A 3400-year-long paleoseismologic record of earthquakes on the southern segment of Anninghe fault on the southeastern margin of the Tibetan Plateau

    A 3400-year-long Paleoseismologic Record of Earthquakes on the southern segment of Anninghe Fault on the Southeastern Margin of the Tibetan Plateau Hu Wang, Yongkang Ran, Yanbao Li, Francisco Gomez, Lichun Chen PII: DOI: Reference:

S0040-1951(14)00230-3 doi: 10.1016/j.tecto.2014.04.040 TECTO 126294

To appear in:

Tectonophysics

Received date: Revised date: Accepted date:

18 March 2013 10 April 2014 27 April 2014

Please cite this article as: Wang, Hu, Ran, Yongkang, Li, Yanbao, Gomez, Francisco, Chen, Lichun, A 3400-year-long Paleoseismologic Record of Earthquakes on the southern segment of Anninghe Fault on the Southeastern Margin of the Tibetan Plateau, Tectonophysics (2014), doi: 10.1016/j.tecto.2014.04.040

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT A 3400-year-long Paleoseismologic Record of Earthquakes on the southern segment of Anninghe Fault on the Southeastern Margin of the Tibetan Plateau

Key Laboratory of Active Tectonics and Volcano, Institute of Geology, China

Earthquake Administration, Beijing 100029, China

Department of Geological Sciences, 101 Geological Science Building, University of

SC

b

RI P

a

T

Hu Wanga), Yongkang Ran a), Yanbao Lia), Francisco Gomezb), Lichun Chena)



MA NU

Missouri-Columbia, Columbia, MO 65211, USA

Corresponding author. Address: 19 Beituchengxilu Road, Beijing 100029, China.

ED

Tel./Fax: +86 10 62009213. E-mail address: [email protected] (Y.-K. Ran),

Abstract The Anninghe fault (ANHF) is an active left-lateral strike-slip fault along

PT

the southeastern margin of the Tibetan Plateau. Previous studies suggested the ANHF was divided into two segments, which herein we named the northern and southern

CE

segments respectively. Multiple trenches were excavated on the northern segment,

AC

revealing well-constrained paleoseismic events by radiocarbon ages. However, until now there is no paleoseismic result on the southern segment, which terminates at the Zemuhe fault (ZMHF) around Xichang where there were multiple historical records of large earthquakes. In this paper, we used high-resolution images for mapping fault traces on the southern segment of the ANHF and found a small depression. Through trenching in the depression, five paleoseismic events are identified and named E1 through E5 from youngest to oldest at 1750 AD-present, 1430-1870 AD, 940-1150 AD, 700-1000 AD, and 1400-500 BC respectively. Comparing with the historical record earthquakes around Xichang, we suggest the latest event E1 is associated with one of the 1850 AD and 1952 AD events, and event E2 is interpreted as the 1536 AD 1

ACCEPTED MANUSCRIPT earthquake, event E4 is possibly associated with the 814 AD earthquake. The average recurrence interval of earthquakes on the southern segment is about 600-800 yr, the

T

interval between E1 and E2 is 416 yr or 314 yr, and 386-596 yr between E2 and E3.

RI P

The interval between E3 and E4 is shorter at 126-336 yr, but much longer at 1314-2214 yr between E4 and E5. These surface-faulting events on the southern

SC

segment of the ANHF appear to be unevenly spaced in time. Furthermore, integrating the paleoseismic sequence of the northern segment of the ANHF, the two segments

MA NU

appear to be ruptured individually or interactively triggered within a narrow time range, or co-ruptured during one paleoseismic event, indicating the ANHF possibly shows a cascading rupturing behavior.

ED

Key words: The Anninghe fault (ANHF), paleoseismology, trenching, cascading rupturing behavior

PT

1 Introduction

The southeastern region of the Tibetan Plateau is one of the most intense areas in

CE

Mainland China on crustal deformation with frequent large earthquakes (Deng et al.,

AC

2003; Xu et al., 2003; Zhang et al., 2003). Historical records show that just in western Sichuan region, located at the junction of three active tectonic blocks -- the Sichuan-Yunnan, Bayan Har, and South China blocks (Fig. 1), there are M ≥ 7 earthquakes occurring on average every 10 to 15 years in this region (Ran et al., 2008a). Such intense earthquake activity has meant that the western Sichuan region has been of great interest to geoscientists, especially after seismic events like the 2001 MW7.8 Kunlun earthquake, 2008 Mw7.9 Wenchuan earthquake, and 2010 Ms7.1 Yushu earthquake occurred in rapid succession along the boundaries of these three active tectonic blocks (Fig. 1). Consequently, understanding a long-time-scale large earthquake behavior on the nature of crustal deformation along the southeastern 2

ACCEPTED MANUSCRIPT margin of the Tibetan Plateau is important for effectively implementing seismic hazard mitigation strategies in this populated region.

T

The Anninghe fault (ANHF) is a left-lateral strike-slip fault that connects to the

RI P

Xianshuihe fault (XSHF) in the north and terminates at the Zemuhe fault (ZMHF) in the south, and all these faults collectively compose a larger, more complex left-lateral

SC

fault system located along the eastern and northern boundaries of the Sichuan-Yunnan block in the southeastern area of the Tibetan Plateau (Fig. 1). From the end of early

MA NU

Pleistocene to the early of middle Pleistocene, the Sichuan-Yunnan block has been extruded in a southeastward direction due to deformation of the Tibetan Plateau, resulting from the collision of the Indian and Eurasian plates (Peltzer and Tapponnier,

ED

1988；Tapponnier et al., 1977, 1982，2001). Previous studies suggested the ANHF was divided into two segments around Mianning (Fig. 2). Pei et al. (1998) first divided the

PT

ANHF into two segments around Mianning based on differences of geometrical styles

CE

on the two fault segments. Wen (2000a) suggested that surface ruptures produced by the historical record earthquake of 1536 AD M71/2 propagated northward from

AC

Xichang and terminated around Mianning, and found the area around Mianning is an abnormal transition zone of gravitational and magnetic field, and proposed the ANHF was divided into two segments around Mianning. Herein, we named the two segments from north to south as the northern and southern segment (Fig. 2) and terminate at the ZMHF around Xichang where there were multiple historical records of 814 AD M7, 1536 AD M71/2, 1850 AD M71/2 and 1952 AD M63/4 earthquakes (Department of Earthquake Disaster Prevention, 1995). Presently, there have been several paleoseismic studies on the ZMHF (He and Ren，2003；Tian et al., 2008；Ren et al.,

2010；Wang et al., 2013), specifically, both Ren et al. (2010) and Wang et al. (2013) 3

ACCEPTED MANUSCRIPT showed deformation evidence of paleoseismic events that are associated with the historical record 814 AD and 1850 AD earthquakes based on multiple trenching and

T

detailed field investigations. However, whether the surface ruptures produced by these

