Earthquake Research in China  2019, Vol. 33 Issue (3): 437-450     DOI: 10.19743/j.cnki.0891-4176.201903005
Analysis of Paleoseismic Events of Dahuzhuang Trench at the Xiadian Fault
DENG Mei1, SHEN Jun1, LI Xi2, DAI Xunye1, LIU Zezhong1, LI Kechang1, JIAO Xuankai1     
1. Institute of Disaster Prevention, Sanhe 065201, Hebei, China;
2. Yunnan Earthquake Agency, Kunming 650224, China
Abstract: The Xiadian Fault is a very important concealed active fault in the Beijing Plain. It is the seismogenic fault of the Sanhe-Pinggu MS8.0 earthquake in 1679. The ancient earthquake sequence in the long historical period is of great significance to understand accurately the activity characteristics of the fault and effectively reduce the earthquake disaster risk in Beijing. We have re-interpreted the Dahuzhuang trench, and identified three layers of buried paleosol, six collapsed wedges and one sand liquefaction event. Further, through the comparison with the landmark strata and paleo-earthquake events revealed by other trenches on the fault, an ancient earthquake sequence with a long historical period of the Xiadian Fault was established:since the 31ka, the Xiadian Fault has 11 occurrences of earthquake events (including the 1679 earthquake), and the average recurrence interval is about 2.8ka. The paleo-seismic sequence also shows that there is an ancient earthquake cluster period from 25ka to 15ka, and there are 5 strong earthquakes in the cluster period. The average recurrence interval is about 2.0ka, which reflects the phase difference of the Xiadian Fault activity.
Key words: The Xiadian Fault     Paleaoearthquake event     Collapse wedge     Sand liquefaction     Symbol layer    

INTRODUCTION

The Xiadian Fault is an important concealed active fault in the northern plain of the North China Plain. It runs toward N50°E, tends to SE, and has a dip angle of 50°-70°. It starts from Pinggu Mafang in the north and ends at Caojiawu in the south, with a total length of more than 120km. (Peng Yimin et al., 1981; Xu Xiwei et al., 2000; Gao Jinghua et al., 2008). On September 2, 1679, the Sanhe-Pinggu MS8.0 earthquake was the most recent surface rupture-type earthquake in the fault, causing a terrible disaster in the east of Beijing area. The earthquake formed a seismic steep fault from the Dongliuhe village in the west, passing through the north of Xiadian Town and ending in Dongxingzhuang village in the east. It is about 10km long, showing the decline of the southeast disk and the ups of the northwest disk with a rupture mode of the right-lateral strike-slip component (Peng Yimin et al., 1981; Meng Xianliang et al., 1983; Xiang Hongfa et al., 1988). With the continuous improvement of the demand for ensuring the social and economic security and sustainable development of the metropolitan area, it is even more important to carry out research on the seismic geological characteristics of the Xiadian fault. Since the early 1980s, many scholars have carried out a lot of research on the seismic geological characteristics of the Xiadian fault and the nature of the fault activity, especially around 2000, with the help of micro-geomorphology, geostrophic method, shallow seismic exploration, and geochemistry. A series of detection techniques such as detection, trenching and drilling have studied the macroscopic epicenter, geological tectonic setting, seismogenic fault and rupture mode of the Sanhe-Pinggu MS8.0 earthquake (Ran Yongkang et al., 1997; Jiang Wali et al., 2000; Xu Xiwei et al., 2000; He Honglin et al., 2008; Liu Baojin et al., 2009; Mao Changwei et al., 2010; Wan Yongkui, 2014). In recent years, the study of stratigraphic sedimentary cycles, core particle size analysis and formation iron oxide content changes around the two disks of the fault reveals the difference between the sedimentary environment and stratum thickness of the two disks of the Xiadian fault (Yang Xiaoping et al., 2012; Zhang Chao et al., 2014; Liu Zhirong et al., 2016). The predecessors have used the trench to study the paleoseismic events of the Xiadian fault, and the ancient earthquake time limits obtained are mostly later than 15ka. In order to obtain a relatively complete paleo-seismic sequence with a long period of the Xiadian fault, this work excavated a single-wall stepped joint large trench with a depth of about 11m at about 700m south of Dahuzhuang. The paleo-earthquake events of the Dahuzhuang trench were identified by using ancient seismic events such as buried paleosol, collapsed wedge and sand liquefaction. Through the landmark stratum between the trenches, which reveals the connection of the paleoseismic events on the same time axis, and the ancient earthquake sequence and the recurrence interval of the strong earthquakes of the Xiadian fault are obtained by successive definition methods.

