The content of this article comes from the 5th issue of the Journal of Surveying and Mapping in 2024 (drawing review number: GS Jing (2024) No. 0950)
Characteristics of Yadong-Gulu fault under the joint constraint of time-series InSAR and GNSS
Zhou Canghai
,, Tian Zhen, Shi Zhen,, Tokan Hayinar
School of Geological Engineering and Geomatics, Chang'an University, Xi'an 710064, Shaanxi, China
Funded by the Fund
National Natural Science Foundation of China (42104003) (42174054); The State Key Laboratory of Earthquake Dynamics (LED2022B02); China Postdoctoral Science Foundation (2022M710012) (2023T160557); Fundamental Research Funds for the Central Universities of Chang'an University (300102262909) (300102262902); Natural Science Basic Research Program of Shaanxi Province (2022JZ-17)
About the Author
第一作者:周苍海(1998—),男,硕士生,研方向为InSAR数据处理与形变监测。 E-mail:[email protected] 通讯作者: 石震 E-mail:[email protected]
summary
As the largest and most active extensional fault in the southeastern part of the Tibetan Plateau since the Cenozoic, the Yadong-Gulu fault plays an important role in regulating and transforming the tectonic deformation and material migration of the Tibetan Plateau. However, due to the lack of geodetic data in the past, there is a lack of quantitative analysis of the overall movement characteristics of the Yadong-Gulu fault. Therefore, this paper collects and processes the sentinel image data from October 2017 to April 2022 in the Yadong-Gulu fault area, and combines the GNSS velocity field to solve the problem of deformation field fusion in different reference frames, and obtains a high-precision and high-resolution interseismic crustal deformation field covering the entire Yadong-Gulu fault. Furthermore, based on the elastic microblock model, the current activity characteristics of the Yadong-Gulu fault were finely determined: the fault as a whole was mainly east-dipping, with an optimal dip angle of 68°, and the current tensile motion was mainly carried out, with a rate of 2~6 mm/a, and gradually increasing from south to north. In addition, the locking depth of the middle and northern sections of the fault is about 14 km. However, the southern section may be affected by the main Himalayan thrust fault, resulting in a shallow lock-up area with a depth of only 4 km. Finally, based on the sliding loss rate of the fault and the historical seismic distribution, the seismic risk of the northern section of the fault is higher, which provides an important reference for the regional geological hazard assessment.
keyword
Yadong-Gulu fault; GNSS; InSAR; tensile rate; Latching depth
This article cites format
周苍海, 田镇, 石震, 托坎哈衣那尔·. 时序InSAR与GNSS联合约束下的亚东-谷露断裂运动特征[J]. 测绘学报, 2024, 53(5): 933-945 doi:10.11947/j.AGCS.2024.20230480ZHOU Canghai, TIAN Zhen, SHI Zhen, TUOKAN Hayinaer. The characteristic of the Yadong-Gulu faults motion constraints by InSAR timeseries and GNSS observations[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(5): 933-945 doi:10.11947/j.AGCS.2024.20230480
Body reading
The Tibetan Plateau is one of the regions with the most intense and frequent lithospheric deformation in the world, and its tectonic movements, deformations, and evolution since the Late Cenozoic are the core issues in the study of continental dynamics [1]. The results show that there are multiple north-south trending normal fault valleys in the area south of the Pangong Lake-Nujiang suture zone, which are distributed on both sides of the Yarlung Zangbo River suture belt from west to east, and control the east-west extensional deformation in southern Tibet to a large extent [2-5]. The Yadong-Gulu fault is the largest and most active of the many north-south extensional faults in the southern part of the Tibetan Plateau, with a total length of about 500 km and a significant south-to-north trend, as shown in Figure 1 [2,6]. The black line in Figure 1(a) shows the location of the Yadong-Gulu fault trace, and the red and black dots show the historical earthquakes around the fault since 1200. Fig. 1 (b) and Fig. 1 (c) are the positions and names of the tracks and frames (T:track, F:frame) in the ascending and descending orbits, respectively. According to the change of fault strike, the fault can be roughly divided into three sections (northern: 30.5°N-31.5°N, middle section: 29°N-30.5°N, southern section: 27.5°N-29°N). Among them, the northern section of the Yadong-Gulu fault has good continuity, and it is also one of the areas with frequent crustal activity and frequent earthquakes in the southern part of the plateau. According to the National Earthquake Administration, since 1200, there have been 15 earthquakes of Mw 6.0 or higher [5-6] in the entire section of the fault, and 12 of these earthquakes of 6 magnitude or higher occurred in the northern and middle sections. Among them, the largest earthquake was the Dangxiong Mw8.0 earthquake in 1411, and the most recent earthquake was the Dangxiong Mw6.6 earthquake in 2008 [7-9], as shown in Figure 1. The southern section of the fault is less seismologically active, but previous studies have shown that the Yadong-Gulu fault is likely to be connected to the main Himalayan thrust fault in the deep part [10-13], so the southern section of the fault also plays an important role in regulating and transforming the crustal tectonic deformation and material transport of the Qinghai-Tibet Plateau.