RI P

two historical record earthquakes propagated northward to the southern segment of the ANHF or not is still uncertain, and this needs to be resolved when analyzing

SC

Holocene faulting behavior of the ANHF. Paleoseismic studies on the ANHF have been mostly conducted on its northern segment (Qian et al., 1990; Wen et al., 2000b,

MA NU

2007; Ran et al., 2008a). Ran et al. (2008a) excavated multiple trenches on the northern segment of the ANHF and measured ages of three recent paleoseismic events (1634-1811 a.BP, 1030-1050 a.BP and 280-550a.BP) from radiocarbon dating, and

ED

then estimated an average of recurrence interval of about 520-660 yr. However, for the southern segment of the ANHF, data on recurrence intervals of paleoseismic

PT

events and rupturing behavior are lacking, only Wen et al. (2007) gave some evidence of surface ruptures that were possibly associated with the historical record 1536 AD

CE

earthquake based on analysis of geomorphic mapping. Additionally, whether the

AC

surface ruptures revealed from trenching on the northern segment of the ANHF propagated southward to the southern segment or not has not been resolved either. Consequently, paleoseismic studies on the southern segment of the ANHF is crucial to explain scientific problems like segmentation of the ANHF, large earthquake behavior and fault interactions between the ANHF and ZMHF, all these above mentioned questions are important to assess risks of large earthquake hazard in the western region of Sichuan and understand crustal deformation along the southeastern margin of the Tibetan Plateau. Regions from Mianning to Xichang along the southern segment of the ANHF are highly populated (Fig. 2). Due to human activities and intense erosion, it is 4

ACCEPTED MANUSCRIPT challenging to find good sites on the fault segment that completely preserve evidence of paleoseismic events and favor for multiple radiocarbon dating. Previous studies

T

including us had attempted to conduct paleoseismic studies along the southern

RI P

segment of the ANHF but all finally failed due to the challenging environment with fast erosion and widespread modification of the landscape by farming. And now we

SC

use high-resolution images (possibly with 1 m resolution from Google Earth) for mapping fault traces of the southern segment of the ANHF and found a small

MA NU

depression located at the Yuehua section about 30 km north of Xichang (Fig. 2). This paper is mainly analyzing paleoseismic events and radiocarbon ages from multiple trenches in the depression, and integrating paleoseismic events revealed on the

ED

northern segment of the ANHF and historical record earthquakes around Xichang to decipher large earthquake behavior since late-Holocene along the ANHF.

PT

2 Late-Quaternary tectonic activity of the ANHF The ANHF connects to the XSHF in the north, and extends southward through

CE

Mianning, Lizhou, and terminates at the ZMHF around Xichang with a total length of

AC

about 150 km (Fig. 2). Presently, left-lateral slip rates of the ANHF have been mainly determined on the northern segment, Ran et al. (2008b) estimated a left-lateral slip rate of about 6.2 mm/yr since Late Holocene and 3.6-4.0 mm/yr since 10 ka BP based on aerial photo interpretations, total station surveys and multiple trenches at Zimakua on the northern segment. Xu et al. (2003) estimated a slip rate of about 6.5±1 mm/yr based on measurement of a gully offset on a terrace and its TL dating in south of Mianning. Additionally, Shen et al. (2005) give a 4±2 mm/yr slip rate on the ANHF from far-field GPS data, Wang et al. (2008) based on recent GPS data measured from more stations and re-estimate that the slip rate on the ANHF is 5.1±2.5 mm/yr. 5

ACCEPTED MANUSCRIPT However, until now there is no paleoseismic result on the southern segment, only there were some historical record earthquakes in this area. For example, the 1536 AD

T

earthquake possibly occurred on the southern segment; moreover, a smaller

RI P

earthquake occurred in 1952 AD with a magnitude of M63/4 near Shilong on the southern segment of the ANHF (Department of Earthquake Disaster Prevention,

SC

1995). So from the aforementioned slip rates and historical records of earthquakes, we suggest that tectonic activity along the ANHF has been quite intense during the late

MA NU

Quaternary.

3 Paleoseismic investigations at the Yuehua section 3.1 Site description

ED

To better assess the seismic hazard along the southern segment of the ANHF, and to investigate the frequency of earthquake recurrence on this fault segment, we

PT

have conducted detailed geologic and geomorphic investigations along the Yuehua section where documented the presence of a nearly NNW-striking fault valley that

CE

favors development of several depression catchments produced by blockage against

AC

the linear ridge due to slip on the fault segment (Fig. 3). When choosing a good trenching site at the Yuehua section, we mainly consider depression catchment that favors continuous sedimentation and without human modification. According to field reconnaissance and investigations with local villagers, we chose a depression that shows a triangular shape with about 60 m in length and 30 m in width (Fig. 3 and 4). The main sediment in the depression might be transported from an inflowing gully G2 and rainfall catchment from the local hill slope against the linear ridge. Actually, trench Tc3 located near the inflowing area from the gully G2, reveals multiple angular types of gravel with different sizes deposited in disorder and shows no bedding characteristics, possibly indicating a quick sedimentation of debris flow. The modern 6

ACCEPTED MANUSCRIPT alluvial fan developed from the gully G1 might deposit near the northern area of the depression. Therefore, in order to avoid disturbance of debris or alluvial fan on

T

identification or sampling for paleoseismic events, we opened five trenches near the

RI P

center of the depression (28°10'30.31444"N，102°11'23.83273"E) (Supplementary Fig.

SC

S1). 3.2 Stratigraphy

MA NU

We opened five trenches in the depression at the Yuehua section across the ANHF (Fig. 4 and Supplementary Fig. S1). Aiming to reveal structures of faults and paleoseismic evidence, four trenches (Tc1, Tc2, Tc3 and Tc4) are nearly-perpendicular excavated across the fault, and to facilitate spatial correlation of stratigraphic units,

ED

trench Tc5 is paralleled to strike of the fault. Specifically, trench Tc3 is to reveal appropriate range of debris from the gully G2 and facilitate to choose other trench

PT

sites within the depression. Tc2 is the longest trench across the depression to constrain

CE

location of the fault trace, considering ground water mainly from the gully G2, trench Tc2 temporarily played a role in water shutoff for other deeper excavations in the

AC

northern area of the depression. However due to shallow depth of the ground water table and quick infiltration, most depth of Tc2 is no more than 2 m. Due to effective pumping, trench Tc1 was nearly 7 m deep and revealed most deformation evidence of paleoseismic events and stratigraphic units. During excavation, each wall was nearly vertically excavated and then systematically cleaned, gridded in squares of mainly 1 m by 1 m, photographed and then printed for detailed field mapping. Some trenches collapsed quickly due to quick infiltration, we mainly use trench Tc1 and local parts of trenches Tc2 , Tc4 and Tc5 to identify paleoseismic evidence. The stratigraphy in the depression is divided into six units, including 15 sub-units and summarized from oldest to youngest in Table 1 and Fig. 5 (letter “U” denotes strata units, seen in Fig. 6 7

ACCEPTED MANUSCRIPT and 7). 3.3 Evidence of faulting events

T

From trenching along the Yuehua section, five paleoseismic events are identified

RI P

and named E5 through E1 from oldest to youngest. Among various indicators for recognizing paleoearthquakes in unconsolidated sediments in a strike-slip fault setting,

SC

the best and clearest evidence for fault rupturing of the ground surface includes scarp formation, scarp-derived colluvium, in-filled and void fissures, and sand blows (Sieh,

MA NU

1978; Weldon et al., 1996; Fumal et al., 2002; Liu-Zeng et al., 2007). Upward fault terminations are generally most effective in identifying the latest faulting event, because subsequent ruptures often follow the same plane (McCalpin et al., 2009).