1 ACTIVE STRUCTURE BACKGROUND

The Beijing Plain is located in the northern part of the North China Plain, mainly including parts of Beijing, Tianjin and Hebei Province. It is a flood plain of the five major river systems of the Daqing River, Yongding River, North Canal, Chaobai River and Ji Canal (Liang Lingjun et al., 2011). It is west to the west of the Taihang Mountains, and the north is bounded by the Jundu Mountain of the Yanshan Mountains. The two mountains meet in the Guangou area of the south exit, forming a unique semi-circular "Beijing Bay" to the southeast facing the Bohai Bay. The altitude of the plain is generally 20m-60m, which is slowly decreasing toward the southeastern Bohai Bay, and the slope is about 5‰ (Zhao Zhonghai, 2009; Xu Qianhua, 2010).

Since the Cenozoics, the Beijing Plain has basically been a sedimentary environment dominated by accumulation. The regional extensional structure promotes the development of normal faults of different scales in the plains of Beijing and a series of asymmetric semi-mantle depression basins controlled by faults, such as Xiadian, Xianghe-Huangzhuang, Nanyuan-Tongxian, Liangxiang-Qianmen and Huangzhuang-Gaoliying faults, the Nankou-Sunhe and Ershili Changshan faults in NW, and Langgu and Dachang depressions (Fig. 1). The Xiadian fault is one of the more active faults in the area. The fault starts from Wangxinzhuang at Pinggu in the north, passes through Mafang and Xiadian, and reaches the Caowuying area in the south, with a total length of 120km (Peng Yimin et al., 1981). The east and west plates of the Xiadian fault are the Dachang depression and the Daxing swell respectively, and the thickness of the Cenozoic strata of the two plates differs by 3, 300m.

Fig. 1 Seismotectonic graph of the Beijing plain and distribution of fault scarp of the Xiadian fault (a) Google imagery of the North China Plain. (b) Seismic tectonic map of the Beijing Plain. (c) Distribution of the steep ridge of the surface rupture zone of the Sanhe-Pinggu MS8.0 earthquake in 1679 F1: Xianghe-Huangzhuang fault; F2: Xiadian fault; F3: 20-mile Changshan fault; F4: Nanyuan-Tongxian fault; F5: Qianmen-Liangxiang fault; F6: Huangzhuang-Gaoliying fault; F7: Nankou-Sunhe fault; Tc1, Tc3 are the locations of the trenches excavated by Xiang Hongfa et al. (1988); Tc2 is the location of the trenches excavated by Ran Yongkang et al. (1997); Tc4 is the location of the trenches in this paper; Tc5 is the location of the trenches excavated by Jiang Wali et al. (2000)
2 DAHUZHUANG TRENCH 2.1 Trench Location

The Dahuzhuang trench is located about 700m south of Dahuzhuang village in Qixinzhuang town, with an altitude of about 10m (Fig. 2). It is a single-wall large trench obtained by stepped excavation at the edge of an artificial sand mining pit. The trench consists of upper and lower parts. The upper part has a maximum depth of 3.5m and a length of about 7m. The lower part has a maximum depth of 8.5m and a length of about 8m. The horizontal distance between the upper and lower grooves is about 4m. The east and west sides of the trench contain relatively obvious seismic steep slopes. Using real-time kinematic (RTK), the heights of the steep ridges on the east and west sides of the trench are 1.72m and 1.52m respectively.

Fig. 2 Position of the Dahuzhuang trench
2.2 Trench Measurement Samples

Age data is the key to defining ancient earthquake events. From sample collection to data analysis, strict compliance with relevant regulations is required to obtain reliable chronological data. The 14C dating samples of the Dahuzhuang trench are strictly in accordance with the requirements for sample collection in active fault detection, and the position, lithology and other relevant parameters of the sample are recorded in detail. The samples were all completed by the Quaternary Age Measurement Laboratory of Peking University. A total of 19 14C dating data are used in the Dahuzhuang trench. The dating data are all corrected values. The specific parameters of the samples are shown in Table 1.