Figure 1
Fig.1 Geological background of the Yadong-Gulu fault and coverage of sentinel images
Fig. 1 Geological background of Yadong-Gulu faults and the Sentinel-1 image coverage
With the continuous development of space earth observation technology, geodetic technology represented by GNSS and interferometric synthetic aperture radar (InSAR) has been widely used in crustal deformation monitoring. For the Yadong-Gulu fault, many scholars have conducted research based on geodetic techniques [3-4,14-20]. Among them, the motion rate of the Yadong-Gulu fault was estimated in Ref. [3-4,19] using GNSS observation data and elastic block model, but the overall motion nature of the fault is not clearly described due to the small number of GNSS stations around the fault. In addition, the horizontal deformation field of this region was obtained based on Envisat satellite image data and InSAR technology in Ref. [17-18,20], but due to the limited data range, most of the deformation field results did not cover the entire fault, so the results still did not depict the fine motion characteristics of the whole fault. In general, although a large number of studies have revealed the strong activity characteristics of the Yadong-Gulu fault, there is still a lack of detailed and segmented studies on the movement rate, latching distribution and seismic hazard. In order to achieve large-scale and high-precision earth observation missions, ESA has launched a number of Sentinel series satellites since 2014, among which the Sentinel-1 satellite is mainly used for all-weather day and night radar imaging. At the same time, Ref. [21] published the most complete GNSS horizontal velocity field in Chinese mainland, which also provides more observation data for the study of the Yadong-Gulu fault. Based on this, based on the Sentinel-1 data (the image range is shown in Fig. 1), this paper further fuses the GNSS velocity field to obtain a high-precision interseismic deformation field covering the entire Yadong-Gulu fault, and studies the current motion characteristics of the Yadong-Gulu fault based on elastic micro-blocks, and further analyzes its locking distribution and seismic hazard, in an attempt to provide an important reference for the assessment of geological hazards.
1 Time-series InSAR technology and data processing
1.1 Temporal InSAR technology and applications
InSAR technology can obtain a wide range and high-precision crustal deformation field quickly and without contact, and has been widely used to monitor crustal deformation, glacial drift, and seismic cycles [22-24]. However, a single interferogram is susceptible to orbit error, atmospheric delay error and elevation error, and its deformation monitoring accuracy is often low. The time-series InSAR technology can effectively remove errors such as atmospheric delay, orbit and elevation, thereby greatly improving the accuracy of regional deformation fields. Therefore, time-series InSAR technology based on multi-frame, long-term interferogram has gradually developed into the main method of high-precision crustal deformation monitoring. Commonly used time series methods include phase-weighted stacking [25], permanent scatterer [26] (PS-InSAR), small baseline set [27] (SBAS-InSAR), and improved small baseline set [28-29] (Pi-rate). The basic idea of the improved small baseline set method is as follows: firstly, the interference diagram with good coherence is adaptively selected based on the minimum spanning tree (MST) algorithm, and the unwrapping error is corrected by using the triangular phase closure loop [29]; Then, the variance-covariance matrix for each interferogram is calculated. Finally, the observations are weighted based on the variance-covariance matrix, and the interferograms are stacked based on the weighted least squares method to suppress the residual atmospheric delay error and orbit error, and the deformation rate is solved [28-29]. Compared with the traditional Small Baseline Set Method (SBAS-InSAR), the proposed algorithm has a stronger ability to correct the unwrapping error, a larger number of effective pixel points, and a higher accuracy of the deformation field. Since the area where the Yadong-Gulu fault is located contains glaciers and a large amount of vegetation, these glaciers and vegetation will lead to poor coherence of the interferogram and reduce the number of effective pixel points [30]. Therefore, the Pi-rate algorithm is used to solve the deformation field in the region, which is more adaptable.
