**4.2 Kok-Bel test site**

The third test site, examined by us, is located approximately at the distance of 30 km to the southeast from Sary-Bulak test site (Fig. 1 and 2). Here during climbing up the pass with the same name from the Ketmen-Tyube depression a systematic right-lateral displacement of small dry gullies and watersheds between them is observed along the TFF zone (Fig. 23). Here we also produced a detailed digital map of the site of the TFF zone by an electronic tachymeter (Fig. 24).

Along one kilometre section of the fault we measured systematic displacement of forms of the modern relief, the upper parts of which are shifted to the right along the fault line in a horizontal direction (Table 3). All measured elements of the relief were small dry valleys of temporary waterways - "says", as well as watersheds between them.

Fig. 23. Displacement to the right of dry rivulets (says) (white faltering lines) and watersheds between them (black faltering lines) along the Talaso-Fergana fault line (a sub horizontal faltering line) to NW from the Cook-Bel pass.

Fig. 24. The Topographic map the Kok-Bel test site, drawn according to shooting by electronic tachometer. Values of displacement of dry waterways beds are shown.


Fig. 23. Displacement to the right of dry rivulets (says) (white faltering lines) and

Fig. 24. The Topographic map the Kok-Bel test site, drawn according to shooting by electronic tachometer. Values of displacement of dry waterways beds are shown.

horizontal faltering line) to NW from the Cook-Bel pass.

watersheds between them (black faltering lines) along the Talaso-Fergana fault line (a sub


Table 3. Horizontal displacement of forms of a relief along the Talas-Fergana fault to NW from the Kok-Bel pass, measured by electronic tachometer

To NW from the Kok-Bel pass we measured 25 valleys of dry gullies (says) and watersheds between them displaced to the right at the distance of up to 120 m. Pay attention at the various representativeness of displacement of different amplitude: some maxima are emphasized whereas intermediate values are absent (Fig. 25). Similar non-uniformity according to Trifonov et al. (1990) indicates the decisive contribution of impulse seismogenic motions into total movement.

To north-east from the the Kok-Bel pass along the Talas-Fergana fault we dug 2 bore pits and cleaned a natural exposure - a scar of a landslide, which was directly in the fault zone (Fig. 26 and 27). Value of horizontal displacement of the dry rivulet and the adjacent watershed, measured by electronic tachymeter, comprised 76 m. We correlated this value with the most ancient date - 4900 ± 230 years. Thus, we receive probable rates of horizontal tectonic movements since mid Holocene equal to 14.81-16.27 mm/year.

Fig. 25. The histogram of amplitudes distribution of Holocenic right-hand-side displacements of dry water currents and watersheds between them on the Talas-Fergana fault to NW from the Kok-Bel pass (only 25 measurements).

The last strong earthquake on the examined site occurred, apparently, about ~275 years ago. An evidence to that are absolute dates of samples SOAN-7021 (270 ± 85 years), SOAN-7024 (370 ± 90 years) and SOAN-7026 (240 ± 50 years), taken in bore pit No 1 and clearing No 3 in the lower parts of the modern soil developed on the underlying «colluvial wedge». Similar dates were received also by Burtman et al. (1987) along TFF line in upper reach of the Chatkal river - 250 ± 50 years (GIN-4299). Though distance between the Kok-Bel test site and the Karakuldzha-Chatkalskaya river valley is about 100 km, but apparently this was the length of the plane of the seismogenic rupture of this age.

In the same bore pit at the depth of 30 cm we found a fragment of ceramic ware. The form is spherical, the fragment was produced on a potter's wheel from well mixed clay with addition of coarse-grained and fine-grained sand. Firing of the fragment was uniform. The fragment was blackened with smoke. The indicated signs are the most typical for Middle Ages epoch-VIII-XII centuries AD (K.S.Tabaldiev, a written communication). For narrowing down the date determination additional artifacts are necessary. On the other hand, manufacturing of similar thin-walled spherical ceramics in Fergana valley and in its foothills is an ancient tradition. Production of ceramic vessels made on a potter's wheel were introduced in I millennium BC. They are characterized by red angob and painted ceramics which we do not see in this fragment. The described fragment of ceramics had been buried during the collapse of the colluvial wedge which occurred during the last strong earthquake. Subsequently on sediments of this wedge the soil the age of which was determined as 270 ± 85 years old (SOAN-7021) was formed.

with the most ancient date - 4900 ± 230 years. Thus, we receive probable rates of horizontal

tectonic movements since mid Holocene equal to 14.81-16.27 mm/year.