Table 1 Dating sample parameters of the Dahuzhuang trench
2.3 Trench Stratigraphic Sequence

We analyze the stratum of the upper and lower parts of the trench, and combine the dating data of the stratum to divide the stratum of the trench into six stratigraphic units, namely soil layer, clay layer, silty clay layer, clay and silt layer, clay layer and silt layer.According to whether the surrounding units contain the signs of paleoseismic events such as collapsed wedges, the sub-layers are further divided. The stratum numbers are marked from new to old and from the upper to the lower plates.

The stratum of the upper trench is mainly composed of 3 layers of soil layer, clay layer and silty clay layer (Fig. 3(b)). U0 is a gray-black soil layer, which is distributed in both plates and contains a large number of plant roots with a layer thickness of 0.4m to 0.5m. U1 is a light yellow-brown clay distributed in both plates with a layer thickness of 0.6m to 0.8m. with a layer thickness of about 0.3m and some snail shell debris. The 14C dating data of the snail shell is (12.86±0.035)ka. U1-1 is gray-yellow clay with a layer thickness of about 0.4m. S2 is a gray-brown paleosol layer with a layer thickness of 0.3m-0.4m. U1-2 is gray-yellow clay with a layer thickness of about 0.6m. S3 is a gray-brown paleosol layer. There are two obvious dislocations near the fault plane. The layer thickness is about 0.4m, rich in plant roots and contains rust-colored rust spots. The 14C dating data of this layer is (15.05±0.045)ka. The above strata are the same set of strata to subdivide the three sedimentary cycles of paleosol and clay. This may be due to the rapid transformation of the sedimentary environment caused by three ancient earthquake events, thus forming three intermittent paleosols. The collapse wedge W1 is under S3 and is a mixture of brown silt and clay. U2 is a light gray-yellow silty clay. The layer thickness of the upper trench is about 2m. The 14C dating data at the top of the layer is (14.98±0.045)ka.

Fig. 3 Picture of the upper trench (a) and comprehensive interpretation diagram (b)

The lower trench mainly consists of five stratum units such as clay layer, silty clay layer and silt layer (Fig. 4(b)). U1-2 is gray-yellow clay. By analysis and comparison, we believe that this layer is the same layer as the upper trench U1-2, and the underlying S3 and W1 of the layer correspond to the stratum of the upper trench. U2-1 is light yellow-brown clay with a near-horizontal bedding thickness of 0.7m. It has a local carbon content, contains a small amount of plant roots, and has vertical rust spots. Two sand veins with a diameter of about 1cm appear in the layer. 14C dating data is now (15.69±0.045)ka, for the phenomenon that there are 2 dating data (sample numbers: XD13-14C-11-1, XD13-14C-11-2) in this layer, we think this is due to the leaching effect causing the carbonaceous particles of the upper young formation to infiltrate into the formation. A pale yellow clay collapse wedge W2 develops at the bottom of the layer. The collapse wedge is located slightly away from the fault plane and has a large thickness at the distal end of the section. This may be because the upper plate is in the process of descending after an earthquake event. Partially formed fault depression ponds with low landslides are formed by sedimentation in a clean water environment. U2-2 is a taupe silty clay with a layer thickness of 0.5-0.8m. The 14C dating data at the top of the layer is (17.13±0.045)ka. U2-3 is a light gray-brown silty clay with a layer thickness of 0.3m-0.6m. The top 14C dating data of this layer is (18.82±0.05)ka, and the bottom is developed with a light gray-brown collapse wedge W3. U2-4 is a gray clay silty sand with a layer thickness of 1.0-1.5m. The upper 14C dating data of this layer is (23.32±0.07)ka. A collapse wedge W4 is developed at the bottom of the layer, and the composition is similar to that of the overlying formation.

Fig. 4 Picture of the lower trench (a) and comprehensive interpretation diagram (b)

U3 is gray-brown clay with a layer thickness of 0.2-0.4m. The 14C dating data of the top layer of the layer is (23.83±0.08)ka, and the underlying collapse wedge W5 is composed of light gray-yellow clay. U4 is dark brown clay. There is no bottom boundary. The top 14C dating data of this layer is (24.31±0.08)ka, and the bottom 14C dating data is (25.44±0.08)ka. The collapsed wedge W6 is located at the bottom of U4, and the lithological composition is dark brown clay, which is relatively loose and has obvious boundaries with the overlying strata.

