN-6531). The age of the above mentioned samples is the minimum age of the earthquake which had occurred there. The samples, which were taken in the top part of the layer, gave radiocarbon ages of 1130 ± 100 years old (SOAN-6528) and

Fig. 17. Sary-Bulak test site and main elements of relief.

50 Earthquake Research and Analysis – Seismology, Seismotectonic and Earthquake Geology

We pointed above, that many researchers reported about displacement of channels of the gullies on first tens meters. Thus V.S. Burtman et al. (1987) cited evidences of strike-slip movements in Karakuldzha-Chatkalskaya and Karakuldzha-Narynskaya river basins (Fig.

V.S. Burtman et al. (1987) point that the same width and morphology of displaced parts of the dry valley above and below the strike-slip fault testifies on a fact that the displacement occurred fast. This circumstance has allowed V.S. Burtman et al. to tie such displacements with earthquakes. Citing on a fact that every valley has its only one ancient continuation, they came

For us it's seemed impossible that one-act horizontal displacement along the strike-slip fault can reach a value of 60 m during one event. World experience of recent strong and catastrophic earthquakes gives us examples of 10 m, maximum 15 m displacements during one event (Strom and Nikonov, 1997). It is not clear also a value of possible creep displacements, their contribution into total value of observed displacement of valley's thalvegs. Most probably a number of the strong earthquakes, occurred along described

The same authors (Burtman et al., 1987) cite data on radiocarbon dating of the samples

In Fig 16 one can observed that at investigated segment of the fault the dated soils give us at least 4 events led to accumulation of loose slope material in a near-fault depression where

Stations ## 22 and 23 are most close to places where samples were collected. There 17 m and 20 m displacements correspondently were measured. Most probably for these total (cumulative) values of displacement there were responsible 3 strong earthquakes, that is 6-7 m during one event in average. These values are in agreement with the world data analysied

The second test site was located in the Sary-Bulak river basin of the Toktogul Region of the Jalal-Abad oblast (Fig. 1 and 2). The mapped area was in the left bank of the upper part of the Sary-Bulak river (the right tributary of the Uzun-Akhmat river). Here we observed the right-lateral displacement of many river beds of small water and dry gullies, as well as

We performed mapping with electronic tachymeter of the area in the upper part of the valley of a large dry stream along the TFF zone (Fig. 17 and 18). Here systematic displacement to the east (to the right) took place at 65,40 m (section А-Б in Fig. 17); 113,06 m (section А-В in Fig. 17) and 336,49 meters (section А-Г in Fig. 17) of the bottom parts of the dry valley. Age of the displaced dry valley, apparently, is the beginning of late Pleistocene,

Bore pits, dug on the left slope of the «legless» valleys (Fig. 19), give a similar picture of the near-surface deposits developed there. At the top there is the soil developed on the loesslike loam. This layer of loam with scattered detritus fragments, probably, is «a colluvial

to a conclusion that observed displacements are the result of one strong earthquake.

segment of the fault, have led to formation of a shift of gullies on few tens meters.

collected by them in pits excavated in the fault zone (Fig. 15 and 16).

soil was formed later. Minimum ages of these events:

2. (1450±40 + 1350±60) / 2 = 1400±50 years, 3. (1150±40 + 1220±50) / 2 = 1185±45 years,

by A.L. Strom and A.A. Nikonov (1997).

since it cuts the mid-Quarterly alluvial surface.

15, Table 2).

1. 2020±50 years,

4. 250±50 years.

**3.4 Sary-Bulak test site** 

watersheds between them.

Fig. 18. A fragment of the upper part of the valley of the Sary-Bulak River.

440 ± 45 years old (SOAN-6530). These age determinations indicate the minimum time periods when there were later soil deformations in the investigated area, apparently, caused by the next seismic motions. These deformations have led to burial of soil fragments, where we have taken samples for determination of absolute age, - to switching on in these fragments of "the geological counter of time».

Fig. 19. Bore pits1 (A) and 2 (B) on Sary-Bulak test site.