ED

Based on above identifying indicators for strike-slip faults, we show stratigraphic evidence for the five paleoearthquakes from oldest to youngest at the site as follows

Event E5

PT

(Fig. 5):

CE

Deformation evidence of older paleoseismic events are poorly unrevealed due to

AC

the limit of depth of trenching, for the event E1, only the northern wall of trench Tc1 revealed the deformation evidence. Both of U1-1 and U1-2 were offset (Fig. 6 and Supplementary Fig. S2). U2-1 contains multiple granular gravel and shows a wedged shape and unconformably overlies U1-2, we interpret U2-1 as a scarp-derived colluvial deposit following an earthquake. Based on above analysis, we suggest event E5 occurred between sedimentation of U1-2 and U2-1. Event E4 Both two walls of trench Tc1 revealed deformation evidence of event E4. On the northern wall, units of U2-2 and U2-3 collectively show a partial ‘Z’-shaped deformation, specifically, some parts of U2-3 consisting of fine sand locally intruded 8

ACCEPTED MANUSCRIPT into U2-4 comprising peat layer. U3-1 is composed of peat containing multiple small gravel and wood, shows wedged shape and unconformably overlies U2-4, interpreted

T

as a scarp-derived colluvial deposit following an earthquake (Fig. 6 and

RI P

Supplementary Fig. S3a). Furthermore, the southern wall of trench Tc1 also showed similar evidence of event E4. Sand unit of U2-3 deformed upward, and shows a

SC

wedged shape containing multiple gravel and unconformably overlies units of U2-3 and U2-4, specifically, stratigraphic contact between U2-3 and U3-1 is evident (Fig.

MA NU

7 and Supplementary Fig. S3b). All of the aforementioned evidence suggests that event E4 occurred before deposition of U3-1 and after that of U2-4. Event E3

ED

Deformation evidence produced by event E3 is well revealed in the trench Tc1. On the northern wall of Tc1, U1-1 was offset and overlain by a wedge-shaped U4-1

PT

containing multiple gravels and is interpreted as scarp-derived colluvial deposit (Fig. 6 and Supplementary Fig. S4). On the southern wall of Tc1, U1-1 and U3-1 was offset

CE

and overlain by a wedge-shaped U4-1, which similarly interpreted as scarp-derived

AC

colluvial deposit following an earthquake (Fig. 7). We suggest that event E3 occurred after deposition of U3-2 and before that of U4-1. Event E2

The northern wall of trench Tc1 revealed evidence of deformation produced by event E2. A fault offset U1-1 and produced a fault scarp that was unconformably overlain by a wedged unit U5-1 consisting of gravel (Fig. 6 and 8b). Moreover, the northern wall of trench Tc2 shows U4-3 appeared to deform with a vertical warping of about 0.5 m (Fig. 8a). Based on deformation of the units, we suggest that event E2 occurred before deposition of U5-1 and after that of U4-3. Furthermore, as secondary paleoseismic evidence, we observed liquefaction in the base and middle area of U4-3 9

ACCEPTED MANUSCRIPT in trench Tc5 (shown around the red pen in Fig. 9), which might suggest there were paleoearthquakes occurred after deposition of U4-3. Even if we do not know whether

T

this liquefaction was produced by the rupture of the southern segment of the ANHF,

RI P

however taking that corresponding deformation evidence of faulting and scarp-derived colluvial deposit revealed in trench Tc1 and Tc2, we are inclined to

SC

suggest these deformation revealed in trenches was produced by event E2. Event E1

MA NU

Deformation evidence produced by the latest event E1 appeared to be a little ambiguous. The southern wall of trench Tc4 shows event E1 ruptured U1-1 near subsurface and produced a fault scarp that was overlain by a scarp-derived deposit

ED

U6-1 (Fig. 10). Stratigraphy of U1-1 within the two sides of the fault is almost identical consisting of alluvial-fan gravel. The western side of U1-1 shows bedding

PT

with an apparent dip angle of about 17°, but the eastern side of U1-1 shows a mixed and loose configuration, which indicates U1-1 near the sub-surface was deformed by

CE

faulting. Furthermore, the northern wall of trench Tc1 also revealed similar

AC

deformation, and the fault nearly ruptured the sub-surface and produced a wedge-shaped deposit U6-1 that mainly consisting of gravel (Fig. 6 and 8b). Combined with these possible evidence in trench Tc1 and Tc4, we suggest the latest event E1 occurred between depositional age of U5-2 and U6-1. Based on stratigraphic deformation evidence, we use a restoration model (Fig. 11) to show deformation associated with the five paleoseismic events. Evidence of event E5 is reliably revealed. We observed faulting deformation, and wedge-shaped scarp-derived deposits on the northern wall of trench Tc1. Events E4 and E3 are also well revealed in trenches with similar paleoseismic evidence. Event E2 is revealed by faulting and scarp-derived colluvial deposit in trench Tc1 and upward warping of 10

ACCEPTED MANUSCRIPT U4-3 in trench Tc2, furthermore, liquefaction revealed in U4-3 in trench Tc5 appears to be secondary evidence for this paleoearthquake. The latest event E1 is a little

T

ambiguous; however we possibly observed a fault ruptures U1-1 near sub-surface

RI P

and develops scarp-derived colluvial deposit in trench Tc1 and Tc4. 3.4 Radiocarbon dating of paleoearthquakes

SC

Numerous charcoal and wood samples were found in trenches, 18 samples including 6 charcoals and 12 woods were sent to Beta Analytic Inc. of America for

MA NU

radiocarbon dating. All charcoal samples were dated by accelerometer mass spectrometer (AMS dating), wood samples, according to measured requirement, some used radiometric dating (Table 2). All ages reported herein are two-sigma

ED

(95.4% confidence limits) calendric ages calibrated with the OxCal 3.10 program (Bronk-Ramsey, 1998; http://c14.arch.ox.ac.uk/embed.php?File=oxcal.html). Most

PT

radiocarbon dates from trenches are generally in correct stratigraphic order, such that younger samples overlie older samples. Nearly ages of all samples are younger than