Compared with the upper plate, the deposition environment of the lower plate is relatively stable, and the formation is easier to divide. U2 is a light gray-yellow silty clay, which is the same layer as U2 in the upper trench. The thickness of the layer is about 3m by analyzing the stratum of the upper and lower trenches. U3 is light gray-brown clay, light gray and yellow silt. The former layer is about 0.3m thick, the latter layer is 0.8m thick, and the silt is well sorted. This layer is very similar to the lithology of U3 on the upper plate. It may be that the difference in water content causes the color of the lower plate to be darker. U4 is dark brown clay with a layer thickness of about 3m and a thin layer of silt in the middle. The top, middle and bottom 14C dating data of this layer are (23.43±0.07)ka, (23.74±0.08)ka, and (31.07±0.14)ka. The bottom is in unconformity contact with the underlying silt layer. U5 is grayish white fine sand with a layer thickness of 2.5m. There is no bottom boundary. There is a diagonal layering. The sorting is good. The gray thin layer of clay is developed in the layer. The thin layer of clay contains rust spots and wrinkles. We believe this is a sand liquefaction.

3 PALEOSEISMIC EVENTS AND COSEISMIC DISPLACEMENT OF THE EXPLORATION TRENCH 3.1 The Paleoseismic Events Revealed by the Trench

In general, earthquakes with magnitudes greater than 6.0 produce significant displacement and deformation on the surface (Meng Xianliang et al., 1983). To fully reveal the paleo-earthquake events, it is necessary to carefully search for the general signs of the paleo-earthquake events in the trench profile. For example, the sedimentary facies in the sedimentary strata before the fault can often have signs such as a binary structure of the collapse wedge, paleosols buried in the upper plate, and sand liquefaction (Ran Yongkang et al., 2014). On the upper plate of the upper trench, we identified three paleoseismic events corresponding to the three buried paleosols S1, S2, and S3 and the paleoseismic events corresponding to the collapsed wedge W1. Six collapsed wedges were identified in the lower trench, one of which was identical to the WI revealed by the upper trench and one sand liquefaction.

According to the 14C dating data, we can only determine that the starting and ending time limits of the three paleoseismic events corresponding to S1, S2, and S3 should be (15-0.31)ka, and the starting and ending time of the last ancient earthquake event S1 is (8-0.31)ka. But since the time limit of S1 is too big, we speculate that the event time limit is about 5ka. S2 and S3 can only determine the period of the event. We speculate that the time limit of S2 should be around 11ka and S3 around 13.8ka.

In view of the deep depth of the Dahuzhuang exploration trench, the surrounding lithology is mostlyclay and silt. The newly-faced clay layer of the trench just excavated has high water content and strong cohesiveness, which makes it difficult to clean and smooth the trench wall. The collapse wedge in the layer is difficult to distinguish at a time, and it is necessary to dry it to see the shape of the collapsed wedge. The silt layer in the trench is relatively loose, and the fresh surface is easily weathered, which makes it difficult to identify the collapsed wedge. We carried out a number of clean-ups of the Dahuzhuang trench and photographed the original photos of the trench walls that were kept for different periods. In order to supplement the integrity of the ancient earthquake events, we determined some less obvious collapse wedges by comparing the photos of different periods and using image enhancement, and limited the start and end time of the paleoseismic events by the age of the upper and lower strata. A collapsed wedge represents an ancient earthquake event. The Dahuzhuang exploration trench reveals a total of six collapsed wedges (Fig. 5). The collapsed wedge number is W1-W6 from new to old, and the corresponding paleoseismic event is E4-E9. W1 is about 2m long and 0.5m thick near the side of the fault. The shape of W2 is different from that of other collapsed wedges. The deposition away from the fault plane is thicker, with a thickness of about 0.5m and a length of about 1.5m. The boundary between W3 and the surrounding strata is not obvious, but it is still possible to speculate that there is a collapse wedge by comparing the profiles at different periods. W4 is about 1.8m long and looks like an obtuse triangle on the section. There is trace of traction deformation on the side of W5 near the fault plane, and there are signs of displacement in the tail. W6 is the collapse wedge at the bottom of the trench. It can clearly see the obvious boundary between the collapsed wedge and the surrounding stratum, but there is a lack of the chronological data of the underlying strata. In addition, there is a sand liquefaction phenomenon at the bottom of the lower plate of the fault. The 14C dating data (31.07±0.14)ka of the overburden can be obtained earlier than 31.21ka, and there was an ancient earthquake, recorded as E10. The M8.0 earthquake in 1679 is recorded as E0.