Bore pit 3 (depth of 155 cm), which was dug directly in the fault zone in the abandoned part of the valley, has shown a more complicated picture (Fig. 20). Here from top we observed a

440 ± 45 years old (SOAN-6530). These age determinations indicate the minimum time periods when there were later soil deformations in the investigated area, apparently, caused by the next seismic motions. These deformations have led to burial of soil fragments, where we have taken samples for determination of absolute age, - to switching on in these

fragments of "the geological counter of time».

Fig. 19. Bore pits1 (A) and 2 (B) on Sary-Bulak test site.

Bore pit 3 (depth of 155 cm), which was dug directly in the fault zone in the abandoned part of the valley, has shown a more complicated picture (Fig. 20). Here from top we observed a significant layer of reclaimed soil with underlying loess-like loam. Under the loam layer there is a layer of buried soil. Contact of the layer of loam and buried soil is very uneven (see Fig. 20).

Fig. 20. Bore pit 3 at the Sary-Bulak test site. Recent reclaimed soil is separated from the buried soil by the «colluvial wedge». Black stains in the drawing are pebbles of average roundedness.

The available section gives us sufficient material for sedimentation history reconstruction and tectonic development of the site in mid-late Holocene. The buried soil started to be formed about 6 thousand years ago (sample SOAN-6536), apparently, after the material displaced during the earthquake from the nearby slope was stabilized in the riverbed of the abandoned valley.

The following sample - SOAN-6534 taken above is an evidence of one more earthquake which occurred about 5 thousand years ago, which traces have been found by us in the neighboring bore pits 1 and 2. The next seismic event took place approximately 2 400 years ago (samples SOAN-6533 and SOAN-6535). It has led to destroying of an ancient soil cover, its coverage by the «colluvial wedge» which is in turn covered by modern reclaimed soil.

The latter seismic event known to us occurred about 400-500 years ago (SOAN-6532). Its traces are reflected also in bore pit 2, which was dug nearby.

It is hardly probable, that in past 6 thousand years there were occurred a displacement at 336.49 meters along the fault (Fig. 17, section А-Г). However displacement at 65.40 m (section А-Б) or 113.06 m (section А-В) is plausible, considering the data published in the world literature (see Strom's and Nikonov's executive summary published in 1997). It is necessary to note, that Burtman et al. (1996) informed on regular displacement of valleys of temporary waterways along the Talas-Fergana fault by 110 ± 10 m in the (from the west to the east) Ustasay, Sary-Bulak, Dzhanaryksay river basins.

Thus, average rate of displacement along the TFF at the Sary-Bulak test site, since middle of Holocene, can reach 10.7-18.5 mm/year. Our results are close to the data received by Burtman et al. (1996) on the site located in several kilometers SE from Sary-Bulak. In the Dzhanaryksay river basin the mentioned authors calculated rate of displacement of 9-12 mm/year during last ~1.5 thousand years on 14-metre displacement of a dry gully.

Let's notice, that the closest age determinations of 1150±40 years old (GIN-4302) and 1220±50 (GIN-4304) were received by Burtman et al. (1987) in the bottom part of modern soil along the TFF line in upper area of the Chatkal river, approximately at the distance of 70 km to NW from Sary-Bulak test site. These data assume not shorter length of the seismogenic rapture.

Probably, one more age determination got by Trifonov et al. (1990) concerns this age group: 1240 ± 60 year ago; it was received by them on the right bank of the Keklikbel river. The latter site is also located at a considerable distance of more than 100 km to SE from the Sary-Bulak test site. Then the plane of the seismogene rupture begins with upper reach of the Chatkal river and stretches for the distance of almost 200 km are to the valley of the Keklikbel river. Theoretically it is possible: we can recollect, that the length of the well studied rupture of the Kebin (Kemin) earthquake of 1911 (М=8,2) in Northern Tian-Shan reached this value (Bogdanovich et al., 1914). However in the valley of Keklikbel river probably also that there was occurred an independent earlier event.

Thus, minimum number of the earthquakes which occurred at this site in Holocene is 5. Their minimum ages are ~ 6 thousand years ago, ~ 5 thousand years ago, ~ 2.4 thousand years ago, ~ 1.1 thousand years ago and 400-500 years ago.