1.2 InSAR data processing and time series analysis
In this paper, a total of 356 Sentinel-1 images from 4 orbits in the study area were selected, with a width of 250 km, a spatial resolution of 5×20 m, a wavelength of 5.56 cm, and a time span from October 2017 to April 2022, as shown in Figure 2.
Figure 2
Figure 2 Data processing and data fusion process
Fig. 2 Data processing and data fusion flowchart
In order to obtain a high-precision InSAR interseismic deformation field, the following settings are made in the data processing process. (1) Due to the large range of Sentinel-1 image data and its high requirements for registration accuracy, this paper uses the enhanced spectral diversity method for registration [31], and introduces the digital elevation model (DEM) as an auxiliary registration for external terrain information. (2) In InSAR data processing, the interference phase includes not only the surface deformation phase, but also the topographic phase, the atmospheric delay phase, the flat phase and the noise phase caused by orbit error. In view of the stable and slow deformation of the earth's crust between earthquakes, in this paper, the surface deformation phase is highlighted and the orbit error is suppressed by expanding the temporal baseline (370~750 d) and decreasing the spatial baseline (less than 80 m) in the process of interferometric pairing[32-34]. For the noise phase in the interference phase, 30×6 multi-view processing is performed on the SAR image, and the adaptive Goldstein filter [35] is used to reduce the phase noise caused by thermal noise and spatiotemporal incoherence. For the topographic phase in the interference phase, DEM data is used to simulate the terrain phase for removal, and the minimum cost flow (MCF) algorithm based on Delaunay triangulation network is used for phase unwrapping [36]. For the atmospheric delay phase in the interference phase, a linear model is established to remove the atmospheric delay phase in the tropospheric phase according to the characteristics of the tropospheric delay vertical stratification [37-38]. According to the visual interpretation, the interferogram without obvious orbit error and atmospheric delay error is selected, and the interseismic deformation field covering the entire Yadong-Gulu fault is solved by the Pi-rate algorithm, as shown in Fig. 3.
Figure 3
Fig.3 The deformation fields of the ascending and descending orbits obtained based on the Pi-rate algorithm
Fig. 3 The deformation field of ascending and descending based on Pi-rate algorithm
2. Combined with GNSS velocity field to unify the deformation field under different reference datums
Since the deformations acquired in InSAR measurements are relative to the internally selected reference point and each frame is performed separately during data processing, the deformation field reference datum obtained for each frame is inconsistent [39-40]. Therefore, it is necessary to unify the reference datum of each deformation field with the GNSS velocity field of the joint region to obtain the self-consistent deformation field under the same reference datum. It is mainly divided into the following steps. (1) Since the existing horizontal data and GNSS vertical velocity field in the Yadong-Gulu area show that the vertical motion around the fault is not obvious [41-42], the influence of vertical deformation is not considered in this paper. Based on the flight information of the Sentinel satellite, the GNSS horizontal velocity field [21] in the Eurasian frame of reference can be converted to the line of sight (LOS) of the satellite by using Eq. (1)
(1) where θ is the angle of incidence of the satellite radar; α is the azimuth of the satellite's flight; VE, VU, VN are the east-west, vertical, and north-south velocity values of the GNSS velocity field, respectively.
(2) Based on the kriging interpolation method [39], the GNSS velocity field in the direction of the satellite line-of-sight is interpolated to the same spatial resolution as the InSAR deformation field, and the GNSS interpolated velocity field consistent with the InSAR observation range and spatial scale is obtained. (3) Fitting the difference between the InSAR deformation field and the GNSS interpolated velocity field based on the quadric surface (Eq. (2)).