Fig. 5 Partial enlargement diagram of collapse wedges (a-f) and sand liquefaction (g)

Table 2 shows that the Dahuzhuang exploration trench revealed a total of 11 earthquake events, including three times of buried paleosol and six times of collapsed wedges. Sand liquefaction corresponds to one earthquake and the M8.0 earthquake of 1679.

Table 2 Main parameters of paleo-earthquake of the Dahuzhuang trench
3.2 Coseismic Displacement

By analyzing the sequence of the Dahuzhuang trench, we obtained the amount of dislocations in the two plates at different time. The amount of dislocations in the U4 layer of the lower trench is 3.2m, and the amount of dislocations in the U3 layer is 2.85m. The difference in the gap between the two layers also supports the existence of W5. The amount of dislocations in the U2 layer cannot be calculated in the trench profile, but there is a significant difference in the thickness of the stratum between the upper and lower plates of the stratigraphic unit. The thickness of the faulted upper stratum is about 4m, and the thickness of the stratum of the lower plate is about 3m. The difference in deposition rate between the two plates (24-15) during this period just indicates that the ancient earthquake occurred in this segment, and the amount of dislocation is about 1m. The amount of dislocations in the paleo-soil layer S of the two trenches in the upper trench is about 1.75m. Therefore, we obtained a cumulative displacement of 3.2m since the Xiadian fault of 24.3ka. The coseismic displacement of the seismic event corresponding to W5 is about 0.35m, and the average of the three earthquake events corresponding to W4, W3 and W2 is the same. The seismic displacement is about 0.33m, the coseismic displacement of the seismic event corresponding to W1 is about 0.1m, and the average coseismic displacement of the three seismic events corresponding to S1, S2 and S3 is 0.48m.

Because of the weathering, erosion and artificial transformation, the vertical deformation caused by earthquakes will be gradually dispersed into a larger area. The faults revealed in the trenches are often only part of the deformation, but the period of the ancient earthquake event can still be identified by the difference of the breaks in different time periods. We used the RTK to obtain a 1.2km long micro-geomorphic line along the north-south road along the trench profile (Fig. 6). Fig. 6 shows that the upper and lower original elevations of the two plates in the measurement area are about 3.6m, and the height of the steep slope of the 1679 earthquake we measured is about 2m from this value, which indicates that the shape variables of the Xiadian fault do not only include the amount of dislocations and the height of the steep ridges in the trench, but also need to consider the surface-shaped variables that are dispersed over a larger range.

Fig. 6 Topographical profile in Dahuzhuang
4 DISCUSSION 4.1 Comparison with Previous Data

Previous studies using the trench technology to study the ancient earthquakes of the Xiadian fault have yielded fruitful results, revealing the multi-phase earthquakes since 20ka. Xiang Hongfa et al. (1988) first used the trenching technique to carry out ancient earthquake research on the Xiadian fault. They excavated a number of trenches in Pangezhuang and Dongliuhetun. Among them, the length of the Pangezhuang No.3 trench was 6.2m, and the maximum depth is 3.78m. The length of the Dongliuhetun trench is 8m and the maximum depth is 3.5m. Ran Yongkang et al. (1997) excavated a large trench with a length of 20m, depth of 6m, upper width of 6m and lower width of 1.5m in the northeast of Dongliuhetun. Jiang Wali et al. (2000) also excavated a number of trenches in Qixinzhuang. The most representative Qixinzhuang No.2 trench has a length of 14m and maximum depth of 4m. Through the analysis and comparison of the lithology and dating data of the above trenches, it is found that all the strata have snail shell fragments and the dating results are all about 15ka.

According to the symbol layer in Table 3 and the corresponding dating data, the following results can be obtained. First of all, except the Dongliuhetun trench, all trenches have snail shell fragments. The 14C dating result of the snail shell fragments of the northeast trench of Dongliuhetun is (11.62±0.15)ka. The 14C dating result of the snail shell fragments of the Dahuzhuang trench is (12.86±0.035)ka. The results of the two snail shell gragments are basically the same. Secondly, all the trenches contain thin carbon chips or snail fragments, and the 14C dating results of the carbon fragments in this layer or nearby layers are around 15ka. Therefore, we believe that the above trenches containing shell fragments and thin layer carbon scraps belong to the iconic micro-stratigraphic unit of the same era. Based on this, we establish the connection of each trench on the time axis.