Fig. 25. The histogram of amplitudes distribution of Holocenic right-hand-side

fault to NW from the Kok-Bel pass (only 25 measurements).

length of the plane of the seismogenic rupture of this age.

determined as 270 ± 85 years old (SOAN-7021) was formed.

displacements of dry water currents and watersheds between them on the Talas-Fergana

The last strong earthquake on the examined site occurred, apparently, about ~275 years ago. An evidence to that are absolute dates of samples SOAN-7021 (270 ± 85 years), SOAN-7024 (370 ± 90 years) and SOAN-7026 (240 ± 50 years), taken in bore pit No 1 and clearing No 3 in the lower parts of the modern soil developed on the underlying «colluvial wedge». Similar dates were received also by Burtman et al. (1987) along TFF line in upper reach of the Chatkal river - 250 ± 50 years (GIN-4299). Though distance between the Kok-Bel test site and the Karakuldzha-Chatkalskaya river valley is about 100 km, but apparently this was the

In the same bore pit at the depth of 30 cm we found a fragment of ceramic ware. The form is spherical, the fragment was produced on a potter's wheel from well mixed clay with addition of coarse-grained and fine-grained sand. Firing of the fragment was uniform. The fragment was blackened with smoke. The indicated signs are the most typical for Middle Ages epoch-VIII-XII centuries AD (K.S.Tabaldiev, a written communication). For narrowing down the date determination additional artifacts are necessary. On the other hand, manufacturing of similar thin-walled spherical ceramics in Fergana valley and in its foothills is an ancient tradition. Production of ceramic vessels made on a potter's wheel were introduced in I millennium BC. They are characterized by red angob and painted ceramics which we do not see in this fragment. The described fragment of ceramics had been buried during the collapse of the colluvial wedge which occurred during the last strong earthquake. Subsequently on sediments of this wedge the soil the age of which was

Fig. 26. Schematic drawing of bore pits No 1 (A) and No 2 (B), which were examined in the Talas-Fergana fault zone at the distance of 1 km to NE from the Kok-Bel pass.

Fig. 27. A schematic drawing of clearing of the natural exposure – a scar, which was formed due to landslide collapse in the zone of the Talas-Fergana fault at the distance of 1 km to NW from the Kok-Bel pass.

The following strong earthquake occurred approximately ~2400 years ago. Absolute date determinations of two samples (SOAN-7022 - 2340 ± 120 years and SOAN-7025 - 2500 ± 100 years) testify to it, the samples were taken in bore pit No 1 and clearing No 3 in the lower part of buried soil. It is not excluded, that this earthquake has left its traces also in the river valleys: Kyldau (dated as 2320 ± 40 years ago according to Trifonov et al. (1990)) and Pchan (dated as 2180 ± 120 years – GIN -7052; 2280 ± 70 years - Beta-47550; 2540 ± 70 years - Beta-47549 according to Burtman et al. (1996)). If it so, the length of this seismogenic rupture reached 100-200 km.

The most ancient earthquake, the traces of which were found at the Kok-Bel test site, took place about 5 thousand years ago to what absolute dating of sample SOAN-7023 testifies.

