Earthquake Reaearch in China  2018, Vol. 32 Issue (1): 64-79
Research of the Differential Uplifting in the Palongzangbu Region Based on the Synthesis of the Watershed Index and Seismic Data
Wang Lin, Zhao Yan, Tian Qinjian, Li Wenqiao, Xu Yueren
Key Laboratory of Earthquake Prediction, Institute of Earthquake Forecasting, CEA, Beijing 100036, China
Abstract: Research on the differential uplifting in the Palongzangbu region is crucial to understanding the tectonic deformation mechanism and establishing the model of the faulted blocks in the region of the east structural knot. In this paper, based on the ASTER-30m DEM, we calculate the hypsometry index (HI) of 19 watersheds in the Palongzangbu region, and study the differential uplifting in this region combining with seismic data, the ground deposition and erosion process. The result shows that the spatial distribution of the HI value can reflect the differential uplifting in the study area. Differential uplifting exists within different zones, and there are 2 relative strong uplifting centers. One center is near the east structural knot, and the other lies between the Lhari fault and Nujiang fault at their converging segment. Also, the watershed evolution, seismic activity, and ground deposition and erosion process are closely related to each other, and they constitute a chain of evidence which reveals the whole process from the tectonic movement underground to the landform evolution on the surface.
Key words: Watershed     Hypsometry Index     GIS     DEM     Differential Tectonic Uplifting     East Structural Knot

INTRODUCTION

The collision between the Cenozoic Indian plate and Eurasian plate causes continuous uplifting of the Himalaya and the Qinghai-Tibetan plateau, where seismicity is frequent, the fracture is well developed and the tectonic deformation is strong. The Himalaya east structural knot on the southeastern margin of the Qinghai-Tibetan plateau is at the forefront of the subduction collision of the Indian plate and the Eurasian plate. The area has strong uplifting by extrusion and well developed fault blocks, and has been a focus area of this study (Cao Jianling et al., 2009; Cui Zhongxiong et al., 2009; Teng Jiwen et al., 2006). The study of the differential uplifting of the area has important scientific significance for understanding the regional tectonic deformation mechanism and the establishment of related faulted block models. The existing research is mainly based on the difference of the age and denudation rate of bedrock with various thermal chronology methods to study the differential uplifting (Zeitler P K et al., 2014; Sun Dongxia et al., 2009; Enkelmann E et al., 2011), but because of the limitations of sample acquisition space, it is difficult to extend to the whole region, and the use of watersheds can make up for this deficiency. In fact, the river landforms in the east structural knot area are also evolving with the uplift of the Qinghai-Tibetan plateau (Wang Zhaoyin et al., 2014), where strong tectonic movement controls the formation and evolution of watersheds in the region (Huang Wenxing et al., 2013). All of these can be used for the study of differential tectonic uplifting in the watershed region.

Watershed is extremely sensitive to the changes of new tectonic movement, lithology and climate, and has a better response (Liang Peng, 2015; Zhang Huiping et al., 2006). The area-hypsometry index (HI) of the watersheds is a good way to judge the evolution characteristics of the watersheds and the landform index corresponding to its tectonic activity. The index, proposed by the American geomorphologist Strahler in 1952 (Strahler A.N., 1952), is usually based on the Digital Elevation Model (DEM). The inversion of the tectonic movement through the HI quantitative analysis is one of the hot spots in the field of active tectonics (Strahler A.N., 1952; Zhao Guohua et al., 2014; Su Qi et al., 2015, 2016; Liang Mingjian et al., 2014; Chu Yongbin et al., 2015; Wang Lin et al., 2008; Zhang Tianqi et al., 2015), but there is not much research on the differential uplifting in the east structural knot region based on the watershed index method.

In this paper, we select the Palongzangbu region, located in the area of the east structural knot and typical of the watersheds, as the research object on the ASTER-30m digital elevation model (DEM) and the watershed geomorphic unit. We study the differential tectonic uplifting in the Palongzangbu region by analyzing the responses of the watershed geomorphic units to tectonic activity with the extraction of the area-hypsometry index on the ArcGIS platform.

The existing geomorphic index research is often focused only on the surface watershed itself. Actually, we can synthesize seismic data appropriately, and combine it with the deposition and erosion process of the surface, and take the landform evolution of the tectonic movement from underground to the surface as a complete process to find the correlation between the two. This is helpful for a more comprehensive analysis and interpretation of related structural problems. Some scholars have used this method to study seismic tectonics (Wang Lin et al., 2016), and this paper is an attempt in this field.

1 REGIONAL GEOLOGICAL AND GEOMORPHIC SURVEY

The Palongzangbu region in this paper is generally located in the Lhari fault between the northeastern part of the east structural knot and the Nujiang fault (Fig. 1(a)). Specifically, it refers to the narrow strip zone in the NW-SE direction between Palongzangbu and Yigonzangbu at the great turn of the Yarlungzangbo River and the upper reaches of the Nujiang (shown as the black line in Fig. 1(b)).