CE

the age of Y055 that is sampled in U1-2 and yielded an age of 1440-1290 BC, this

AC

sample is a fingernail-sized granular charcoal and is reliable to represent deposition age of U1-2, moreover, another fingernail-sized granular charcoal Y057 was collected in U2-1 and yielded an age of 600-400 BC, which shows a correct stratigraphic sequence. Only a charcoal sample Y104 collected in U5-2 in the northern wall of trench Tc1 yielded an age of 1980-1750 BC and is the oldest than other samples. This sample is a tiny piece of charcoal and possibly interpreted to be a reworked sample and does not represent deposition age of U5-2. Additionally, we consider some wood samples collected in U3-1 of trench Tc1 to be possibly reworked. Wood samples YM004, YM015 and YM011 in U3-1 yielded ages of 330-570 AD, 430-670 AD and 630-720 AD respectively. However, a wood sample 11

ACCEPTED MANUSCRIPT YM021 in U2-4 yielded an age of 660-780 AD and is younger than that of three samples in U3-1, though the age of YM011 partly overlaps that of YM021, to be

T

more reliable, we cancel the three samples in U3-1 and use another wood sample

RI P

YM018 in the overlain unit of U3-2 yielded an age of 940-1030 AD to constrain occurrence ages of paleoseismic events. A wood sample YM019 collected in U4-2 in

SC

the southern wall of trench Tc1 yielded an age of 600-780 AD that is older than that of the wood sample YM018 in U3-2, we interpret the sample YM019 to be possibly

MA NU

reworked and do not consider it. Samples with much closer dating ages within error of several tens of years from the same stratigraphic unit, we put dating data in a phase correction when establishing a sequence correction during processing of

ED

OxCal 3.10 program (Dawson et al. 2003). For example, we put samples (Y014, Y018, YM013, and YM020) from U4-2 into a phase. After deleting possibly

PT

reworked samples of YM015, YM004, YM011, YM019, YM016, Y104 and AM007,

12).

CE

we use OxCal 3.10 program to further constrain ages of paleoseismic events (Fig.

AC

As previously discussed, event E5 occurred after deposition of the sample Y055 in U1-2, which has calibrated ages of 1440-1290 BC (2σ). We can also constrain that the age of event E5 occurred before deposition of the sample Y057 in U2-1 that has calibrated ages of 600-400 BC (2σ). Using OxCal for further analysis, the age of event E5 is well constrained between 1400-500 BC (Fig. 12). In keeping with this kind of scenario, ages of all events can be constrained by dating samples from scarp-derived colluvial deposits and faulted or disturbed units below. Event E4 occurred before deposition of U3-1 and after that of U2-4, and multiple wood samples constrain the age of event E4 occurred between 700-1000 AD. Event E3 occurred before deposition of U4-1 and after that of U3-2, and events E4 and E5 12

ACCEPTED MANUSCRIPT respectively occurred before deposition of U5-1, U6-1 and after that of U4-3, U5-2. Event E3 is constrained by samples YM018 and YM005 from U3-2 and U4-1

T

respectively and dated between 940-1150 AD. Event E2 constrained by samples

RI P

AM010 in U5-1 and Y002 in U4-3 and dated between 1430-1870 AD. Event E1 is just constrained on its lower bound by the sample AM010 and dated as 1750

SC

AD-present. 4 Discussions

MA NU

4.1 Comparison with historical records of large earthquakes around Xichang

Based on multiple statistics on surface ruptures and corresponding magnitudes of

ED

earthquakes summarized by Wells and Coppersmith (1994), there possibly would be surface ruptures with magnitude greater than M5.6 for strike slip faults. Moreover,

PT

taking the summarized equation for strike-slip fault (M=5.16( ± 0.13)+1.12( ±

CE

0.08))*lg(SRL), M denotes magnitude of an earthquake, SRL represents the corresponding surface rupture length) concluded by Wells and Coppersmith (1994),

AC

there would be surface ruptures with a length of about 16-45 km for a magnitude of M63/4 earthquake. According to the historical records, the latest two historical record events occurred around Xichang where the ANHF intersects with the ZMHF are 1850 AD and 1952 AD earthquakes (Department of Earthquake Disaster Prevention, 1995). For the 1850 AD event, though many researchers considered the ZMHF as the responsible structure from field surface-rupture investigations and radiocarbon dating results combined with trenching (He and Ren, 2003; Tian et al., 2008；Ren et al.,

2010；Wang et al., 2013), whether the surface ruptures propagated northward to the southern segment of the ANHF is still unknown. The Yuehua trenching site is just 13

ACCEPTED MANUSCRIPT located about 30 km north of Xichang (Fig. 2). Therefore, it is possible that the deformation revealed in the trenches was produced by the 1850 AD M71/2 event.

T

Alternatively, for the 1952 AD event, the historical record showed that the possible

RI P

epicenter was located near Shilong along the southern segment of the ANHF (Fig. 2; Department of Earthquake Disaster Prevention, 1995). Until now, there are no

SC

corresponding studies on this event, even whether it produced surface ruptures or not is questionable. However, the Yuehua trenching site is about 28 km south of the

MA NU

epicenter area of Shilong and could be in the estimated range of the possible surface rupturing produced by the 1952 AD event according to the aforementioned formulas. Consequently, for the latest event E1 revealed in the trenches, we propose that there

ED

are two possibilities associated with the two historical record earthquakes and needs further detailed studies to distinguish which historical event is associated with the

PT

latest event E1.

For the historical record 1536 AD earthquake, the epicenter area was suggested

CE

to be located around the southern segment of the ANHF (Department of Earthquake

AC

Disaster Prevention, 1995), however due to intense rainfall as much as 800-1000 mm per year in this area (Wen et al., 2007), surface ruptures might be easily eroded away; besides, human activities modified the areas near faults to be farmland, both resulting fault traces to be ambiguous. Only Wen et al. (2007) found some possible evidence of the surface ruptures with left lateral offsets of greater than 1 m and suggested it associated with the historical record 1536 AD earthquake. Herein, based on the age constraint of 1430-1870 AD for event E2 from trenching, we are more likely to suggest it associated with the historical record 1536 AD earthquake. For the historical record 814 AD earthquake, due to long years after its occurrence, we just know that the epicenter area for this event may be located in south 14

ACCEPTED MANUSCRIPT of Xichang (Department of Earthquake Disaster Prevention, 1995), Wang et al. (2013) revealed the deformation evidence for this historical event based on multiple trenches

T

on the northern section of ZMHF and radiocarbon dating. The event E4 is constrained

RI P

between 700-1000 AD from radiocarbon dating for units U2-4 and U3-2, actually, this event could have been narrower constrained by units U2-4 and U3-1, so the true age

SC

for this event may be closer to the older end of the constrained range. Herein, from trenching results and historical records around Xichang, we suggest that event E4 is

MA NU

possibly associated with the historical record 814 AD event, indicating a possible co-rupturing event on the southern segment of the ANHF and ZMHF. For the other two events E3 and E5, there is no historical record earthquake around Xichang during