(2) where D is the fitting quadric; a, b, c, d, e, and f are the parameters to be found for the surface. X and Y are the coordinates of the cell point in the deformation field.
The fitted difference surface is added back to the original InSAR deformation field to correct the deformation differences caused by different reference datums [43], and finally form a unified high-precision and high-resolution ascending and descending deformation fields under the Eurasian reference frame, as shown in Figure 4. Studies have shown that the long-wavelength trend and local short-wavelength deformation in the deformation field are not affected by this method of correction [40,44].
Figure 4
Fig.4 Deformation fields of ascending and descending orbits in the Eurasian reference frame
Fig. 4 The deformation field of ascending and descending under Eurasian reference frame
3 Analysis of regional deformation field characteristics and reliability
3.1 Regional deformation field characteristics
As mentioned above, the vertical movement of the crust in this area is not obvious, and the InSAR observations are not sensitive to the north-south deformation [45], so the interseismic crustal deformation field (Fig. 4) after the unified reference mainly reflects the east-west motion characteristics of the Yadong-Gulu region. As can be seen from Figure 4, the InSAR deformation field of the ascending orbit is negative, and the deformation field of the descenting is positive. However, the deformation on the west side of the fault is mostly smaller than that on the east side of the fault in the deformation fields of the rising and descending orbits, so it is confirmed that the tensile motion characteristics of the fault are obvious [2-5]. In the northern and middle sections of the Yadong-Gulu fault, the color difference between the two sides of the fault is more obvious than that in the southern section, which may indicate that the tensile rate between the northern and middle sections is greater than that in the southern section. In addition, as mentioned above, the area contains areas of glaciers and vegetation, which will lead to poor coherence of the interferogram [30], and the blank area in Figure 4 is caused by the incoherence of the interferogram. Although there may be unwrapping errors in these blank areas, which reduces the accuracy of the observations, there are not many incoherent areas on the whole, which can better constrain the overall motion characteristics of the Yadong-Gulu fault.
3.2 Reliability Analysis
In order to verify the reliability of the above data fusion, this paper further compares the self-consistent InSAR deformation field with the GNSS velocity field. Firstly, a total of 28 GNSS stations in the region were selected by using equation (1) to convert the GNSS velocity to the direction of the satellite line-of-sight [21]. Then, the InSAR deformation value within 3.5 km was selected with the GNSS station as the center of the circle[46], and the mean value was obtained and compared with the GNSS velocity value. The correlation coefficient R2 is above 0.9, and the root mean squared error (RMSE) is less than 0.3 mm/a, indicating that the accuracy of data fusion is high, and the histograms in Figure 5(a) and Figure 5(b) represent the difference between the selected InSAR velocity value and the GNSS station velocity value, and the overall value is within 1 mm/a, indicating that the two are in high agreement. In addition, in order to further verify the possible differences between the InSAR deformation field and the GNSS velocity field, the deformation profile AA′ across the fault is further plotted (Fig. 5(c), Fig. 5(d), and the profile location is shown in Fig. 4). The profile results show that the GNSS deformation rate and InSAR deformation value are in good agreement on the whole, which further confirms the reliability of the data fusion method in this paper.
Figure 5
Figure 5 Results of reliability analysis
Fig. 5 Results of reliability analysis
4 Inversion of the elastic microblock model
4.1 Elastic microblock model
The most common method used to study the characteristics of fault motion in the interseismic crustal deformation field obtained based on space geodetic technology is to establish a velocity profile across the fault, and project the velocity profile (GNSS velocity field or InSAR deformation field) parallel or perpendicular to the direction of the fault, so as to calculate the strike-slip or tensile rate of the fault [17-18,20]. However, this method requires dense and uniform spatial distribution of observation data, clear distribution and geometric characteristics of faults, and the distribution of observation stations and fault investigation in the Yadong-Gulu region cannot meet the above requirements at present. In fact, the differential motion of the blocks on both sides of the fracture reflects the rate of motion of the fracture [3-4,19]. It has been shown that in the vicinity of active faults, the interseismic crustal deformation field obtained based on space geodetic technology includes the overall rotation of the block, the elastic deformation caused by fault blocking, and the permanent strain inside the block, while the strain inside the block can often be ignored for the elastic microblock model [47-49]. Therefore, estimating the velocity of fracture based on the elastic microblock model is also an effective method. In addition, the model can also estimate the motion parameters of rigid blocks and the distribution of the locking strength of the fracture at the same time, which provides an important reference for determining the kinematic characteristics of the fracture and evaluating the seismic risk.