 Fig. 1 Tectonic position of the study area (a) and the main faults and earthquake distribution in the study area (b) (F1 is the Nujiang fault, F2 is the Lhari fault, F3 is the Yarlung Zangbo River fault, F4 is the Mêdog fault and F5 is the Apalong fault)

The area is in the intersection of the mountains of Himalaya, Gangdisê, Nyainqêntanglha. It belongs to the Lhasa massif in terms of the tectonic location (Wang Meng et al., 2008). In terms of the new tectonic zoning it belongs to the uplifting area of the Nyainqêntanglha-Gaoligongshan fault block in the forefront of the tectonic subduction of the Indian plate to the Eurasian plate and is a new tectonic movement region (Lei Yongliang et al., 2008) with various well developed surrounding faults. In addition to the Nujiang and Lhari faults, there are also the Yarlung Zangbo River fault, Mêdog fault and Carina fault (Tang Fangtou et al., 2010; Song Jian, 2011) (Fig. 1). The development and evolution of the watershed geomorphic units in the study area are directly influenced by the Nujiang and Lhari faults on its sides.

The Nujiang fault is the southeast segment of the regional Bangong Lake-Nujiang fault tectonic belt, and is a right spin strike slip fracture along the Nujiang watersheds (Wang Yanzhao et al., 2015). The fault zone has long been controlled by the continental collision of the Indo-Eurasian plate, has undergone many complex tectonic movement processes, has multiple periods of activity, and has formed a towering orogenic belt along the fault zone, reflecting a strong neotectonic movement. With the uplift of the eastern part of the Indian plate from the late tertiary to the north-east of the region, this fault zone is dominated by the right spin-slip and squeeze-thrust (Song Jian, 2011; Li Guangtao, 2008).

The Lhari fault is the squeezed southwest boundary of the main body of the Qinghai-Tibetan plateau (Song Jian, 2011; Ren Jinwei et al., 2000). The fault is divided into two parts in the east and west with Tangmai as the boundary. The western part begins from the west of Lhari and extends along Yigongzangbu to Tangmai. The eastern part divides itself in the vicinity of Tangmai into northern and southern branches. The northern branch extends from Tangmai to Rawu along Palongzangbu. The southern part extends after Tangmai along the Gongri GAV valley in the southeast and goes into Myanmar from Chayu. The Lhari-Tangmai fault west of Tangmai developed along the Yigongzangbu River valley, and was a right spin-slip and thrust in the late Quaternary activity. Some scholars believe that the northern branch fault east of Tangmai is not very active, and the southern branch fault is right spin positive activity, which controls the development of the Gongri GAV valley (Ren Jinwei et al., 2000; Song Jian et al., 2013).

There are two main water systems in the surrounding area of the Palongzangbu and Nujiang. Palongzangbu is a primary tributary of the Yarlung Zangbo River watersheds and has two sources in the east and west. The eastern source is called Palongzangbu, and the western source is Yigongzangbu. The two form an obvious reverse river (Fig. 1(b)) with the eastern source as the main one. The eastern source develops from the Ya Nong Glacier in the northern slope of Azagonra in the east of the Nyainqentanglha Mountains, running from east to west through Rawu, Yupu, Songzong, and Bomi. The western source Yigongzangbu is the largest tributary of the Palongzangbu, which originated in the southern slope of the Nyainqentanglha Mountains in Lhari County of Nagqu, Tibet and runs from the northwest to southeast, joining the Palongzangbu near Tangmai. The Palongzangbu turns to the southeast and joins the Yarlung Zangbo River. The whole main stream is 1540m-4876m in altitude with the length of 266km (Shan Juping, 2007). The Nujiang is one of the great rivers in southwest China, originating from Guiges on the south side of the Tanggula in the Qinghai-Tibetan plateau. It runs deep into the Qinghai-Tibetan plateau, through the plain shallow valleys in eastern Tibet in the direction of northwest to the southeast, and when it runs into Yunnan Province, it makes a southern turn into Myanmar where it is named the Salween. It finally injects into the Andaman Sea of the Indian Ocean. The total length from the origin to the sea mouth is 3240km, and the Chinese part is 2013km. The area under research is mainly located in the upper reaches of the Nujiang River.

2 DATA AND PRE-TREATMENT 2.1 Data

The data used in this paper mainly includes DEM and seismic data. DEM data is ASTER GDEM data proposed by NASA and the Japanese Ministry of Economic Industry (METI) in 2009. In December 1999, the advanced satellite thermal radiation and emission gauges produced by METI on the Satellite TERRA obtained the stereo-contrast with the vertical downward imaging sensor and the rear vision imaging sensor through the infrared band and generated DEM. It covers all land areas between the Earth's surface from 83°N to 83°S, accounting for 99% of the Earth's land area. The spatial resolution of ASTER GDEM is 30m×30m, and each ASTER image is 4200 rows×4100 columns, approximately corresponding to the 60km×60km region on the Earth. This data has been widely used in the analysis and study of structural geomorphology (Chang Xiaoli et al., 2014; Chen Qiguang et al., 2014; Pike R.J. et al., 1971). Seismic data comes from a total of 8, 860 M≥1.0 earthquakes collected from the earthquake Catalogue of China Earthquake Networks Center in 1970-2016, and the data includes the information of time, latitude and longitude, magnitude and depth.