ED

the constrained age ranges of the two events. For event E5, it is too old and suggested to be a prehistoric earthquake. And for event E3, we suggest it may be lacking in

PT

historical record.

the ANHF

CE

4.2 Characteristics of paleoearthquake sequence on the southern segment of

AC

Five events are named E5 through E1 from oldest to youngest and further constrained at 1400-500 BC, 814 AD, 940-1150 AD, 1536 AD and 1850 AD (or 1952 AD). The average recurrence interval of earthquakes on the southern segment of the ANHF is about 600-800 yr, the interval between E1 and E2 is 416 yr or 314 yr, and 386-596 yr between E2 and E3. The interval between E3 and E4 is shorter at 126-336 yr, but much longer at 1314-2214 yr between E4 and E5. These surface-faulting events on the southern segment of the ANHF appear to be unevenly spaced in time. 4.3 Comparison with large earthquakes between the southern and northern segments of the ANHF Presently, there are some paleoseismic results on the northern segment of the 15

ACCEPTED MANUSCRIPT ANHF. Ran et al. (2008a) excavated multiple trenches at Zimakua and constrain ages of three recent paleoseismic events at 1634-1811 a BP, 1030-1050 a BP and 280-550

T

a BP with each coseismic left-lateral displacement of about 3 m estimated from

RI P

measurement of small terraces. Moreover, they excavated some trenches at Ganhaizi on the northern segment and revealed four events at 1536 AD M71/2 event,

SC

1768-1826 a BP, 2755-4108 a BP, 4108-6593 a BP. Herein, we calibrated these ages into calendar ages (Fig. 13). The latest event revealed at Ganhaizi on the northern

MA NU

segment of the Anninghe fault was associated with the historical record 1536 AD earthquake, the occurrence age of the latest event revealed at Zimakua was constrained between 1400-1670 AD, which is very close to that revealed in the

ED

Yuehua trenches and possibly interpreted as the 1536 AD earthquake. According to formulas summarized by Wells and Coppersmith (1994), when magnitude is 7.5, a

PT

strike-slip faulting earthquake may produce surface ruptures of about 69-237 km, with an average length of about 153 km. The ANHF is about 150 km long that is

CE

well consistent with the calculated length of surface ruptures, so we suggest that the

AC

historical record 1536 AD event may produce surface ruptures on both two segments of the ANHF. However, some paleoseismic events occurred individually on the two segments. For example, at Zimakua and Ganhaizi on the northern segment, an event constrained between 124-316 AD may be well corresponded but there is no deformation evidence on the southern segment in that age range. Additionally, the latest event revealed at Yuehua on the southern segment, event E5 has not been identified on the northern segment. Furthermore, some events show very close occurrence ages. An event constrained at 900-920 AD at Zimakua is close to the events constrained at 940-1130 AD or 814 AD revealed in Yuehua trench. Comparing paleoseismic sequences between the northern and southern segments of the ANHF, 16

ACCEPTED MANUSCRIPT the two segments appear to be ruptured individually or interactively triggered within a narrow time range, or co-ruptured produced by one earthquake, indicating the

T

ANHF possibly shows a cascading rupturing behavior.

RI P

5 Conclusions

Based on the detailed paleoseismological approach, multiple radiocarbon dates,

SC

and geomorphic interpretation at the Yuehua section, we first give paleoseismic documentation on the southern segment of the ANHF along the southeastern

MA NU

boundary of the Tibetan Plateau. Five events occurred during the past 3400 years and appear to be unevenly spaced in time. Comparing paleoseismic sequence between the northern and southern segments of the ANHF, the two segments appear to be ruptured

ED

individually or interactively triggered within a narrow time range, or co-ruptured produced by one earthquake, indicating the ANHF possibly shows a cascading

Acknowledgment

PT

rupturing behavior.

CE

This work has been funded by a Special Fund from China Earthquake

AC

Administration (Grant No. 200808016) and China Active-Fault Survey Project--The South-North Seismic Zone (201108011). We would like to express great thanks to William J. Cochran for his enthusiastic help in improving the English of the manuscript. Thanks for China Scholarship Council and China Earthquake Administration for sponsoring the first author to study in University of Missouri-Columbia for one year.

References Bronk-Ramsey, C., 1998. Radiocarbon calibration and analysis of stratigraphy: the OxCal program. Radiocarbon 37 (2), 425-430. 17

ACCEPTED MANUSCRIPT Dawson, T.E., McGill, S.F., Rockwell, T.K., 2003. Irregular recurrence of paleoearthquakes along the central Garlock fault near El Paso Peaks, California.

T

Journal of Geophysical Research 108 (B7), 2356-2385.

RI P

Deng, Q.D., Zhang, P.Z., Ran, Y.K., Yang, X.P., Min, W., Chu, Q.Z., 2003. Basic characteristics of active tectonics of China. Science China Earth Sciences 46 (4),

SC

356-372.

Deng, Q.D., 2007. Map of Active Tectonics in China. Seismological Press, Beijing.

MA NU

Deparment of Earthquake Disaster Prevention, China Earthquake Administration (DEDP, CEA), 1995. The catalogue of Chinese historical strong earthquakes, 514 pp., Seismological Press, Beijing, China.

ED

Fumal, T.E., Weldon, R.J., Biasi, G.P., Dawson, T.E., Seitz, G.G., Frost, W.T., Schwartz, D.P., 2002. Evidence for Large Earthquakes on the San Andreas Fault

PT

at the Wrightwood, California, Paleoseismic Site: A.D. 500 to Present. Bulletin of the Seismological Society of America 92 (7), 2726-2760.

CE

He, H.L., Ren, J.W., 2003. Holocene earthquakes on the Zemuhe Fault in

AC

Southwestern China. Annals of Geophysics 46 (5), 1035-1051. Liu-Zeng, J., Klinger, Y., Xu, X.W., Lasserre, C., Chen, G.H., Chen, W.B., Tapponnier, P., Zhang, B., 2007. Millennial Recurrence of Large Earthquakes on the Haiyuan Fault near Songshan, Gansu Province, China. Bulletin of the Seismological Society of America 97 (1B), 14-34. McCalpin, J.P., 2009. Paleoseismology. 473 pp., Academic Press, San Diego, California, USA. Pei, X.Y., Wang, X.M., Zhang, C.G., 1998. Basic character of segmentation of the Quaternary movement on the Anninghe fault. Earthquake Research in Sichuan 4, 52-61 (in Chinese with English abstract). 18

ACCEPTED MANUSCRIPT Peltzer, G., Tapponnier, P., 1988. Formation and evolution of strike-slip faults, rifts, and basins during the India-Asia collision: an experiment approach. Journal of

T

Geophysical Research 93 (B12), 15085-15117.

RI P

Qian, H., Wu, X.G., Ma, S.H., Cai, C.X., Tian, H., 1990. Prehistorical earthquakes on the north segment of the Anninghe fault and their significance to seismological

SC

research. Earthquake Research in China 6 (4), 43-49 (in Chinese with English abstract).