Table 3 Comparison of symbol layers in respective trenches
4.2 Successive Definitions of Ancient Earthquakes

Zhang Peizhen et al. (2005) used the successive method to analyze the paleoseismic eventsin each section of the Haiyuan fault. This paper compares the paleoseismic eventsof the five trenches in the Xiadian fault and obtains 11 paleoseismic events by successive definition methods from new to old, recorded as Ea-Ek (Fig. 7). Among them, Ea is the Sanhe-Pinggu MS8.0 earthquake in 1679. Eb is revealed by Tc3 and Tc5, and the event time limit is limited by Tc3, which is (4.73±0.43)ka BP. Ec is defined by Tc2 and Tc5, and the event time limit is limited by Tc5. (7.07±0.95)ka BP. Ed is jointly revealed by Tc1, Tc2, Tc3 and Tc5, and Tc2 defines the upper limit of the event. Tc5 defines the lower limit of the event, so the time limit of this event is (11.62±0.15)ka BP. Ee and Ef are revealed only by Tc4, and the event time limits are (15.37±0.275)ka BP and (16.52±0.78)ka BP. Eg is jointly revealed by Tc2 and Tc4, and Tc2 limits the upper limit of the event, and Tc4 limits the lower limit of the event, so the time limit of Eg is (21.63±1.625)ka BP. Eh and Ei are revealed only by Tc4, and the event time limits are (23.57±0.18)ka BP and (24.07±0.16)ka BP. Ej and Ek expose the upper limit by Tc4, which are 25.52ka BP and 31ka BP respectively. The time limits of the two events are about 25.52ka and 31ka respectively. From the distribution of 11 paleoseismic events on the time axis, the average recurrence interval since 31ka was 2.82ka. In addition, there is a cluster period in the ancient earthquakes of Xiadian fault, that is, five strong earthquakes occurred within (25-15)ka from the present, with an average recurrence interval of 2ka.

Fig. 7 Comparison of paleoseismic in different zones of the Xiadian fault
4.3 Collapse Wedge Evolution Model

The section morphology revealed by the Dahuzhuang exploration trench and the wild fault outcrop is mainly dominated by single dislocations, and most of the multi-period ancient earthquakes continue to be discontinued along the main section. In this paper, the evolution model of the collapse wedge and the stratum is drawn by combining the lithologic characteristics of the Dahuzhuang trench and the lithology and age of the stratum (Fig. 8). The overall evolution pattern is as follows: An earthquake occurred after the completion of the U3 formation deposit, and W6 was deposited near the section of the fault on the upper plate, and then the upper plate continued to accept deposition to form U3-2. The earthquake that occurred after the deposition of U3-2 caused the fracture to continue to fall and accumulated W5 and U3-1. The earthquake that occurred afterwards accumulated W4, and both plates were subjected to deposition to form U2-4. The next earthquake formed a collapse wedge W3, and the upper plate continued to accept deposition to form U2-3. Later, due to the change of the macroscopic sedimentary environment, U2-2 was deposited on the two plates. The subsequent earthquake once again shifted the formation below U2-2, forming a low-lying terrain under the traction of the upper plate and accumulating W2. When W2 filled the low-lying land of the upper plate. The two plates were again deposited due to the change of the depositional environment U2-1. The next earthquake formed W1, and both plates received sediment to form U1-5 and U1-4.

Fig. 8 Schematic diagram of the collapse of the Dahuzhuang trench and the evolution of the stratum
5 CONCLUSION

Through the analysis of the paleoseismic events in the Dahuzhuang exploration trench on the Xiadian fault earthquake surface rupture zone, we obtained the following conclusions:

In the Dahuzhuang and other trenches, we have found the symbol layer containing the shell fragments and thin carbon chips. The 14C dating data of the shell fragments or carbon chips in the symbol layer basically match, indicating that the symbol layer is a micro-stratigraphic unit that exists in a macroscopic environment. This paper also reveals 11 paleoseismic events since the Xiadian fault 31ka, with an average recurrence interval of 2.8ka. However, there is an ancient earthquake clustering period from 25ka to 15ka, and there were 5 strong earthquakes in the cluster period, with an average recurrence interval of 2ka.