2.2 Pretreatment

For DEM data, the preprocessing steps, such as water network extraction and watershed unit extraction, should be carried out before the watershed landform index is calculated. This step is primarily based on the various features of the hydrology toolset on the ArcGIS Platform. To get the seismic data, it is necessary to transform the earthquakes from the catalogue based on their latitude information into the formats for various analysis and vector point treatments, specifically using GIS software to read two columns of longitude and latitude in the earthquake catalogue and generate vector points according to the longitudinal and latitudinal coordinates, and keep the time, magnitude and location information in the attribute table of the vector points.

2.2.1 Drainage Network Extraction

The hydrology tool set based on the ARCGIS platform computes the DEM data using the functions of the fillings, flow directions and flow accumulations and the cumulative flow can be obtained. The water network with different density can be extracted from the cumulative flow through different flow thresholds. The greater the threshold, the more sparse the water network. Because our goal is to study the area within the scope of the higher level of the watershed units of Palongzangbu, Nujiang, Maidizangbu, and Sangqu, in order to further extract the water network, as long as the extraction of the water network can include the target watershed unit of the main water system, we aim not at a fixed value. The paper sets the threshold to be 2000.

The method of calculating the water network level adopts Strahler's classification method (Strahler A.N., 1952). This method is to define the water system without any tributaries in the river networks as the first level, the intersection of the two first levels forms the second level, and so with the remaining levels until at the river network outlet. After extracting the grid river network, and combining it with the remote sensing image, the vectorization is performed to smooth the river network which is actually not in conformity with reality, and finally get a system diagram of the research area and its neighboring areas (Fig. 2). We further select the main sections of the Palongzangbu, Nujiang, Maidizangbu, and Sangqu etc., which are located within the scope of the study area (shown in the light blue thick segment position in Fig. 2).

 Fig. 2 Extraction results of water network and watershed unit in the research area
2.2.2 Watershed Unit Extraction

To obtain a more accurate watershed unit, the location of the water outlet of each watershed unit needs to be determined first. The outlet is the location of flowing water and sediment in the catchment area of the whole watershed, which eventually flows out of the watershed. Normally, the catchment point is located at the end of the main water system in the watershed and is also the lowest elevation point in a single watershed (Cui Zhongxiong et al., 2009). The water outlet of each watershed unit in the study area is located at the intersection where the main stream joins a higher level watershed unit. The water network extracted from the previous section is transformed into an independent reach through the Stream Link Function, and the intersection of the lower reaches into the upper reaches represents the outlet of the watershed unit as the main stream of the lower reach. Then, all of the nodes in the independent reach layer can be used as outlets to extract the corresponding whole watershed unit (Fig. 2). Our goal, however, is to study the higher-level watershed units in the main stream reaches of the Palongzangbu, Nujiang, Maidizangbu, and Sangqu within the research area. Thus, we need to make an appropriate integration of the lower-level watershed units within the research area. The final result of integration is shown in Fig. 2.

Eventually, we got 19 major watershed units, of which 6 run into the Palongzangbu and are named p1-p6, 11 run into the upper reaches of the Nujiang and are named n1-n11, and 1 runs into the Maidizangbu and Sangqu respectively and is named m1 and s1. The 19 watersheds have a symmetrical distribution on both sides of the main watershed of the study area, which can be divided into two groups by dividing the main watershed. Group Ⅰ is on the southwestern side of the dividing line with labels of m1, p1-p6 and s1. Group Ⅱ is on the northeastern side of the dividing line with labels of n1-n11. On the one hand, the scale of these watershed units is controlled by the fault block size. On the other hand the evolution of these watershed units is affected by the fault block tectonic movement and fault block boundary fault activity.