MA NU

Ran, Y.K., Chen, L.C., Cheng, J.W., Gong, H.L., 2008a. Late Quaternary surface deformation and rupture behavior of strong earthquake on the segment north of Mianning of the Anninghe fault. Science China Earth Sciences 51 (9),

ED

1224-1237.

Ran, Y.K., Cheng, J.W., Gong, H.L., Chen, L.C., 2008b. Late Quaternary geomorphic

PT

deformation and displacement rates of the Anninghe fault around Zimakua. Seismology and Geology 30 (1), 86-98 (in Chinese with English abstract).

CE

Ren, Z.K., Lin, A.M., Rao, G., 2010. Late Pleistocene-Holocene activity of the

AC

Zemuhe Fault on the southeastern margin of the Tibetan Plateau. Tectonophysics 495 (3-4), 324-336. Shen, Z.K., Lu, J.N., Wang, M., Burgmann, R., 2005. Contempory crustal deformation around the southeast borderland of the Tibetan Plateau. Journal of Geophysical Research 110 (B11409). doi:10.1029/2004JB003421. Sieh, K.E., 1978. Prehistoric large earthquake produced by slip on the San Andreas Fault at Pallett Creek, California. Journal of Geophysical Research 83 (B8), 3907-3939. Tapponnier, P., Molnar, P., 1977. Active faulting and tectonics in China. Journal of Geophysical Research 82 (20), 2905-2930. 19

ACCEPTED MANUSCRIPT Tapponnier, P., Peltzer, G., Le Dain, A.Y., Armijo, R., Cobbold, P., 1982. Propagating extrusion tectonics in Asia: New insights from simple experiments with

T

plasticine. Geology 10 (12), 611-616.

RI P

Tapponnier, P., Xu, Z.Q., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G., Yang, J.S., 2001. Oblique Stepwise Rise and Growth of the Tibet Plateau. Science 294

SC

(5547), 1671-1677.

Tian, Q.J., Ren, Z.K., Zhang, J.L., 2008. Study of paleoearthquakes by combined

MA NU

trench on Zemuhe fault around Daqingliangzi, Xichang, Sichuan. Seismology and Geology 30 (2), 400-411 (in Chinese with English abstract). Wang, H., Ran, Y.K., Li, Y.B., Gomez, F., Chen, L.C., 2013. Holocene

ED

paleoseismologic record of earthquakes on the Zemuhe fault on the southeastern margin of the Tibetan Plateau. Geophysical Journal International.

PT

Doi: 10.1093/gji/ggs095.

Wang, Y.Z., Wang, E.N., Shen, Z.K., Wang, M., Gan, W.J., Qiao, X.J., Meng, G.J., Li,

CE

T.M., Tao, W., Yang, Y.L., Cheng, J., Li, P., 2008. GPS-constrained inversion of

AC

present-day slip rates along the Sichuan-Yunnan region, China. Science China Earth Sciences 51(9), 1267-1283. Weldon, R.J., McCalpin, J.P., Rockwell, T.K., 1996. Paleoseismology of strike-slip tectonic environments, in Paleoseismology, J. McCalpin (Editor), Academic Press, San Diego, 271-329. Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement. Bulletin of the Seismological Society of America 84 (4), 974-1002. Wen,

X.Z.,

2000a.

Character

of

rupture

segmentation

on

the

Xianshuihe-Anninghe-Zemuhe fault zone, western Sichuan. Seismology and 20

ACCEPTED MANUSCRIPT Geology 22 (3), 239-249 (in Chinese with English abstract). Wen, X.Z., Du, P.S., Long, D.X., 2000b. New evidence of paleoearthquakes and date

T

of the latest event on the Xiaoxiangling mountain segment of the Anninghe

RI P

fault zone. Seismology and Geology 22 (1), 1-8 (in Chinese with English abstract).

SC

Wen, X.Z., Ma, S.L., Lei, X.L., Yasuto, N., Tsutomu, K., Chen, Q., 2007. Newly found surface rupture remains of large historical earthquakes on and near the

MA NU

transition segment of the Anninghe and Zemuhe fault zones, western Sichuan, China. Seismology and Geology 29 (4), 826-833 (in Chinese with English abstract).

ED

Xu, X.W., Wen, X.Z., Zheng, R.Z., Ma, W.T., Song, F.M., Yu, G.H., 2003. Newly tectonic deformation and its dynamic source of the Sichuan-Yunnan Block.

PT

Science China Earth Sciences 33 (S1), 151-162 (in Chinese). Zhang, P.Z., Deng, Q.D., Zhang, G.M., Ma, J., Gan, W.J., Min, W., Mao, F.Y., Wang,

CE

Q., 2003. China continental strong earthquake activities and active blocks,

AC

Science China Earth Sciences 33 (S1), 12-20 (in Chinese).

21

T

ACCEPTED MANUSCRIPT

RI P

Figure captions

SC

Fig.1. Tectonic and topographic map of the eastern Tibetan Plateau. The white rectangle shows the study area of the ANHF. Three red stars represent epicenters of

MA NU

the 2001 MW7.8 Kunlun earthquake, 2008 MW7.9 Wenchuan earthquake, and 2010 MS7.1 Yushu earthquake. Red small circles show epicenters of earlier documented (historical and instrumental) earthquakes. White small rectangles denote cities. KLF,

ED

Kunlun Fault, GZ-YSF, Ganzi-Yushu Fault, XSHF, Xianshuihe Fault, LMSF, Longmenshan Fault, ANHF, Anninghe Fault, ZMHF, Zemuhe Fault, DLSF,

PT

Daliangshan Fault, XJF, Xiaojiang Fault, HHF, Honghe Fault, JSJF, Jinshajiang Fault. The locations of active faults and earthquake epicenters are from a tectonic activity

AC

CE

map of China (Deng et al., 2007).

Fig. 2. Distribution of the ANHF and the northern section of ZMHF. Distribution of the ANHF are modified from Ran et al. (2008a). Red lines denote faults, grey shadows show areas of small basins with high population, red open circles represent possible epicenters of historical record earthquakes and numbers beside circles are corresponding occurrence ages and magnitudes, small green rectangles show locations of paleoseismic trench sites from Ran et al. (2008a). The two small yellow rectangles represent coseismic displacement produced by 1536AD earthquake from Wen et al. (2007). The ANHF was divided into two segments around Mianning, the blue rectangle is the trenching site at Yuehua section located about 30 km north of 22

ACCEPTED MANUSCRIPT Xichang.

T

Fig. 3. Photo and high-resolution image (GoogleEarth) around the trenching site at

RI P

Yuehua section. (a) Red arrows show fault traces of the ANHF, a white open circle denotes a brown horse with about 1.5 m tall，the trenches were excavated in the

SC

depression. (b) A high-resolution image (GoogleEarth) shows the trenching site and

MA NU

its simple interpretation. Red arrows show the ANHF. The depression catchment is possibly transported from the gully G2 and rainfall catchment from the local hill slope against the linear ridge, this process is schematically shown with yellow lines.