The earliest (~ 6 thousand years) and latest age determinations (400-500 years ago) coincide with the absolute data received at the Kara-Bura test site, located at the distance of 90 km to NW from the Sary-Bulak test site. This fact assumes similar length of the seismogenic rupture formed twice in the middle and in the end of the Holocene.

Data of absolute age (400-500 years ago), received on the Kara-Bura and Sary-Bulak test sites coincide also with archeoseismologic data on the destroyed caravanserai located in an middle part of the Kara-Bura river valley (Korjenkov et al., 2009).

#### **4. Middle part of the Talas-Fergana fault**

#### **4.1 Investigations in Ustasay-Karasu interfluve**

In a distance of 6 km south of the Naryn river the Talas-Fergana fault limits the Ketmentyube depression filled with the Toktogul reservoir. Features of vertical movements are analogous to previous ones. Further, up to the Kokbel pass, the fault plane passes on the left bank of a nameless channel which is drained to the Naryn river. The whole low left slope of the valley is cut with short (up to 500-700 m) small gullies which were probably formed in the Late Pleistocene-Holocene (Chediya, 1986). The whole slope with all gullies is cut by the Talas-Fergana fault. In a profile of the gullies' watersheds one can clearly observe

It is hardly probable, that in past 6 thousand years there were occurred a displacement at 336.49 meters along the fault (Fig. 17, section А-Г). However displacement at 65.40 m (section А-Б) or 113.06 m (section А-В) is plausible, considering the data published in the world literature (see Strom's and Nikonov's executive summary published in 1997). It is necessary to note, that Burtman et al. (1996) informed on regular displacement of valleys of temporary waterways along the Talas-Fergana fault by 110 ± 10 m in the (from the west to

Thus, average rate of displacement along the TFF at the Sary-Bulak test site, since middle of Holocene, can reach 10.7-18.5 mm/year. Our results are close to the data received by Burtman et al. (1996) on the site located in several kilometers SE from Sary-Bulak. In the Dzhanaryksay river basin the mentioned authors calculated rate of displacement of 9-12

Let's notice, that the closest age determinations of 1150±40 years old (GIN-4302) and 1220±50 (GIN-4304) were received by Burtman et al. (1987) in the bottom part of modern soil along the TFF line in upper area of the Chatkal river, approximately at the distance of 70 km to NW from Sary-Bulak test site. These data assume not shorter length of the

Probably, one more age determination got by Trifonov et al. (1990) concerns this age group: 1240 ± 60 year ago; it was received by them on the right bank of the Keklikbel river. The latter site is also located at a considerable distance of more than 100 km to SE from the Sary-Bulak test site. Then the plane of the seismogene rupture begins with upper reach of the Chatkal river and stretches for the distance of almost 200 km are to the valley of the Keklikbel river. Theoretically it is possible: we can recollect, that the length of the well studied rupture of the Kebin (Kemin) earthquake of 1911 (М=8,2) in Northern Tian-Shan reached this value (Bogdanovich et al., 1914). However in the valley of Keklikbel river

Thus, minimum number of the earthquakes which occurred at this site in Holocene is 5. Their minimum ages are ~ 6 thousand years ago, ~ 5 thousand years ago, ~ 2.4 thousand

The earliest (~ 6 thousand years) and latest age determinations (400-500 years ago) coincide with the absolute data received at the Kara-Bura test site, located at the distance of 90 km to NW from the Sary-Bulak test site. This fact assumes similar length of the seismogenic

Data of absolute age (400-500 years ago), received on the Kara-Bura and Sary-Bulak test sites coincide also with archeoseismologic data on the destroyed caravanserai located in an

In a distance of 6 km south of the Naryn river the Talas-Fergana fault limits the Ketmentyube depression filled with the Toktogul reservoir. Features of vertical movements are analogous to previous ones. Further, up to the Kokbel pass, the fault plane passes on the left bank of a nameless channel which is drained to the Naryn river. The whole low left slope of the valley is cut with short (up to 500-700 m) small gullies which were probably formed in the Late Pleistocene-Holocene (Chediya, 1986). The whole slope with all gullies is cut by the Talas-Fergana fault. In a profile of the gullies' watersheds one can clearly observe

mm/year during last ~1.5 thousand years on 14-metre displacement of a dry gully.

probably also that there was occurred an independent earlier event.

rupture formed twice in the middle and in the end of the Holocene.

middle part of the Kara-Bura river valley (Korjenkov et al., 2009).