3 AREA-HYPSOMETRY VALUE CALCULATION 3.1 Principle Introduction

The area-hypsometry value HI mainly includes the two aspects of the area-hypsometry value and the area-hypsometry curve. As shown in Fig. 3, the area-hypsometry curve takes the relative area ratio (a/A) of the watershed unit as the horizontal axis, and the relative height ratio (h/H) is the longitudinal axis. The area below the curve is the area-hypsometry value (HI), and its significance is to describe the residual rate of the three-dimensional original surface erosion by the two-dimensional area-hypsometry curve (Su Qi et al., 2015; Zhang Tianqi et al., 2015). According to Davis' geomorphic cycle theory, Strahler believes that the area-hypsometry curve is divided into 3 forms; convex-shaped, S-shaped and concave-shaped, corresponding respectively to the geomorphic evolution of infancy, prime and old age (Zhang Tianqi et al., 2015). In the tectonic active area, the surface erosion degree is low, the development of the water system is often disturbed, and the watershed has infancy characteristics. The HI value is high (HI>0.60), and the HI curve presents a convex shape. Conversely, in the tectonic stability area, the surface erosion degree is high, the tributary river system more maturely developed, and the watershed has old age characteristics. The HI value is lower (HI < 0.35), and the curve presents a concave shape. When the tectonic uplift and the surface erosion balance each other, the river develops in the prime period, the area-hypsometry value is medium (0.35 < HI < 0.60), and the curve presents a S-shape (Zhao Guohua et al., 2014). In order to distinguish the geomorphic development stages of each watershed in a more detailed way, this paper makes a more detailed division of the geomorphic development stage, and further divides the prime age (0.35 < HI < 0.60) into 3 phases: the old prime period (0.35 < HI < 0.43), the middle prime period (0.43 < HI < 0.51) and the young prime period (0.51 < HI < 0.60). The old prime period and the young prime period are closer to the old age and the infancy stages.

 Fig. 3 Definition and calculation method of area-hypsometry value and curve
3.2 Calculation Process and Results

Based on the algorithm and module of GIS platform, the results of the area-hypsometry curve of Group Ⅰ and Group Ⅱ are calculated as shown in Fig. 4(a) and Fig. 4(b).

 Fig. 4 The result of area-hypsometry curve of Group Ⅰ and Group Ⅱ

As you can see from the HI integral curve in Fig. 4, m1 in Group Ⅰ is a low value concave type, p1, p2, p4 are middle value types, p3, p5, p6, s1 are high value convex types, and Group Ⅱ from n1 to n11 gradually change from a low value concave type to a high value convex type.

In order to quickly and accurately calculate the HI value of each watershed unit, this paper uses a simple calculation method of elevation fluctuation ratio (E) approximately equal to area-hypsometry (HI), which is presented by Pike R.J. et al., (1971), and the formula is as follows:

 $E \approx \left({{H_{{\rm{mean}}}} - {H_{\min }}} \right)/\left({{H_{\max }} - {H_{\min }}} \right) = HI$ (1)

Where Hmean, Hmax and Hmin correspond to the average elevation, maximum elevation and minimum elevation of the watershed unit, respectively.

The parameters in the corresponding Formula (1) of each watershed unit and the final HI results are shown in Table 1.

Table 1 Statistical results of the HI value of Groups Ⅰ and Ⅱ watershed units

From the calculation results, only n1 is in the old strong period, and is very close to the old age. Except for n1, all watershed units are in different degrees in the prime and infancy periods, of which m1, p2, p4, n2, n3 are in the middle prime period, and p1, p3, p5, s1, n4-n5 are in a young prime period, and p6, n6-n11 are in the infancy period.

4 COMPREHENSIVE ANALYSIS OF DIFFERENTIAL TECTONIC UPLIFTING 4.1 Synthesis of Data

In order to synthesize the HI index, seismic and surface deposition, and erosion data for the analysis of the differential tectonic uplifting, this paper sets up a projection surface P in a three-dimensional space which is vertical and axial parallel to the long axis of the study area (shown in the black line in Fig. 5(a)). The above data is projected at the same time in the surface P (Fig. 5(b)).

 Fig. 5 Plane distributions (a) and projection results (b) of the HI values, seismic data, surface depositions and erosion data in watershed units in the study area

With the HI index, we generate the respective geometric centers of each watershed unit within Group Ⅰ and Group Ⅱ (shown in black squares in Fig. 5(a)) and assign the HI value of the watershed itself as the attribute value to the central point of the watershed. Then the plane position of these points, the HI attribute values are projected on the surface P as transverse and ordinate, and finally the corresponding projection points in GroupsⅠ and Ⅱ are lined up with each other, and respectively form the HI value variation curve of each watershed unit in Groups Ⅰ and Ⅱ on the surface P (corresponding to the black graph in the range of the HI value 0.36-0.74 on the vertical axis in Fig. 5(b)). Similarly, for seismic data (shown in Fig. 5(a)), the plane position and depth value of these points are projected on the surface P as horizontal and vertical coordinates respectively. Then we find out the magnitude of the earthquake point density (corresponding to the seismic red outline map and blue-tuned density map on the vertical axis in Fig. 5(b) with seismic depth values of 0-13km). The density map reflects the number of earthquakes within the unit area and the overall level of magnitude. The high value region of the layer can be regarded as zones with frequent earthquake activities and high seismic energy release.