#### **4.3 Investigations in the Karasu-Kyldau interfluve**

Further south-east the zone of the fault crosses the Toktogul water reservoir and stretches along the Karasu river valley (Fig. 1 and 28). However well-expressed displacements (up to

Fig. 27. A schematic drawing of clearing of the natural exposure – a scar, which was formed due to landslide collapse in the zone of the Talas-Fergana fault at the distance of 1 km to

The following strong earthquake occurred approximately ~2400 years ago. Absolute date determinations of two samples (SOAN-7022 - 2340 ± 120 years and SOAN-7025 - 2500 ± 100 years) testify to it, the samples were taken in bore pit No 1 and clearing No 3 in the lower part of buried soil. It is not excluded, that this earthquake has left its traces also in the river valleys: Kyldau (dated as 2320 ± 40 years ago according to Trifonov et al. (1990)) and Pchan (dated as 2180 ± 120 years – GIN -7052; 2280 ± 70 years - Beta-47550; 2540 ± 70 years - Beta-47549 according to Burtman et al. (1996)). If it so, the length of this seismogenic rupture

The most ancient earthquake, the traces of which were found at the Kok-Bel test site, took place about 5 thousand years ago to what absolute dating of sample SOAN-7023 testifies.

Further south-east the zone of the fault crosses the Toktogul water reservoir and stretches along the Karasu river valley (Fig. 1 and 28). However well-expressed displacements (up to

NW from the Kok-Bel pass.

reached 100-200 km.

**4.3 Investigations in the Karasu-Kyldau interfluve** 

tens of meters) along the fault one can observe south-west and north-east from the valley. Watersheds and spring valleys were displaced from 40 to 225 m and >2 km. Not far from the Kok-Bel pass, where main highway from the Fargana Valley crosses the Ketmen'-Tyube Depression which is occupied by the Toktogul water reservoir at present time, right-lateral displacements on 300 m are well expressed. It is especially well visible in south-western slope of a dry valley stretched parallel to the fault trace.

Fig. 28. A topographic map of the Talas-Fergana Fault Fragment along the Karasu River Valley (Burtman et al., 1996). A black line is a fault trace. Isoline intervals – 200 m.

In the south-east of the Kokbel pass the Talas-Fergana fault cuts the lower part of the right slope of the Karasu river valley for the distance of 6 km. The zone is characterized by strongly crushed Devonian deposits and numerous springs, that provokes to wide development of landslides. If to go upward on the Middle Pleistocene terraces of the left slope of the Karasu river one can observe a clearly expressed right-lateral slip along the fault. Here the length of galleys (therefore, time of their development) increases, and the slip amplitude is up to 15-20 m (Chediya, 1986). Analysis of geomorphologic levels does not allow determination of vertical displacement amount.

Further the fault passes on the left slope of the Karasu valley. In two kilometers of the place there is a seismic rockslide (considering its characteristic form) occurred along the fault and dammed the half of the valley. Here on the left slope of the Karasu river there are also gullies large in the length and time of development (since the Middle Pleistocene). One can clearly observe right-lateral displacement of lower parts of the gullies along the fault plane. If the displacement in small galleys reaches 15-20 m, the large ones are characterized by 100 m displacement (Chediya, 1986). If it is so, we can speak of long-term strike-slip fault which was active for the most part of the Quaternary. For example, E.Ya. Rantcman and G.N. Pshenin (1967) reported a 750 m displacement of a Middle Pleistocene moraine in the upper reaches of the Karasu-Eastern river.

A cross zone of the Shaldyrak river by the fault is very interesting. Meeting the fault, the river turns at the angle 90 and flows along the fault for the distance of 2 km, then it turns again and inflows into the Karasu river. The bending of river channel as well as a watershed between its lower parts and the Karasu valley agrees with the scheme of the right-lateral displacement for 2 -2.5 km which could occur during the Pleistocene. However, it would be irresponsibly to infer that basing only on one river bending. Comparison of hypsometrical elevation of the terrace QII1 on both sides of the fault testifies to thrusting of north-eastern wall for 150 m (Chediya, 1986).

Above the Shaldyrak river mouth up to the Karasu lake the fault passes on the central part (near-bed) of the valley (Fig. 29); so we cannot speak of any evidences of displacement here.

South-east of the "Karasu" fragment the Talas-Fergana Fault crosses a high-elevated area where a moraine was displaced on 30 м (Burtman, 1964).