The Dahuzhuang exploration trench reveals the ancient earthquake sequence longer than 15ka, and also found three earthquakes corresponding to three buried paleosols since 15ka or on the symbol layer, perfecting the completeness since 31 ka. The sequence of paleoseismic events provides a basis for analyzing the activity characteristics of the Xiadian fault. With the development of the Sanhe City Urban Activity Fault Detection Project and the attention of more researchers to the Xiadian fault, it is necessary to establish a sound network of research results to more accurately identify the activity of the Xiadian fault.

REFERENCES
Gao Jinghua, Xu Mingcai, Rong Lixin, Chai Mingtao, Wang Guangke, Wang Xiaojiang, Liu GuanjunGao Jinghua, Xu Mingcai, Rong Lixin, Chai Mingtao, Wang Guangke, Wang Xiaojiang, Liu Guanjun. Activity of the southwest segment of Xiadian fault investigated by seismic reflection profiling[J]. Seismology and Geology, 2008, 30(2): 497-504 (in Chinese with English abstract).
He Honglin, Min Wei, Haraguchi THe Honglin, Min Wei, Haraguchi T. Testing geo-slicer on the rupture of the M8.0 Sanhe-Pinggu earthquake of 1679[J]. Seismology and Geology, 2008, 30(1): 289-297 (in Chinese with English abstract).
Jiang Wali, Hou Zhihua, Xiao Zhenmin, Xie XinshengJiang Wali, Hou Zhihua, Xiao Zhenmin, Xie Xinsheng. Stady on paleoearthquakes of Qixinzhuang trench at the Xiadian fault, Beijing plain[J]. Seismology and Geology, 2000, 22(4): 413-422 (in Chinese with English abstract).
Liang Lingjun, Yang Zhongshan, Bai GuoyingLiang Lingjun, Yang Zhongshan, Bai Guoying. Discussion on the division method of water system in the plain area of Beijing based on GIS[J]. Beijing Water, 2011(6): 4-6 (in Chinese).
Liu Baojin, Hu Ping, Meng Yongqi, Feng Shaoying, Shi Jinhu, Ji JifaLiu Baojin, Hu Ping, Meng Yongqi, Feng Shaoying, Shi Jinhu, Ji Jifa. Research on fine crustal structure using deep seismic reflection profile in Beijing region[J]. Chinese Journal of Geophysics, 2009, 52(9): 2264-2272 (in Chinese with English abstract).
Liu Baojin, Zhang Xiankang, Chen Yong, Feng Shaoying, Ji Jifa, Yuan Hongke, Zuo YingLiu Baojin, Zhang Xiankang, Chen Yong, Feng Shaoying, Ji Jifa, Yuan Hongke, Zuo Ying. Research on crustal structure and active fault in the Sanhe-Pinggu Earthquake (M8.0) Zone based on single-fold deep seismic reflection and shallow seismic reflection profiling[J]. Chinese Journal of Geophysics, 2011, 54(5): 1251-1259 (in Chinese with English abstract).
Liu Zhirong, Shen Jun, Huang Jingyi, Chen YaofeiLiu Zhirong, Shen Jun, Huang Jingyi, Chen Yaofei. Grain size analysis of the Late Pleistocene Sediments in Sanhe County, Hebei Province[J]. Acta Geologica Sinica, 2016, 90(5): 997-1005 (in Chinese with English abstract).
Mao Changwei, Ding Rui, Gong Zheng, Zhang ShiminMao Changwei, Ding Rui, Gong Zheng, Zhang Shimin. GPS survey of the surface fault scarp of 1679 Sanhe-Pinggu M8 earthquake[J]. Bulletin of the Institute of Crustal Dynamics, 2010, 11-18 (in Chinese with English abstract).
Meng Xianliang, Du Chuntao, Wang Rui, Liu ShipingMeng Xianliang, Du Chuntao, Wang Rui, Liu Shiping. Seismic fault zone of the Sanhe-Pinggu earthquake in 1679[J]. Earthquake, 1983(3): 18-23 (in Chinese).
Peng Yimin, Li Dingrong, Xie Zhenzhao, Wang Ande, Liu QingsiPeng Yimin, Li Dingrong, Xie Zhenzhao, Wang Ande, Liu Qingsi. Some features of contemporaneous faults in Beijing plain and their significance[J]. Seismology and Geology, 1981, 3(2): 57-64 (in Chinese with English abstract).