For deposition data, we select the boundary lines of the Quaternary strata located in the study area in the 1:2.5 million regional geological map (as shown in the irregular orange lines in Fig. 5(a)) to form the planar position of the nodes of these lines and take them as the horizontal axis. Because of the lack of the stratigraphic depth data, we set the nodal ordinates to 0, and project them onto the surface P (corresponding to the horizontal orange thick lines in the"Quaternary stratigraphic range"mark in Fig. 5(b)). The projection points in the horizontal range are the horizontal distribution of the Quaternary strata. Similarly, for the erosion data, we select the main water system in the study area (water systems marked with the names in Fig. 5(a)), and then project the plane positions and elevation values of the nodes of these lines onto the surface P as transverse and ordinate respectively. Finally, the projection points are connected in lines to obtain the riverbed section of the main water system (corresponding to the blue sections on the longitudinal axis of the riverbed profile in the range of 2, 094m-5, 011m in Fig 5(b)).

What needs to be explained here is that in this paper, the projection results of various data in the direction of the vertical axis have been zoomed multiple times and translated horizontally overall, so that they can be displayed in an orderly fashion for easy synthesis and comparative analysis.

As can be seen from Fig. 5(b), there are two relative high value centers in the curve of the HI values of the watershed units, which are located near p3 in Group Ⅰ and n7-n10 within GroupⅡ. The same positions also exist in the two Quaternary sedimentary centers and the two riverbed erosion centers, and correspond to the two underground centers with frequent seismic activity and earthquake energy release.

4.2 Analysis of Influence Factors of HI Value Spatial Distribution

The main factors that affect the HI values in the high and low spatial distributions include tectonics, lithology and climate (Liu Feifei et al., 2016; Zhao Hongzhuang et al., 2010).

From the point of view of lithology, the Mesozoic Jurassic and Cretaceous strata are mainly developed in the whole research area, only in the local areas of the ancient, Carboniferous, Permian, Triassic, Eogene and Neogene strata, the Ordovician and Devonian strata appear sporadically (Fig. 6(a)). The Jurassic and Cretaceous strata in this study area are mainly of granitic diorite, marble, quartzite, angular flash rock, gneiss, gravel plate, sandstone and other types. The Carboniferous and Permian strata are mainly silty slate, gravel-bearing sandstone-like angle flashover schist, gneiss and other types. The Triassic strata are mainly quartzite, silty slate, angular flash rock, gneiss and a small number of neutral volcanic rocks and other types. Overall, the research area is mainly composed of granite and mixed gneiss.

 Fig. 6 Quantitative statistical chart of stratigraphic distribution and the difference of HI values in different lithologic formations in the study area (a) is a sketch map of the strata in the study area on the basis of the 1/2.5 million geological map, (b) is a quantitative statistical chart of HI mean levels, range intervals and area proportions in each watershed unit for different lithology formations; and (c) is a scatter graph for the size of HI ranges of different lithologic formations with the change of stratum area

Fig. 6(b) sums up the HI mean levels domain ranges and area proportions in each watershed unit in different ages and lithology strata. Fig. 6(c) is a scatter chart of HI ranges with strata size changes in different ages and lithologic strata. The colors of the strips and blocks in Fig. 6(b) and Fig. 6(c) correspond to the same colors in Fig. 6(a) as well as to the linked age strata.

The positions and lengths of the strip below Fig. 6(b) indicate the HI ranges of the corresponding age strata. In fact, strata of the same age are often divided into several secondary stratigraphic units with different area and HI values in the watersheds of GroupⅠand Group Ⅱ in Fig. 5(a) (Fig. 6(a)). The pie chart above Fig. 6(b) accounts in sector sizes and colors for the ratio of relative areas and HI values of these secondary units. The ranges of HI values of these secondary stratigraphic units determine the field range and size of HI. The mean HI of these secondary units, which is calculated by the weighted area, can be used as the whole level of the HI values in formation, and is represented by black crosses in the stratigraphic strip. From the statistical results of Fig. 6(b) and 6(c), the HI mean levels and ranges of the different ages and lithologic strata are more affected by the distribution position and scope of strata: The greater the stratigraphic distribution is, the closer the HI value range is to the maximum width, and the more stable the HI mean level is. The more local the stratigraphic distribution, the higher the fluctuations of the HI range and mean-level, the narrower the range of the concentration-type distributed strata, and the wider the range of the scattered strata. In general, it is not apparent that a particular age or lithologic formation control has formed a phenomenon that is unusually higher or lower than the HI value of other formations. In fact, each lithologic stratum contains a subordinate unit with different HI value level, and the lithology condition does not have the spatial distribution of the HI value of the watershed unit in the study area, which can have a systematic and significant effect, and is not a major factor in controlling the difference in the spatial distribution of HI values. It is also noteworthy that there are two Quaternary sedimentary centers in the research area. One is located inside and around the Palongzangbu p3 watershed, and the other is densely dispersed in patches in the Nujiang watershed units n7-n11 and the Sangqu s1 watershed unit. (Fig. 5(a), Fig. 6(a)).