**4. Middle part of the Talas-Fergana fault 4.1 Investigations in Ustasay-Karasu interfluve** 

years ago, ~ 1.1 thousand years ago and 400-500 years ago.

the east) Ustasay, Sary-Bulak, Dzhanaryksay river basins.

seismogenic rapture.

that a straight depression is located along the fault. Mainly it filled by the colluviumalluvium material of few meters thickness, which is shelved from upper part of the slope, i.e. it is the amplitude of the Holocene thrusting of the Kochkortobe uplift. Comparison of the Early Pleistocene surface (E in Fig. 14) gives the value about 150 m (Chediya, 1986).

Although a trace of the fault in the Ustasay-Zhanaryksay region is less expressed, than in south-east or further north-west, sections of well expressed displacements are distribute along whole zone of the fault, and right-lateral displacement is evident (Fig. 21). Maximum values of displacements of watersheds and spring valleys near station # 14 reach 110 10 m (Burtman et al., 1996).

Burtman et al. (1996) have measured a displacement of a small gully - 142 m in the field station #14 (Fig. 21). A pit in upper part of the gully near the fault trace reaches hard rocks in a depth of 0.5 m under the soil of 0.45 m thickness. Organic part of the soil - 0.1 m from the whole layer (depth 0.35-0.45 m) gives the radiocarbon age of the seismic event equal 144030 years ago.

We have studied segments of the Talas-Fergana Fault zone in 2 km south-east from the Aktaybulak-Korumtokay interfluve – in a region of the Ustasay-Sarybulak pass. Everywhere we measured right-lateral displacement of spring beds and watersheds between them on a value of first ten meters (Fig. 22 a). Eastwards, a displacement of the spring valleys and watersheds between them reaches hundreds meters (as, for example, in the right bank of the Sarybulak River). It is necessary to mention also a vertical uplifting of southwestern limb of the fault on a value of 20 m. In tie with lateral movements along the fault, in some places the consequent spring valleys, flowing down the slope, are blocking by so-called barrier ridges (Fig. 22 b), in front of which there are forming a local depressions, where fine material is accumulated.

Fig. 21. Topographic map of a fragment of the Talas-Fergana Fault in the Ustasay-Dzhanaryksay interfluve (modified after Burtman et al., 1996). Black line marks a trace of the Talas-Fergana Fault. A location of the field station # 14 is shown. Contour interval in 200 m.

Fig. 22. a. Right-lateral displacement of a channel of a dry gully (white arrows) and watershed ridge along the line of the Talas-Fergana Fault (shown by the dotted line).

Fig. 22. b. Barrier ridge (a shepherd tent is by foot of it) in a zone of the Talas-Fergana Fault (shown by the dotted line) in the Aktaybulak-Korumtokay interfluves.

Parameters of the Strong Paleoearthquakes Along the Talas-Fergana Fault, the Kyrgyz Tien Shan 57

Fig. 22. c. A structure of the Talas-Fergana Fault zone in the Aktaybulak-Korumtokay interfluves. Fault gouge, filled the fault zone, is overlaid by folded alluvial deposits, in which fragments blue clays laying below are found.

In 500 m south-east of previous field station an irrigation canal exposes the fault zone. In south-western wall of the canal there are exposed (up-down):


56 Earthquake Research and Analysis – Seismology, Seismotectonic and Earthquake Geology

Fig. 22. a. Right-lateral displacement of a channel of a dry gully (white arrows) and watershed ridge along the line of the Talas-Fergana Fault (shown by the dotted line).

Fig. 22. b. Barrier ridge (a shepherd tent is by foot of it) in a zone of the Talas-Fergana Fault

(shown by the dotted line) in the Aktaybulak-Korumtokay interfluves.


Last two horizons one can observe in Fig. 22 c. A layer, consisting of clays and loams, is representing itself fault gouge, filled the fault zone.