4.2 Parameter settings for elastic microblock models
In this paper, an elastic microblock model in the study area is established with the Yadong-Gulu fault as the boundary, and the TDEFNODE software is mainly used for inversion [47-48]. In order to improve the inversion efficiency, it is first necessary to downsample the InSAR interseismic deformation field [40,44]. After downsampling, the ascending deformation field is 9232 data points, and the descending deformation field is 9335 data points. Since there is no strong evidence of Yadong-Gulu fault inclination and dip angle, two fault models are established in this paper, east dipping and west, respectively, and the fault dip angle is set to 40°~80° according to the research results of literature [5,9]. After repeated trial calculations, it is determined that the fracture tendency is eastward, and the optimal dip angle is 68°, as shown in Figure 6.
Figure 6
Fig.6 Relationship between root mean square error and inclination angle of inverted residuals
Fig. 6 The relationship between the RMSE errors of the inversion residuals and the inclination
In this paper, after determining the inclination angle and tendency of the fault, two sets of tests are further set up for comparison: test 1 is the inversion of a single GNSS velocity field; Experiment 2 is the inversion of the GNSS velocity field and the InSAR interseismic deformation field. The residuals of GNSS observations are shown in Table 1. By comparing the GNSS observed residuals, it can be found that the fitting residuals of GNSS stations can be reduced to a certain extent after the addition of InSAR interseismic deformation field, but the improvement is not obvious due to the small number of GNSS stations. Table 1 Residual values of GNSS observations from test 1 and test 2 Tab. 1 GNSS residuals between experiment 1 and experiment 2
In addition, the TDEFNODE software will grid the fault surface, so in this paper, five lattice points with the same depth are set along the fault trend, which are 0.01, 10, 18, 25, and 30 km, respectively, and 10 nodes along the fault strike (as shown in Fig. 7(e)), a total of 5×10 nodes. In addition, the motion parameters of the blocks on both sides of the fracture need to be inverted, so at least 50 parameters need to be inverted in this paper. However, the elastic microblock model established in this paper only includes 72 GNSS stations, and if only the GNSS velocity field is used for inversion, the constraints on the inversion parameters may be insufficient. Therefore, in this paper, the joint constraint of the GNSS velocity field and the InSAR interseismic deformation field is selected for the inversion, and the inversion residuals are shown in Figure 7. It can be seen from the inverted residual plot that the overall residual is small, indicating that the fitting accuracy of the model is high.
Figure 7
Fig.7 Inversion residuals
Fig. 7 Inverse residuals
In this paper, the tensile rate of the Yadong-Gulu fault is 2~6 mm/a, and gradually increases from south to north, as shown in Fig. 8 [3-5,15,19-20].