Further to the south-east along the TFF zone (the Kuroves and Keklikbel rivers' basins) V.G. Trifonov et al. (1990) reported on 32 structures shifted to the right for 36 m (see Fig. 30 c). Displayed maximums of displacement are not so contrasting as those reported by V.G. Trifonov et al.(1990) in the south-east (Fig. 30 a, b), and not correlated to them. V.G. Trifonov et al. (1990) supposed that either other impulses of movements occurred here or creep processes were prevailing. The second supposition is more probable due to abundant elastic clay-shale rocks in the portion of the fault. Movements along the fault and damming caused formation of small depressions with peat bogs on the right slope of the Keklikbel river valley in the surface of a moraine deposited by a small Late Pleistocene glacier. The radiocarbon age of one of samples collected in the bottom of one of the depressions is 1240±60 years (9 in Fig. 31). In the immediate vicinity of the depression one can observe only general displacement of the late Pleistocene moraine occurred during much larger time interval. However, in the north-west there is a number of Late Holocene displacements evidences with prevailing amplitude 10-12 m. If the radiocarbon age of the depression is characterized the displacement, the average slip rate does not exceed 0.8-1 cm/year, i.e. close or a little higher than that on the site to the south-east of the described one (Trifonov et al., 1990).

Fig. 29. Straight-line segment of the Karasu-Eastern River from the Shyldyrak massif to Karasu rockslide. The photograph was made from the rockslide body toward north-west.

displacement for 2 -2.5 km which could occur during the Pleistocene. However, it would be irresponsibly to infer that basing only on one river bending. Comparison of hypsometrical elevation of the terrace QII1 on both sides of the fault testifies to thrusting of north-eastern

Above the Shaldyrak river mouth up to the Karasu lake the fault passes on the central part (near-bed) of the valley (Fig. 29); so we cannot speak of any evidences of displacement here. South-east of the "Karasu" fragment the Talas-Fergana Fault crosses a high-elevated area

Further to the south-east along the TFF zone (the Kuroves and Keklikbel rivers' basins) V.G. Trifonov et al. (1990) reported on 32 structures shifted to the right for 36 m (see Fig. 30 c). Displayed maximums of displacement are not so contrasting as those reported by V.G. Trifonov et al.(1990) in the south-east (Fig. 30 a, b), and not correlated to them. V.G. Trifonov et al. (1990) supposed that either other impulses of movements occurred here or creep processes were prevailing. The second supposition is more probable due to abundant elastic clay-shale rocks in the portion of the fault. Movements along the fault and damming caused formation of small depressions with peat bogs on the right slope of the Keklikbel river valley in the surface of a moraine deposited by a small Late Pleistocene glacier. The radiocarbon age of one of samples collected in the bottom of one of the depressions is 1240±60 years (9 in Fig. 31). In the immediate vicinity of the depression one can observe only general displacement of the late Pleistocene moraine occurred during much larger time interval. However, in the north-west there is a number of Late Holocene displacements evidences with prevailing amplitude 10-12 m. If the radiocarbon age of the depression is characterized the displacement, the average slip rate does not exceed 0.8-1 cm/year, i.e. close or a little higher than that on the site to the south-east of the described one (Trifonov et

Fig. 29. Straight-line segment of the Karasu-Eastern River from the Shyldyrak massif to Karasu rockslide. The photograph was made from the rockslide body toward north-west.

wall for 150 m (Chediya, 1986).

al., 1990).

where a moraine was displaced on 30 м (Burtman, 1964).

Fig. 30. Histograms of amplitude distribution for the Holocene right-lateral deviation of channels and other small forms of the relief along the Talas-Fergana Fault (modified from V.G.Trifonov et.al., 1990): a – from the Kok-Kiya pass to the Biruza river (points show probable strong earthquakes); б – in the Pchan and Kyldou river valleys; *в* - in the Kuroves and Keklikbel river valleys. The horizontal axis – amplitude of displacement; the vertical one – number of shifted forms of the relief.