Fig. 7 is the average rainfall distribution in 500m resolution of the study area and the surrounding areas provided by the National Earth System Scientific Data Sharing Platform. It can be seen that except for the slightly higher rainfall in the southeast p4 and the slightly lower rainfall in n5, the rainfalls of all watersheds are around 500mm-600mm. This shows that the more uniform climatic conditions do not make the spatial difference of the HI value.

 Fig. 7 Distribution of rainfall in the study area and its surroundings (This data is provided by the National Earth System Scientific Data Sharing Platform)

In summary, the tectonic uplifting should be the main factor affecting the spatial distributions of HI values. In conclusion (Fig. 5(a)), the location very close to n1 of the old age is relatively stable in structure. All other watershed units in the prime age or infancy except for n1 are in overall tectonic uplifting, which shows consistency with the overall extrusion uplifting of the eastern tectonic knot and the surrounding area. However, even if the whole is in tectonic uplifting, the degree of uplifting strength between them is still different. The tectonic uplifting degrees of the m1, p2, p4, n2, n3 in the middle prime age are relatively weak, those of the p1, p3, p5, s1, n4-n5 in the young prime period are medium, and those of the p6, n6-n11 in infancy age are relatively strong.

Of course, the most unusual and eye-catching is the two HI relative high value centers mentioned above. There is no coincidence that they correspond to two Quaternary sedimentary centers, two riverbed erosion centers, and two underground seismic activity centers, which reflect strong coupling laws and show an intrinsic connection. Under the tectonic background of the strong collision of the Indian plate and the Eurasian plate, in the two center areas with high HI values, the fault blocks receive a relatively strong push action, along with the relatively high earthquake frequency and energy release level. This fault section also is more active. On the one hand this kind of state lasts for a long time, the extrusion deformation unceasingly accumulates, and forms strong tectonic uplifting, and ultimately controls and affects the development and evolution of the surface watersheds, and the direct response of the watersheds is high HI. On the other hand, the fissures and joints on the surface medium are also more developed. The uplift stratum is more susceptible to weathering and denudation, the abundant source makes the Quaternary sedimentary layer more developed, and the main riverbed is more prone to destruction from water erosion, which forms a low-lying terrain area in the riverbed section. Conversely, in the surrounding area of relative high HI value center, the size of the block may not be as strong as in the high HI value area. The frequency and energy release levels of earthquakes are not as high as the high HI values. This segment of the fault is not as active as the high HI value areas. The tectonic uplifting is not as active as the high HI value zones. However, the direct response of the watersheds shows a lower HI value level, while the Quaternary sedimentary layer is not very well developed, and the main riverbed suffers from the erosion of the geological battalion.

From the point of view of dynamics and kinematics, the strong tectonic uplifting of p3 is mainly due to the strong push of the eastern horn of the Indian plate near the east structural knot, and in the north-east of the area, the tectonic uplifting near n7-n10 is located at the junction of the Lhari and Nujiang faults, and the sliding direction of the two strike-slip faults is generally caused by squeezing (Zhu Zhicheng, 1999). Thus, it is inferred that the uplifting in the region is caused by extrusion at the intersection of the same right-slip Lhari, Nujiang faults.

In general, by interrelating the spatial variation law of the watershed tectonic response (HI value) and the spatial distribution the deposition/erosion process of deep earthquakes, we can more completely reveal the causal process of the deep tectonic movement to the surface geologic landform. This reflects the coupling mechanism between these two aspects and improves the rationality of the analysis and the credibility of the conclusion, which is the significance and aim of the comprehensive method of this paper, and also a new approach and attempt.

5 CONCLUSION

Based on the GIS platform, this paper uses the ASTER-DEM data to calculate the area-hypsometry value (HI) of the main watershed units in the Palongzangbu region, and by combining this with the seismic distribution in this region, we obtained the following understandings:

(1) The research area is in an overall structural uplifting. There are differences in the uplifting degrees in different places, among which there are two relatively strong uplifting centers. One is located near the east structural knot, and is generated by the strong push northward in the eastern horn of the Indian plate. The other is located at the junction of the Lhari fault and the Nujiang fault. It is generated by the squeezing caused by the two faults along the direction of the right spin strike.

(2) The evolution of the watersheds, the seismic activities, the ground depositions, and the erosions are interrelated. In an area of strong fault-block tectonic movement, significant tectonic uplifting caused by the intense earthquake activity, the concentrated release of energy, the continuous movement of faults, and the long-term accumulation of strain, directly control the evolution of the watersheds and makes systematic increase (or decrease) of the index of the watersheds sensitive to tectonic movement. At the same time, the fissures and joints of the surface media are more developed, and the strength of the erosion resistance is significantly reduced. The uplifting stratum is more easily weathered and eroded, and the thick Quaternary sediments formed in the periphery become the sedimentary centers, while the main channel of the fault passage is more prone to erosion, forming a low-lying erosion center.