FIGURE 8
Fig.8 Comparison of fracture motion rates obtained by different methods
Fig. 8 Comparison of faults extensional slip rate obtained by different methods
5 Discussion
5.1 Comparison with previous studies
As the largest and most active extensional fault in the southeastern part of the Tibetan Plateau since the Cenozoic, the Yadong-Gulu fault plays an important role in regulating and converting the tectonic deformation and material migration of the Tibetan Plateau. The results of the geodetic survey method are basically concentrated on the northern and middle sections of the Yadong-Gulu fault, and the tensile rate of the Yadong-Gulu fault in the N69°W direction is (5.9±0.7) mm/a based on the GNSS velocity field as the observation data in Ref. [3]. Subsequently, the block boundary was revised on the basis of Ref. [3] and the GNSS velocity field in the Himalayas was fused to further determine the tensile rate of the Yadong-Gulu fault to be (6.6±2.2) mm/a, which is basically consistent with the results of this paper. In addition, the GNSS velocity field was used as the observation data in Ref. [19] to study the Yadong-Gulu fault through the negative dislocation block model, and the results showed that the tensile rate of the Yadong-Gulu fault was 1.2~6.0 mm/a, and the overall trend was increasing from south to north, which was basically consistent with the results of this paper. However, the tensile rate of the Yadong-Gulu fault is estimated to be only (2.3±1.0) mm/a based on a similar method in Ref. [14], which is lower than that of this paper and other studies. The possible reasons for this are as follows: (1) Ref. [14] sets the latching depth of the fault zone to 0 when performing the rate estimation, which leads to a low rate estimation; (2) The number of GNSS stations in this study is too small, especially in the northern part of the fault, which leads to insufficient constraints on the model. In addition, Envisat image data and GNSS velocity field were used in Ref. [20] to analyze the crustal motion characteristics in the central and southern Tibetan Plateau. The east-west deformation field is in the direction of N80°W, and the tension rate of the Yadong-Gulu fault is (4.5±0.5) mm/a in the region north of 29.6°N. In general, the research results of the Yadong-Gulu fault obtained in this paper are basically consistent with the existing research results in the northern and middle sections, and the research results are more refined by adding more measured data. In addition, this paper further fills the gap in the interseismic motion characteristics of the southern section of the Yadong-Gulu fault based on the spatial geodetic method. In the southern section of the Yadong Gulu fault, there is a lack of research results based on geodetic methods, so this paper chooses to compare the research results obtained with geological methods. In terms of geology, the entire section of the Yadong-Gulu fault was measured in Ref. [5] based on the topographic data obtained by dating and 3D laser scanners, and the corresponding tensile rate was obtained. Among them, the tensile rate of the southern section is 0.8~1.3 mm/a, and the tensile rate of the northern section increases sharply from (3.1±0.6) mm/a to (6.0±1.8) mm/a from south to north. The above results are less different from the fracture movement rate inferred in this paper in the northern and middle sections, but large differences in the southern section, as shown in Figure 8. Ref. [2] Based on the vertical data of stratigraphic changes and topographic fluctuations in the fault, the tensile rate of the Yadong-Gulu fault in the southern section is (1.4±0.8) mm/a. This means that there may be systematic differences between the rate of fracture motion determined by geology and the results obtained based on geodetic methods. The reasons for this discrepancy may be: (1) geological methods are often limited to a few points of fault when performing rate determination; However, this paper is based on the micro-block model, and the motion difference between the blocks on both sides of the fault is constrained by geodetic data, so as to estimate the overall motion rate of the fault, which may be more macroscopic. (2) For the Yadong-Gulu fault, which is not clear in the field, the geological method only investigates some points, and the possibility that the result is a branch fault of the Yadong-Gulu fault cannot be ruled out [2,5-6]. (3) The sampling range and time span of geodetic methods and geological methods in estimating the rate of fault movement are quite different, which may also lead to the above differences [2-3,5]. However, in general, the difference between the geological method and the results of this paper in the southern section of the rift valley is not significant[5], and the characteristics of the increase of the tensile rate of the fault from south to north are consistent, which also confirms the rationality of the research results in this paper.
5.2 Fracture and latching distribution
Figure 9(a) shows the fault locking distribution obtained by the microblock model, and it can be found that most of the locking depths in the northern and middle sections of the Yadong-Gulu fault are greater than 14 km. In the southern section of the fault, the locking depth is shallow, only about 4 km. In Ref. [49], the occlusion depth of the entire Tibetan Plateau fault was estimated by a spherical model, and the average latching depth was 16 km, which is consistent with the latching depths of the northern and middle sections obtained in this paper. However, some scholars have studied the coseismic rupture of the 2008 Dangxiong Mw 6.6 earthquake [7-9], and believe that the coseismic rupture of this earthquake occurred at a depth of about 10 km, which confirms that the locking depth in the middle of the Yadong-Gulu fault should be greater than 10 km.