On the right slope of the Keklikbel river valley V.G. Trifonov et al. (1990) found right-lateral displacement of channels for 60, 70-80, 135, 230, 300, 450-500 and 700-800 m. Shifted trough valleys and moraine deposits filling the troughs are of the greatest interest. The Late Pleistocene moraine is shifted for 135 m to the right. It is unclear if the time of its depositing is only the Holocene or the end of the Pleistocene too. More clear evidence is shifting of the Late Pleistocene moraine for 700-800 m relatively its trough valley. This moraine of about 10 m thick covers a depression cut in the surface of the beginning Middle Pleistocene (40-50 m), i.e. it was formed in the end of the Middle Pleistocene. Obviously, the moraine displacement can occur in the Late Pleistocene and Holocene, i.e. for the last 100 000 years, that gives the average slip rate 0.7-0.8 cm/year, which is close to the slip rate of the Late Holocene displacement between the Pchan and Kyldou river valleys. It is necessary to add that the mentioned amplitudes of displacements are accompanied by thrusting of the south-western wall of the fault. For displacement 135 m the thrusting is up to 5-6 m, for 300 m – 8-10 m, for 450-500 m and 700-800 m – several tens meter. Thus, the vertical component is 10-30 times less than the horizontal one (Trifonov et al., 1990).

Fig. 31. Sections of near-fault depressions and their radiocarbon ages within the Talas-Fergana Fault (modified from V.G.Trifonov et.al., 1990). 1-3 – the first portion of the fault: 1 - 3970±40 years, left bank of the Bolsun river in the north-west of the Kok-Kiya pass. 2 - 15800±1300 years, 3 - 4590±100 years, the both ages are from a section on the right bank of the Chitta-Severnaya river in the north-west of the Dzhilangach river; 4-7 – the second portion of the fault: 4 - 3740±600 years, interfluve of the Biruza and Pchan rivers in the north-west of the pass between them. 5 - 3150±40 years, to the north-west of the previous point, 6 - 2640±600 years, upper reaches of the Pchan river in the south of the Chityndy pass, 7 - 2320±40 years, upper reaches of the Kyldou river valley; 8 – the third portion of the fault: 1510±60 years, interfluve of the Urumbash and Kuroves rivers; 9 –the firth portion of the fault: 1240±60 years, right bank of the Keklikbel river; 10-14 – the sixth portion of the fault (from V.S.Burtman et.al., 1987): 10 -1220±50 years, upper reaches of the Chatkal river, 11 - 2020±50 years, 12 - 1150±40 years and 13 - 1350±60 years, in 1 km to the north-west of the previous place, 14 - 1450±40 years, in 1,5 km to the north-west of the previous place. 1 – detritus, 2 – sand, 3 – clay, loam, sandy loam, 4 – the same rocks rich in organic material, and peat bogs; 5 – basement rocks.

Right bends of large river valleys (Kugart, Kaldama, Kyzylsu, Urumbash, Kuroves and oth.), described by V.G. Trifonov et al. (1990), allows inferring their right-lateral shifting for 3-4 km. These authors studied some of the bending channels and reconstructed their initial form. Minimal value of the right deviation is: for the Kongurtobe river – 2.8 km, the Kaldama river – 1.8 km, the Molasu river – 1.8 km, the Dzhindisu river – 2 km and the Urumbash river – 1.5 km. Since the mentioned river valleys were formed in the Early Pleistocene (Kostenko et al., 1972; Makarov, 1977), duration of slip processes is hundreds thousand years.

#### **4.4 Investigations in the Kuroves-Malasu interfluve**

In mentioned region the Talas-Fergana Fault goes along a foot of north-eastern slope of the Fergana Range. South-western limb of the fault, composed by Paleozoic rocks of the range, is trusted on many hundreds of meters, and north-eastern limb, covered by the Cenozoic deposits, is dropped on the same value. Here, in significant degree there are developed gravitation formations, the scars of which are in north-eastern slope of the Fergana Range, and landslide bodies descend into the Kazarman (Kugart) depression.