The paper puts the spatial range of the same or partially overlapping of various data into the same coordinate system, and then into the appropriate location to intercept the section, so that all kinds of data have accurate real space relative position relationships. There will be no artificial, subjective factors caused by the relative position of the deviation. This method is an effective method for comprehensive analysis of deep data.

This paper has been published in Chinese in the journal of Earthquake, Volume 37, Number 3, 2017.

REFERENCES
 Cao Jianling, Shi Yaolin, Zhang Huai, Wang Hui. Numerical simulation of GPS observed clockwise rotation around the eastern Himalayan syntax in the Tibetan Plateau[J]. Chinese Science Bulletin, 2009, 54(8): 1398–1410. Chang Xiaoli, Zhou Yang, Luo Mingliang. Study on drainage network extraction and hypsometric integral in Yanhe watershed based on ASTER GDEM[J]. Resource Development & Market, 2014, 30(12): 1475–1477. Chen Qiguang, Shao Zhaogang, Han Jian'en, Meng Xiangang, Yu Jia, Wang Jin. Analysis of geomorphologic characteristics of the Yamzho Yumco region based on ASTER-GDE[J]. Journal of Geomechanics, 2014, 20(3): 304–316. Chu Yongbin, Zhu Lidong, Chen Wei, Zhang Jianguo, Yang Wenguang, Tao Gang. Geomorphology and evolution of the upper Shule River[J]. Quaternary Sciences, 2015, 35(2): 465–474. Cui Zhongxiong, Pei Shunping. Study on Pn velocity and anisotropy in the uppermost mantle of the Eastern Himalayan Syntaxis and surrounding regions[J]. Chinese Journal of Geophysics, 2009, 52(9): 2245–2254. Enkelmann E., Ehlers T. A., Zeitler P. K., Hallet B. Denudation of the Namche Barwa antiform, eastern Himalaya[J]. Earth and Planetary Science Letters, 2011, 307(3/4): 323–333. Huang Wenxing, Wang Guocan, Wang An, Cao Shuzhao, Cao Kai. The morphology of the Yalung Zangbo River in the Great Canyon Region and its implications[J]. Geological Bulletin of China, 2013, 32(1): 130–140. Lei Yongliang, Zhong Dalai, Ji Jianqing, Jia Chengzao, Zhang Jin. Fission track evidence for two Pleistocene uplift-exhumation events in the eastern Himalayan Syntaxis[J]. Quaternary Sciences, 2008, 28(4): 584–590. Li Guangtao. Tectonics Activity of the Nujiang Fault Zone in Quaternary in the Southwest Yunnan[D]. Master's thesis. Beijing: Institute of Earthquake Science, CEA, 2008 (in Chinese with English abstract). Liang Mingjian, Zhou Rongjun, Yan Liang, Zhao Guohua, Guo Hongmei. The relationships between Neotectonic activity of the middle segment of Dari fault and its geomorphological response, Qinghai Province, China[J]. Seismology and Geology, 2014, 36(1): 28–38. Liang Peng. The Extraction of Terrace and Geomorphology Index in the South Longmenshan Fault and Its Insight into Tectonic Implications[D]. Master's thesis. Beijing: Institute of Earthquake Science CEA, 2015 (in Chinese with English abstract). Liu Feifei, Du Guoyun. Catchment geomorphic properties of active tectonics in Yishu fault zone[J]. Quaternary Sciences, 2016, 36(2): 453–463. Pike R. J., Wilson S. E. Elevation-relief ratio, hypsometric integral, and geomorphic area-altitude analysis[J]. Geological Society of America Bulletin, 1971, 82(4): 1079–1084. DOI:10.1130/0016-7606(1971)82[1079:ERHIAG]2.0.CO;2. Qian Cheng, Cui Tianri, Li Linchuan, Chen Huijun, Qin Tao, Lu Lu. Application of ASTER-GDEM data in the geomorphic characteristics analysis of northern Great Higgnan Ling Mountains[J]. Journal of Geomechanics, 2013, 19(1): 82–92. Ren Jinwei, Shen Jun, Cao Zhongquan, Wang Yipeng. Quaternary faulting of Jiali fault, Southeast Tibetan Plateau[J]. Seismology and Geology, 2000, 22(4): 344–350. Shan Juping. Research on the Morphological Features of Yarlung Zangbo River Based on DEM[D]. Master's thesis. Beijing: China University of Geosciences (Beijing), 2007 (in Chinese with English abstract). Song Jian, Tang Fangtou, Deng Zhihui, Cao Zhongquan, Zhou Bin, Xiao Genru, Chen Weitao, Ge Weipeng. Study on current movement characteristics and numerical simulation of the main faults around Eastern Himalayan Syntaxis[J]. Chinese Journal of Geophysics, 2011, 54(6): 1536–1548. Song Jian, Tang Fangtou, Deng Zhihui, Xiao Genru, Chen