ED

Fig.4. Surveyed landforms around trenching site at the Yuehua section. The landform

CE

interval is 1 m.

PT

was measured by GPS RTK. Five trenches were excavated in the depression. Contour

Fig. 5. Stratigraphic units revealed from trenching. The right column summarizes

AC

evidence of paleoseismic events in trenches and the related figures.

Fig.6. Map of the northern wall of trench Tc1 and its interpretation. Red lines represent faults, and black lines show stratigraphic contacts, specifically while in dashed, U denotes units. Four yellow rectangles show pictures of Fig. 8b, S2, S3a and S4, respectively. Blues lines with numbers show sampling locations, labeled by sample number and corresponding corrected radiocarbon ages (2σ). Letters ‘Y’ and ‘YM’ represent charcoal and wood samples. All samples were processed in Beta Analytic Inc.

23

ACCEPTED MANUSCRIPT Fig.7. Map of the southern wall of trench Tc1 and its interpretation. Red lines represent faults, and black lines show stratigraphic contacts, specifically while in

T

dashed, U denotes units. A yellow rectangle shows pictures of Fig. S3b. Blues lines

RI P

with numbers show sampling locations, labeled by sample number and corresponding corrected radiocarbon ages (2σ). Letters ‘YM’ represent wood samples. All samples

SC

were processed in Beta Analytic Inc.

MA NU

Fig.8. Close-up views of deformation revealed in trench Tc1 and Tc2. (a) U4-3 shows upward warping with about 0.5 m, a dashed red line denotes inferred fault. The yellow line with number shows sampling locations and corresponding corrected

ED

radiocarbon ages (2σ). (b) The northern wall of trench Tc1 shows two branch faults offset U1-1 respectively and produced two scarps that were overlain by scarp-derived

PT

units U5-1 and U6-1, which are interpreted to be followed by two paleoseismic

CE

events.

AC

Fig.9. Map of the local western wall of trench Tc5 and its interpretation. White lines show stratigraphic contacts. The base and middle area of U4-3 reveals liquefaction, especially intense near the red pen. U denotes stratigraphic unit. The inset map shows the layout of all trenches; the thick blue line in the map denotes location of this wall.

Fig.10. Local enlarged photo of the southern wall of trench Tc4. This figure shows the latest event E1 ruptured U1-1 near subsurface and produced a fault scarp, which was overlain by a scarp-derived deposit U6-1. Areas of U1-1 on the right of the fault show bedding characteristics with an apparent dip angle of 17° (shown in dashed white lines). However, the other areas of U1-1 on the left of the fault show a mixed and 24

ACCEPTED MANUSCRIPT loose configuration. The inset map shows the layout of all trenches; the thick blue line

T

in the map denotes location of this wall.

RI P

Fig.11. Schematic restoration of stratigraphy on the northern wall of trench Tc1. (a) Interpretation of the northern wall of trench Tc1. (b) Scarp-derived deposit U6-1

SC

developed. (c) Event E1 occurred and produced a fault scarp. (d) U5-2 developed. (e) Scarp-derived deposit U5-1 developed. (f) Event E2 occurred and produced a fault

MA NU

scarp. (g) Units of U4-2 and U4-3 developed. (h) U4-1 developed as a scarp-derived colluvial deposit after Event E3. (i) Event E3 occurred. (j) Units of U3-1 and U3-2 deposited. (k) Event E4 occurred and offset all units, especially the sand of U2-3

ED

deformed with a ‘Z’ shape. (l) Units of U2-2, U2-3 and U2-4 deposited continuously. (m) Scarp-derived deposit U2-1 developed after event E5. (n) Event E5 occurred and

PT

offset U1-2 and produced a fault scarp. (o) A possible configuration before deformation of U1-2. Due to depth of trenching, peat units and deformation evidence

CE

of older events in the depression may have not been completely revealed. The black

AC

question mark denotes those unrevealed older units, and the yellow question marks represent unrevealed portion of U1-2.

Fig.12. Results of OxCal analysis of radiocarbon dates from the trenching site. Lines with no fill are prior probability distributions, and solid curves are posterior distributions after OxCal analysis. Phases are summed probability for units with multiple radiocarbon dates. Number after each radiocarbon date is an agreement index, indicating the extent of overlap between prior and posterior distribution. Overall agreement index for this model is 97.5%. Red vertical lines represent the timing of historical record earthquakes around Xichang, solid ones suggest reliable correlation 25

ACCEPTED MANUSCRIPT between paleoseismic events and historical earthquakes while dashed ones show uncertainties. Red curves denote five events with related time ranges, the youngest

T

two events are associated with the historical 1850 AD or 1952 AD event, and 1536

RI P

AD event. Event E4 is associated with historical 814 AD earthquake.

SC

Fig.13. Paleoseismic sequence from trenching at different sites along the ANHF and compared with historical record earthquakes. The thick black horizontal dashed line

MA NU

denotes approximate boundary between the northern and southern segment of the ANHF around Mianning. Grey rectangles represent ranges of age constraints of Paleoseismic events, red rectangles show possible comparison of paleoseismic events

ED

on the two segments of the ANHF. Paleoseismic results at Zimakua and Ganhaizi are modified from Ran et al. (2008a). The most recent earthquake revealed at Yuehua on

PT

the southern segment did not rupture the northern segment, but for the second oldest one that is associated with the 1536 AD earthquake possibly ruptured the whole

CE

ANHF. For the third or fourth oldest events, there was at most one event could

AC

correspond with the paleoseismic event on the northern segment, however we are not sure which one of the two events may rupture the northern segment and are illustrated with dashed red lines. The Yuehua trench does not show evidence of plaeoseismic events during 124-316 AD when an earthquake occurred along the northern segment. For much older paleoseismic events, their timing are not well constrained.