Figure 9
Fig.9 Lock-in distribution and sliding loss distribution of Yadong Gulu fault
Fig. 9 The locking distribution and slip rate deficit distribution of Yadong-Gulu faults
In the southern section of the fault, the results of this paper show that the lock-up area is shallow, only 4 km. On the one hand, in the fitted residual plot of the elastic microblock model, most of the stations with large GNSS residuals are located in the southern section of the fault (red box in Fig. 7(e)). It is not difficult to see that the GNSS residuals in this area are dominated by northward movements, suggesting that the crustal movement in this area may be largely controlled by the main Himalayan thrust fault in the south. Therefore, the shallow locking depth of the southern section of the Yadong-Gulu fault obtained in this paper may also be affected by the main Himalayan thrust fault. Previous studies have shown that the Yadong-Gulu fault is likely to be connected to the main Himalayan thrust fault at depth [10-13], but the model in this paper does not consider the main Himalayan thrust fault, so it may underestimate the lock-up depth of the southern section of the Yadong-Gulu fault.
5.3 Seismic Hazard Analysis
A large number of studies have confirmed that the seismic risk of active faults is positively correlated with the slip loss caused by locking on the fault surface [47-48,50-52], and the slip loss on the fault surface can be obtained by multiplying its movement rate by the locking coefficient. In this paper, the movement rate and latching distribution of the Yadong-Gulu fault are inverted based on the elastic microblock model, and the slip loss can be further calculated, as shown in Fig. 9(b). It can be seen that the sliding loss of the northern section of the fault is the most obvious, followed by the middle section, and the southern section is the smallest. This means that the seismic risk is higher in the northern and middle sections of the fault than in the southern section alone, which is consistent with the historical seismic distribution in the area: since 1200, there have been 15 earthquakes of Mw 6.0 or higher around the Yadong-Gulu fault, but these earthquakes have mostly occurred in the northern and middle sections and less frequently in the southern section (Fig. 1(a)). In general, the extensional movement of the northern section of the fault is the most obvious (Fig. 8), and the slip loss rate is relatively high (Fig. 9(b)). Previous studies have shown that the activity of the northern section may be affected to some extent by the collapse strike-slip fault [5]. In addition, the lock-in depth is about 14 km, and the recent earthquake of Mw 6.0 or higher has been 62 years ago, so the northern section has the conditions for moderate and high-intensity earthquakes, and the risk of future earthquakes is greater. In the middle of the fault, the overall slip loss rate is also relatively high, with 14 earthquakes above Mw 5.0 in the region since 1200 (Fig. 1), but the 2008 Dangxiong Mw 6.6 earthquake has released a large number of seismic moments [7-9], so the future seismic risk in the middle section is lower than that in the northern section. However, in the southern section of the fault, the overall sliding loss rate is low, and since 1200, most of the earthquakes have not occurred directly on the Yadong-Gulu fault, but have been distributed in a distant area [6], so the seismic risk in the southern section of the fault is low as a whole. However, there is a large area of slip loss at the southernmost part of the fault (Fig. 9(b)), which may be caused by the connection of the Yadong-Gulu fault with the main Himalayan thrust fault at depth [10-13]. On the other hand, due to the continuous collision and compression of the India plate, the strong crustal convergence rate (about 16 mm/a) occurs in the main Himalayan thrust fault, so the seismic risk in this region remains unnegligible.
6 Conclusion
Based on the long-term series of sentinel image data, this paper effectively suppresses the unwrapping error, residual atmospheric delay error and orbit error in InSAR data processing through the Pi-rate algorithm, successfully extracts the signal of inter-seismic crustal motion, and further combines with the GNSS velocity field to realize the unification of the deformation field reference frame, and obtains the high-precision and high-resolution interseismic crustal deformation field covering the entire section of the Yadong-Gulu fault. The elastic microblock model is used to make up for the lack of research results based on geodetic methods in the southern section of the fault, and the following conclusions are obtained: the overall tendency of the Yadong-Gulu fault is eastward, and the optimal dip angle is 68°. At present, the activity is mainly east-west extension, and the activity rate gradually increases from south to north, and the overall activity rate is 2~6 mm/a. Based on the elastic microblock model, the latching depth in the northern and middle sections of the Yadong-Gulu fault is about 14 km. The locking depth of the southern section is shallow, only 4 km. According to the sliding loss rate of the Yadong-Gulu fault and the distribution of historical earthquakes, it is concluded that the northern section of the fault has the greatest seismic risk, but the southernmost part of the fault is still not negligible due to the influence of the main Himalayan thrust fault.
First trial: Zhang Lin review: Song Qifan
Final Judge: Jin Jun
information