#### **5. South-eastern part**

#### **5.1 Investigations in the Kyldou-Dzhilangach interfluve**

There are a few evidences of Holocene slip along the TFF between the Pchan and Kyldou river valleys, since here the fault passes along river beds. V.G. Trifonov et al. (1990) found right-lateral displacement of small channels for 17-20, 34-37, 60-66 m. The radiocarbon age of a sample collected from loams composing a gulley terrace shifted for 17-20 m (?) is 151060 years (section 8 in Fig. 31). This means that the average slip rate is less than 1.1-1.3 cm/year.

Fig. 31. Sections of near-fault depressions and their radiocarbon ages within the Talas-Fergana Fault (modified from V.G.Trifonov et.al., 1990). 1-3 – the first portion of the fault: 1 - 3970±40 years, left bank of the Bolsun river in the north-west of the Kok-Kiya pass. 2 - 15800±1300 years, 3 - 4590±100 years, the both ages are from a section on the right bank of the Chitta-Severnaya river in the north-west of the Dzhilangach river; 4-7 – the second portion of the fault: 4 - 3740±600 years, interfluve of the Biruza and Pchan rivers in the north-west of the pass between them. 5 - 3150±40 years, to the north-west of the previous point, 6 - 2640±600 years, upper reaches of the Pchan river in the south of the Chityndy pass, 7 - 2320±40 years, upper reaches of the Kyldou river valley; 8 – the third portion of the fault: 1510±60 years, interfluve of the Urumbash and Kuroves rivers; 9 –the firth portion of the fault: 1240±60 years, right bank of the Keklikbel river; 10-14 – the sixth portion of the fault (from V.S.Burtman et.al., 1987): 10 -1220±50 years, upper reaches of the Chatkal river, 11 - 2020±50 years, 12 - 1150±40 years and 13 - 1350±60 years, in 1 km to the north-west of the previous place, 14 - 1450±40 years, in 1,5 km to the north-west of the previous place.

1 – detritus, 2 – sand, 3 – clay, loam, sandy loam, 4 – the same rocks rich in organic material,

Right bends of large river valleys (Kugart, Kaldama, Kyzylsu, Urumbash, Kuroves and oth.), described by V.G. Trifonov et al. (1990), allows inferring their right-lateral shifting for 3-4 km. These authors studied some of the bending channels and reconstructed their initial form. Minimal value of the right deviation is: for the Kongurtobe river – 2.8 km, the Kaldama river – 1.8 km, the Molasu river – 1.8 km, the Dzhindisu river – 2 km and the Urumbash river – 1.5 km. Since the mentioned river valleys were formed in the Early Pleistocene (Kostenko et al.,

In mentioned region the Talas-Fergana Fault goes along a foot of north-eastern slope of the Fergana Range. South-western limb of the fault, composed by Paleozoic rocks of the range, is trusted on many hundreds of meters, and north-eastern limb, covered by the Cenozoic deposits, is dropped on the same value. Here, in significant degree there are developed gravitation formations, the scars of which are in north-eastern slope of the Fergana Range,

There are a few evidences of Holocene slip along the TFF between the Pchan and Kyldou river valleys, since here the fault passes along river beds. V.G. Trifonov et al. (1990) found right-lateral displacement of small channels for 17-20, 34-37, 60-66 m. The radiocarbon age of a sample collected from loams composing a gulley terrace shifted for 17-20 m (?) is 151060 years (section 8 in Fig. 31). This means that the average slip rate is less than 1.1-1.3 cm/year.

1972; Makarov, 1977), duration of slip processes is hundreds thousand years.

and landslide bodies descend into the Kazarman (Kugart) depression.

**4.4 Investigations in the Kuroves-Malasu interfluve** 

**5.1 Investigations in the Kyldou-Dzhilangach interfluve** 

and peat bogs; 5 – basement rocks.