Weitao. Late quaternary movement characteristic of Jiali fault in Tibetan Plateau[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2013, 49(6): 973–980. Strahler A. N. Hypsometric (area-altitude) analysis of erosional topography[J]. Geological Society of America Bulletin, 1952, 63(11): 1117–1142. DOI:10.1130/0016-7606(1952)63[1117:HAAOET]2.0.CO;2. Su Qi, Yuan Daoyang, Xie Hong, Shao Yanxiu. Geomorphic features of the Danghe River drainage basin in western Qilian Shan Mountain and its tectonic implications[J]. Quaternary Sciences, 2015, 35(1): 48–59. Su Qi, Yuan Daoyang, Xie Hong, Shao Yanxiu, Liang Mingjian. Geomorphic features of the Shule river drainage basin in Qilianshan and its insight into tectonic implications[J]. Seismology and Geology, 2016, 38(2): 240–258. Sun Dongxia, Ji Jianqing, Zhang Zhicheng, Gong Junfeng, Chen Jianjun, Qing Jianchun, Zhong Dalai. AFT dating of detrital apatites from the Yarlung Zangbo Great Canyon: Implications for its distinct geomorphological evolution[J]. Chinese Science Bulletin, 2009, 54(23): 3738–3747. DOI:10.1360/972008-2143. Tang Fangtou, Song Jian, Cao Zhongquan, Deng Zhihui, Wang Min, Xiao Genru, Chen Weitao. The movement characters of main faults around Eastern Himalayan Syntaxis revealed by the latest GPS data[J]. Chinese Journal of Geophysics, 2010, 53(9): 2119–2128. Teng Jiwen, Wang Qianshen, Wang Guangjie, Xu Ya, Zhang Xuemei. Specific gravity field and deep crustal structure of the 'Himalayas east structural knot'[J]. Chinese Journal of Geophysics, 2006, 49(4): 1045–1052. Wang Lin, He Zhongtai, Ma Baoqi. Geomorphic evolution and its implication for the fault activity in the Daihai drainage basin[J]. Quaternary Sciences, 2008, 28(2): 310–318. Wang Lin, Zhou Qingyun, Wang Jun, Li Wenqiao, Zhou Lianqing, Chen Hanlin, Su Peng, Liang Peng. The research of the seismogenic structure of the Lushan earthquake based on the synthesis of the deep seismic data and the surface tectonic deformation[J]. Seismology and Geology, 2016, 38(2): 458–476. Wang Meng, Liu Yan, He Yanbo, Wei Dong. Geomorphic characteristics of the Himalayan mountains and its tectonic implications: New insights from SRTM digital elevation model and precipitation data[J]. Chinese Journal of Geology, 2008, 43(3): 603–622. Wang Yanzhao, Wang Min, Shen Zhengkang, Shao Desheng, Shi Faqi. Present-day slip rates and potential earthquake risks along the Nujiang fault[J]. Seismology and Geology, 2015, 37(2): 374–383. Wang Zhaoyin, Yu Guoan, Wang Xuzhao, Liu Le. Impact of techtonic motion on sediment storage and morphology of Yalu Tsangpo Valley[J]. Journal of Sediment Research, 2014(2): 1–7. Zeitler P. K., Meltzer A. S., Brown L., Kidd W. S. F., Lim C., Enkelmannet E. Tectonics and Topographic Evolution of the Namche Barwa and the Easternmost Lhasa Block, Tibet[C]. In: Nie J., Hoke G. D., Horton B (Editors). Towards an Improved Understanding of Uplift Mechanisms and the Elevation History of the Tibetan Plateau. Boulder: Geological Society of America, 2014. 507: 23-58. http://www.researchgate.net/publication/284504864_Tectonics_and_topographic_evolution_of_Namche_Barwa_and_the_easternmost_Lhasa_block_Tibet Zhang Huiping, Yang Nong, Zhang Yueqiao, Meng Hui. Geomorphology of the Minjiang drainage system (Sichuan, China) and its structural implications[J]. Quaternary Sciences, 2006, 26(1): 126–135. Zhang Tianqi, Wang Zhen, Zhang Xiaomin, Hu Sheng, Lü Honghua, Zheng Xiangmin. Hypsometric integral analysis of the Vrümqi River drainage basin and its implications for topographic evolution[J]. Quaternary Sciences, 2015, 35(1): 60–70. Zhao Guohua, Li Yong, Yan Zhaokun, Yan Liang, Li Jingbo, Liang Mingjian, Ma Chao, Zheng Lilong.. Tectonic geomorphology analysis of piedmont rivers of the middle Mt.2014. Longmenshan based on hack profile and hypsometric integral[J]. Quaternary Sciences, 2014, 34(2): 302–311. Zhao Hongzhuang, Li Youli, Yang Jingchun, Yang Jingchun, Lu Honghua. Influence of area and space dependence for hypsometric integral and its geological implications[J]. Geographical Research, 2010, 29(2): 271–282. Zhu Zhicheng. Structural Geology[M]. Wuhan: China University of Geosciences Press, 1999