26

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 1

27

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 2

28

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 3

29

AC

Figure 4

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

30

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 5

31

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 6

32

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 7

33

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 8

34

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

Figure 9

35

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

Figure 10

36

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 11

37

AC

CE

PT

ED

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 12

38

MA NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

Figure 13

39

ACCEPTED MANUSCRIPT Table 1

U2-3 U2-4 U3-1 U3-2 U4-1

U4-2

U4-3 U5-1 U5-2 U6-1 U6-2

RI P

SC

MA NU

A1

ED

U2-2

PT

U2-1

CE

U1-2

AC

U1-1

Description This unit is dominantly yellow-dust with a little rusty color, and local portions near faults are grey-green due to the penetration of ground water. This unit mainly consists of angular or sub-angular gravel in different sizes that are mainly 4-10 cm in diameter and some are large as about 20 cm. The gravel in the base of this unit show bedding with westward dipping, specifically, the dip angle is about 20° in the base of the northern wall of trench Tc1. However, the bedding becomes gradually weak and ambiguous upward. This unit was revealed only on the northern wall of trench Tc1. The unit is composed of grey-black peat that locally contains gravel with diameter of about several millimeters. This unit contains multiple angular gravel and peat, shows a wedged shape and unconformably overlies on U1-2. The unit is interpreted as scarp-derived colluvial deposit. This unit is grey-black peat and represents a low-energy water-rich environment. The uppermost stratigraphic contact is stable and continuous. This unit is a lenses-shaped deposit of the front portion of the modern alluvial fan and mainly consists of poorly-sorted angular gravel with different sizes. The thickness of the unit is thinning westward and gradually pinches out. This unit is light-grey clay-rich fine sand and represents a low-energy fluvial sedimentation. The thickness of U2-3 in the northern wall is stable but is thinning eastward and gradually pinches out within the southern wall. This unit is black peat and represents a low-energy water-rich environment. This unit is black peat and contains tiny gravel that is abundant near fault, and locally shows individual large gravel. The unit is interpreted as scarp-derived colluvial deposit. This unit is grey-black peat with a stable thickness and represents a low-energy water-rich environment. Within the northern wall of trench Tc1, U4-1 is composed of granular gravel with different sizes mixed with grey-black peat, and contains multiple wood and charcoals. The unit is interpreted as scarp-derived colluvial deposit. This unit is grey-black peat and represents a low-energy water-rich environment. In the southern wall of trench Tc1, this unit was divided into subunits, of which a grey-white subunit was folded and possibly shows locally disturbed deformation. This unit is grey-white fine-sandy peat and represents a low-energy water-rich environment. The unit contains a lenses-shaped deposit A2 that consists of poorly-sorted angular gravel with different sizes from the front portion of the modern alluvial fan. This unit is interpreted as scarp-derived colluvial deposit that consists of mixed gravel. This unit shows three colorful sub-layers that are dark green-grey sandy clay locally containing small gravels, light green-grey sandy clay, and rusty or green fine-sandy clay from bottom the up. A brown-yellow wedge-shaped deposit that mainly comprised mixtures of gravel and sandy clay, which is interpreted as scarp-derived colluvial deposit. A brown-yellow deposit composed of rigid sandy clay. Some modern knitting wool is observed in this unit in the northern wall of trench Tc1, indicating that U6-2 might be modified.

T

Unit No.

40

ACCEPTED MANUSCRIPT Table 2 Radiocarbon samples from trenching at the Yuehua section. Amount

Calendar Years (cal BP)

Radiocarbon 13C/12C

Lab No.

of Carbon

Age

Description 1σ

(o/oo) (Years B.P. ±σ)

(mg)

1420-1370 BC 300363

2.8

-25.8

3100±30

Angular charcoal

1440-1290 BC 1350-1310 BC

U1-2,Tc1

RI P

Y055^

Unit Sampled

2σ

T

Sample

Angular charcoal

660-780 AD

wood

U2-4,Tc1

wood

U3-1,Tc1

fragment, solid

750-680 BC (18.9%)

720-690 BC Y057^

300364

3.3

-24.5

2430±30

670-640 BC (5.4%)

540-410 BC

U2-1,Tc1 fragment, solid

SC

600-400 BC (71.1%)

670-720 AD YM021^

300379

3.7

-27.7

1290±30

YM011^

305189

3.7

-25.5

MA NU

740-770 AD

1350±30

630-720 AD (87.6%)

645-680 AD

740-770 AD (7.8%)

300374

-27.1

1470±60

540-650 AD

430-670 AD

wood

U3-1,Tc1

YM004

300371

-26.9

1610±50

400-540 AD

330-570 AD

wood

U3-1,Tc1

YM018^

300376

-28.9

1050±30

975-1020 AD

wood

U3-2,Tc1

1020-1160 AD

wood

U4-1,Tc1

810-1040 AD

wood

U4-1,Tc1

wood

U4-2,Tc1

wood

U4-2,Tc1

YM013

Y018^

300375

300378

-24.6

3.8

300373

300362

-26.8

-26.3

-23.9

5

940-1030 AD (84.6%)

1080-1160 AD 890-920 AD

1080±50

720±30

890-920 AD (10.8%)

1020-1050 AD

950±30

PT

YM020^

4.7

CE

YM016

300372

AC

YM005^

3.8

ED

YM015

940-1020 AD 1220-1310 AD (90.0%) 1265-1290 AD 1360-1390 AD (5.4%) 1295-1370 AD

610±50

1280-1420 AD 1380-1400 AD 1300-1330 AD

Angular charcoal -25.0

610±30

1340-1370 AD

1290-1410 AD

U4-2,Tc1 fragment, solid

1380-1400 AD Angular charcoal

Y014^

300361

2.3

-25.3

470±30

1420-1445 AD

1405-1460 AD

U4-2,Tc1 fragment, solid

YM019

300377

-27.8

650-720 AD

600-780 AD (94.3%)

740-770 AD

790-810 AD (1.1%)

1420-1450 AD

1410-1470 AD

1330±50

wood

U4-2,Tc1

Angular charcoal Y002^

300360

1.3

-23.9

460±30

U4-3,Tc1 fragment, solid

AM010^

Y104^

300370

300366

4.4

2.4

-26.8

-24.4

200±30

1650-1680 AD

1640-1690 AD (25.5%)

1760-1810 AD

1720-1810 AD (51.6%)

1930-1960 AD

1920-1960 AD (18.3%)

1950-1870 BC

2020-1990 BC (3.3%)

Small piece of

1850-1810 BC

1980-1750 BC (92.1%)

charcoal

wood

3550±40

U5-1,Tc2

U5-2,Tc1

41

ACCEPTED MANUSCRIPT 1800-1770 BC 131.8±0.3 300359

3.1

-27.2

wood pMC

Note: All samples were processed at the Beta Analytic Inc. Miami, Florida USA. All of raw

RI P

T

radiocarbon ages are calibrated by OxCal 3.10 (Bronk-Ramsey 1998). The calibration calculates probability distributions for raw radiocarbon ages with associated uncertainties (reported by the lab facility). Radiocarbon ages BP relative to 1950. All samples typically undergo the

SC

acid-alkali-acid (AAA) method before radiocarbon dating.

* Letters ‘Y’ and ‘Z’ in sample column represent charcoal samples, ‘YM’ and ‘AM’ denote wood

MA NU

samples.

Marks ‘^’ represent samples were measured by AMS dating, and no marked samples were

CE

PT

ED

measured by standard radiometric dating.

AC

AM007^

42

U5-2,Tc2

ACCEPTED MANUSCRIPT

T

Highlights

RI P

1. We first conduct paleoseismic study on the southern segment of the ANHF.

2. Five paleoseismic events show unevenly spaced behavior of large earthquakes.

AC

CE

PT

ED

MA NU

SC

3. The ANHF may show a cascading rupturing behavior.

43