**5. South-eastern part** 

On the second section – in the Pchan and Kyldou river valleys (see Fig. 30 b) V.G. Trifonov et al. (1990) marked 41 structures shifted to the right up to 45 m. Measured maximums are comparable with maximum displacements on the first site (from the Kokkiya pass up to the Biruza river), but differ (with 2 exclusions) by 10-20% larger slip amplitudes. This fact reflects increasing intensity of movements from the first section to the second one at a similar regime of development, i.e. prevailing of seismogenic movements.

For calculation of slip rates on the both sites V.G. Trifonov et al. (1990) investigated sections of near-fault depressions formed in result of damming of the channels by shifted gulley slopes and upslope facing scarps occurred simultaneously with the slip events. Sediments at the bottom of the depression correspond to the time of the movements (Fig. 31). They show slip rate about 0.5 cm/year for the first section (the sample age is 397040 years at total slip 19 m) and about 0.7 cm/year for the second one (3760600 years at total slip of the gulley on 271 m). The values agree with other age determinations on the first and second section (see Fig. 31). Thus, in section 2 the age of a sample, collected at the bottom of a 1.5 m scarp synchronous to the 40 m slip along the northern branch of the fault, is 158001300 years, that gives the rate slip along the branch 0.25 cm/year. In sections 5-7 the age of samples from upper layers is 315040, 2640600 and 232040 years at total slip 903, 251 and 23-24 m, respectively. Moreover, the first measurement was made in a gulley characterizing by a right-lateral displacement of the channel for 7-8 m occurred after filling the near-fault depression with sediments. The displacement gives the probable limit of slip rate 0.24 cm/year (Trifonov et al., 1990).

Between the Kokkiya pass and the Biruza river valley (the right inflow of the Pchan river) V.G. Trifonov et al. (1990) found 75 water channels and other forms of the relief shifted up to 50 m to the right. Different distribution of various amplitudes should be noted: there are 11 or 12 maximums while the intermediate values either are absent or represented by single shifted forms (Fig. 30 a). The irregularity was noted for the first time by R.E. Wallace (1968) at studying of Late Holocene slip along the San Andreas Fault in California. V.G. Trifonov (1985) studied the same phenomenon by the example of the Hangai, Kobdin and Dolinoozersk active faults in Mongolia, where the morphological features of displacement and character of the revealed irregularity are similar to those found on the studied site of the Talas-Fergana Fault. Obviously, here the irregularity testifies to prevailing impulse seismogenic movements in the total slip. The amplitude difference of maximum values on neighboring sites, i.e. slip amount at a single impulse of movement, varies from 3 to 6 m; the average value is 4-4.5 m.

It is interesting that the TFF is not active at present time according to GPS data (Midi and Hugger, 2001), and the maximum right-lateral displacement along the fault is estimated as 2-3 mm/year for the north-western portion and almost zero for the south-eastern one. Apparently, the main slip along the fault was caused by strong movements (several meters) during earthquakes. During the intervals between the earthquakes the fault is blocked and is characterized by strain accumulation, which drops once upon several hundredsthousands years.

A fragment of the Talas-Fergana Fault "Kyldau" (Burtman et al., 1996) demonstrates a series of right-lateral displacements from 12 m for small gullies to 125 (25) m for watersheds (Figs. 32 and 33). In a pit section located in north-western slope of the watershed, displaced on 125 (25) m, in a fault zone a black soil of 0.3 m thickness overlays a brown soil of 0.6 m thickness. The last one overlays a clay and sand strata (Fig. 34). Contact surface of a brown soil and clay-sand layer is tilted on about 20 toward north-east. Radiocarbon dating gives an age 39624132 years В.Р. (point 11, table 1), which is most ancient from the investigated samples in this part of the Talas-Fergana Fault. Seismotectonic deformations in the Talas-Fergana Fault zone are in association with gravitation formations, probably seismically-induced.

Fig. 32. Topographic map of the Talas-Fergana Fault along the Kyldau River Valley (Burtman et al., 1996). Black line is a fault trace. Figures point location of the field stations 10 and 11.

Fig. 33. The Talas-Fergana Fault goes along the right slope of the Kyldau river valley – south-western slope of the Yangyzkyr ("Lonely") ridge and displaces the watershed right on tens of meters.