Ran Yongkang, Deng Qidong, Yang Xiaoping, Zhang Wanxia, Li Rucheng, Xiang HongfaRan Yongkang, Deng Qidong, Yang Xiaoping, Zhang Wanxia, Li Rucheng, Xiang Hongfa. Paleoearthquakes and recurrence interval on the seismogenic fault of 1679 Sanhe-Pinggu M8.0 earthquake, Hebei and Beijing[J]. Seismology and Geology, 1997, 19(3): 193-201 (in Chinese with English abstract).
Ran Yongkang, Li Yanbao, Du Peng, Chen Lichun, Wang HuRan Yongkang, Li Yanbao, Du Peng, Chen Lichun, Wang Hu. Key techniques and several cases analysis in paleoseismic studies in mainland China(3): Rupture characteristics, environment impact and paleoseismic indicators on normal faults[J]. Seismology and Geology, 2014, 36(2): 287-301 (in Chinese with English abstract).
Wan Yongkui, Shen Jun, Yu Xiaohui, Dai Xunye, Wang Lei, Shao Bo, Li XiangWan Yongkui, Shen Jun, Yu Xiaohui, Dai Xunye, Wang Lei, Shao Bo, Li Xiang. The slipping rates and ancient earthquake recurrence interval at Xiadian Fault, Beijing Plain[J]. Journal of Institute of Disaster Prevention, 2014, 16(3): 38-45 (in Chinese with English abstract).
Xiang Hongfa, Fang Zhongjing, Xu Jie, Li Rucheng, Jia Sanfa, Hao Shujian, Wang Jingbo, Zhang WanxiaXiang Hongfa, Fang Zhongjing, Xu Jie, Li Rucheng, Jia Sanfa, Hao Shujian, Wang Jingbo, Zhang Wanxia. Seismotectonic background and recurrence interval of great earthquakes in 1679 Sanhe-Pinggu M=8.0 earthquake area[J]. Seismology and Geology, 1988, 10(1): 15-28 (in Chinese with English abstract).
Xiang Hongfa, Fang Zhongjing, Zhang Wanxia, Jia Sanfa, Li RuchengXiang Hongfa, Fang Zhongjing, Zhang Wanxia, Jia Sanfa, Li Rucheng. Preliminary study on the late Quaternary activity of concealed faults in the Beijing Plain[J]. Acta Seismologica Sinica, 1993, 15(3): 385-388 (in Chinese with English abstract).
Xu Xiwei, Ji Fengju, Yu Guihua, Chen Wanbin, Wang Feng, Jiang WaliXu Xiwei, Ji Fengju, Yu Guihua, Chen Wanbin, Wang Feng, Jiang Wali. Reconstruction of paleoearthouake sequence using stratigraphic records from drill logs: a study at the Xiadian fault, Beijing[J]. Seismology and Geology, 2000, 22(1): 9-19 (in Chinese with English abstract).
Xu Qianhua. A Study about the Relationship between Quaternary strata and the new tectonic movement within Beijing Plain[D]. Beijing: China University of Geosciences, 2010 (in Chinese with English abstract).
Yang Xiaoping, Cao Jinghu, Chen XianchengYang Xiaoping, Cao Jinghu, Chen Xiancheng. The iron oxide changes in drilling cores from the two walls of Xiadian active fault[J]. Seismology and Geology, 2012, 34(4): 659-671 (in Chinese with English abstract).
Zhang Chao, Shen Jun, Zhao Jiangtao, Zhou Chenyang, Li QingZhang Chao, Shen Jun, Zhao Jiangtao, Zhou Chenyang, Li Qing. Research on sedimentary cycles of Xiadian Fault near Beijing[J]. Journal of Institute of Disaster Prevention, 2014, 16(2): 49-57 (in Chinese with English abstract).
Zhang Peizhen, Min Wei, Deng Qidong, Mao FengyingZhang Peizhen, Min Wei, Deng Qidong, Mao Fengying. Paleoearthquake rupture behavior and recurrence of great earthquakes along the Haiyuan fault, northwestern China[J]. Science in China Series D: Earth Sciences, 2005, 48(3): 364-375. DOI:10.1360/02YD0464
Zhao ZhonghaiZhao Zhonghai. Analysis of the neotectonic activity and the bad geological phenomenon caused by it in Beijing plain areas[J]. Resources Environment & Engineering, 2009, 23(3): 282-286 (in Chinese with English abstract).