Fig. 34. Stratigraphic sections of pits excavated in different localities along the Talas-Fergana Fault by Burtman et al., 1996 (See also Table 2).

#### **5.2 Investigations in the Dzhilangach-Pchan interfluve**

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

this part of the Talas-Fergana Fault. Seismotectonic deformations in the Talas-Fergana Fault

Fig. 32. Topographic map of the Talas-Fergana Fault along the Kyldau River Valley (Burtman et al., 1996). Black line is a fault trace. Figures point location of the field stations 10 and 11.

Fig. 33. The Talas-Fergana Fault goes along the right slope of the Kyldau river valley – south-western slope of the Yangyzkyr ("Lonely") ridge and displaces the watershed right on

tens of meters.

zone are in association with gravitation formations, probably seismically-induced.

Southern most part of the Talas-Fergana Fault, where there were collected samples for determination of the absolute age (Burtman et al., 1996), includes a segment of the fault located between Dzhilangach and Pchan river basins (Fig. 35). Both in the north and in its middle parts the Talas-Fergana Fault looks as a lengthy up-slope facing scarp of 10-30 m width and down to 5 m depth (Fig. 36). In given locality there are clear evidences of a horizontal component along the fault: displacement of gullies (Fig. 37 and 38) and watersheds (Fig. 39) among them on a value of few tens of meters. Thus along the

Fig. 35. Topographic map of a fragment of the Talas-Fergana Fault in the Pchan-Dzhilangach interfluves (Burtman et al., 1996). Black line is a fault trace. Numbers are localities of pits where samples for absolute age were collected. Isoline interval – 200 m.

Fig. 36. On whole its length a line of the Talas-Fergana fault is expressed as lengthy upslope facing scarp (shown by pointers) which cuts any forms of relief. A region of the Dzhilangach pass.

Fig. 37. A region of the Dzhilangach pass. One can observe right-laeral displacement of gullies on 40-50 meters. Arrows placed in left slopes of mentioned gullies.

Fig. 36. On whole its length a line of the Talas-Fergana fault is expressed as lengthy upslope facing scarp (shown by pointers) which cuts any forms of relief. A region of the

Fig. 37. A region of the Dzhilangach pass. One can observe right-laeral displacement of

gullies on 40-50 meters. Arrows placed in left slopes of mentioned gullies.

Dzhilangach pass.

Dzhilangach river valley a value of horizontal displacement according to geomorphologic data is from 19 to 45 m (See Table 1), in Chitty-Western river basin – 40 m, in the Birguzy river basin – 27-35 m, in the Pchan river basin – from 21 to 90 m.

Fig. 38. Close view on a zone of the Talas-Fergana Fault. Left slope of the Dzhilangach river valley. One can observe right-lateral displacement of gullies. For a scale pay attention for a group of people in the center of photograph.

Fig. 39. Displacement of gully channels (solid line) and watersheds along the Talas-Fergana Fault line (dashed line) in a region of the Dzhilangach pass. "B" marks a watershed located higher a slope; "H" point its continuation down the slope. As a scale look on two shepherd tents higher and right of "H" in the center of photograph.

It is clear that such values of displacements cannot be attributed to only one earthquake. Indeed absolute dates (table 1) testifies on minimum eight earthquakes occurred in 15800, 4590, about 3955, about 3095, about 2635, 2230,1940, 1720 years ago.

In tie with mainly strike-slip movements along the fault, often in its zone there are formed specific deformation forms: so-called tectonic swells (Fig. 40), which are like a barrier ridges for consequent drainage and near-fault depressions – places of accumulations of fine material from adjacent slopes. In those depressions one has to conduct pit excavations for samples collection for absolute dating.

Fig. 40. A view toward north-west along the Talas-Fergana Fault. A region of the Semiz river mouth. Here the springs – tributaries o of the river flowing down toward south-west are dammed by the tectonic swells (shown by pointers) serving as barrier or "shatter" ridges for consequent drainage.
