Earthquake Reaearch in China  2017, Vol. 31 Issue (2): 169-178
The North-South Seismic Belt: Vertical Deformation Velocity Gradient Research
Liu Liwei, Ji Lingyun, Zhao Qiang
The Second Monitoring and Application Center, CEA, Xian 710054, China
Abstract: The vertical deformation gradient can reflect the rate of vertical change in unit distance, and the vertical deformation velocity gradient can reflect the strength of the earths crust tectonic activities. In this paper, using long period leveling data combined with GPS data, the vertical deformation gradient values are calculated. Leveling data and GPS data are two different means of monitoring deformation, but the result is approximately the same vertical deformation gradient. The results show that the spatial distribution of the vertical deformation velocity gradient and tectonic distribution has an obvious correlation. The most significant gradient anomalies along the North-South Seismic Belt are Xianshuihe fault, Longmenshan fault and Xiaojiang-Zemuhe fault, while the second gradient anomalies in the northeastern Qinghai-Tibetan plateau are Zhuanglanghe fault and Lenglongling fault. The Menyuan MS6.4 earthquake in 2016 occurred in this abnormal area. However, according to the vertical deformation high gradient area distribution, there is also the possibility of an earthquake occurrence in the Tianzhu and Jingtai area. The area of convergence of three major fault zones is the strongest tectonically active region of the North-South Seismic Belt.
Key words: Gradient     North-South     Seismic Belt     Vertical deformation     Strong Earthquake activty

INTRODUCTION

The North-South Seismic Belt and its surrounding areas, bordered between the east and west tectonic systems in the Chinese mainland, characterized by significant relative movement and deformation of tectonic active blocks and frequent strong earthquakes, are sensitive to overall tectonic dynamic environment and stress field changes (Ding Guoyu et al., 1993). Therefore, the North-South Seismic Belt has become an area most frequently studied for the relationship between precursor anomalies and earthquake occurrence, and also a key area for earthquake prediction and risk tracking. Crustal deformation monitoring and research are effective means of earthquake prediction. Currently, crustal vertical movement velocity obtained by calculation with local leveling data and GPS data has only relative significance, because this velocity is always by reference to a standard velocity. For the selection of this standard velocity, although we can make assumptions based on prior knowledge to make it closer to actual conditions, this kind of standard is presently subjective due to lack of constraints from external conditions. Thus, the velocity value itself can hardly be used as an objective criterion to measure crustal movement. Therefore in the study of crustal movement and seismic activities, the change of velocity (in direction and size) might be more meaningful than the value itself. In the study of the relationship between crustal vertical movement and earthquake activities, its generally considered that earthquakes are likely to occur along deformation gradient zones, especially in the intersection of active tectonics, the transitional zone of the deformation gradient belt and areas with strong crustal differential movement. The deformation gradient is the ratio of the difference of vertical deformation between two adjacent points to the distance between the two points, reflecting the degree of deformation changes in unit distance, therefore, the gradient of vertical deformation velocity can reflect the intensity of crustal tectonic activities (Guo Liangqian et al., 2007; Zhang Yingzhen et al., 1992; Zhang Jing et al., 2013).

The vertical deformation velocity gradient can reflect the spatial-temporal range of crustal tectonic activities, depending on spatial-temporal range of the observation data. Because there has been no large scale level monitoring data in recent years, we cannot get the latest achievement of gradient of vertical deformation velocity. The currently available result of the vertical deformation velocity gradient obtained from regional leveling data is calculated by the use of the national leveling velocity results from the periods 1951-1982 and 1951-1990 (Zhang Yingzhen et al., 1992; Zhang Zusheng et al., 1996; Guo Liangqian et al., 2001), and the leveling deformation data utilized was from 25 years ago, which may not be able to reflect the current situation of crustal deformation gradient changes. Since 2010, with the accumulation of data from the geophysical observation project of the China Earthquake Administration, we have obtained the latest leveling observation results in areas such as Yunnan, Sichuan, Shaanxi, Gansu and Ningxia, along with leveling data accumulated from routine earthquake monitoring along the northeast edge and the North-South Belt in recent years, which has to a great extent renewed the leveling data along the North-South Seismic Belt.

1 THE RESULTS OF THE VERTICAL CRUSTAL MOVEMENT VELOCITY GRADIENT

Since the 1990s, with the rapid development of GPS space surveying techniques and the continuous accumulation of monitoring data, the use of GPS technology has helped us get a wealth of horizontal crustal movement information in the Chinese mainland (Zhang Peizhen et al., 2002; Wang Min et al., 2003, Jiang Zaisen et al., 2003). The use of constantly updated GPS results makes it possible to study the relationship between dynamic changes of horizontal crustal movement and earthquakes. Current research related to earthquake forecasting mainly manifests in the studies of the relationship between the locations of strong earthquakes and horizontal crustal movement difference area, strain accumulation area and their dynamic changes (Guo Liangqian et al., 2011; Jiang Zaisen et al., 2007; Yang Guohua et al., 2007). With the continuous improvement of GPS observation and processing technology, the precision of GPS in calculated results of vertical deformation has also been significantly improved. Liang Shiming et al. (2013) calculated GPS vertical velocity field in the Tibetan Plateau and its east edge using GPS observation data from 1998-2013, but has not used the GPS vertical results to calculate the gradient of vertical deformation velocity results till now.

Vertical results obtained from regional leveling data and GPS data have their own characteristics: ① Difference in the spatial distribution of the observation points. The results of vertical deformation velocity obtained from leveling data are limited by leveling monitoring networks. Leveling monitoring networks are sparser in West China than that in the eastern region, while GPS stations are more densely distributed in the eastern margin of the Tibetan Plateau; ② Different accuracy. The first-order precision leveling observation can have vertical precision of 0.1mm, while although the precision of GPS in horizontal direction can reach the millimeter scale, its accuracy in the vertical direction is poor; ③ Different data accumulating time. GPS emerged in the late 1990s, and so far its accumulated data only has a history of more than ten years, while leveling observation in China started in the 1950s, and data is easily collected and arranged and can be traced back to the 1970s. Therefore, the combined use of spatial and temporal distribution characteristics of leveling data and GPS vertical results, complementing and verifying each other, to comprehensively explore the gradient of vertical deformation velocity along the north-south belt, enables us to have a more comprehensive understanding of the tectonic movement along the North-South Seismic Belt.

1.1 Data Application

In this article, the vertical movement velocity field used to calculate the gradient of vertical deformation rate is based on more than 40 years of (1970-2011) high precision leveling observation data, using vertical movement velocity results obtained from GPS stations in the study area as prior constraints, which in combination with the linear dynamic adjustment model is applied to deal with the acquired crustal vertical movement velocity field along the North-South Seismic Belt (Hao Ming, 2013). The vertical deformation velocity field obtained by GPS surveying takes ITRF2008 as the reference framework, and makes use of the GPS vertical velocity field in the Tibetan Plateau and its eastern margin calculated from GPS observation data from 1998-2013 (Liang Shiming et al., 2013). Both leveling and GPS results have excluded the effects of great earthquakes, and thus can reflect long-term crustal tectonic movement features, and leveling lines and GPS survey stations used in the study are spread almost all over the North-South Seismic Belt (Fig. 1).

 Fig. 1 The North-Sorth Seismic Belt leveling, GPS roadmap
1.2 Calculation Method of Gradient of the Vertical Deformation Velocity

The vertical deformation gradient is dimensionless, reflecting the degree of deformation changes in unit distance. Vertical deformation gradient is directional, and gradient values vary in different directions at the same point. When calculating gradient values at each point, we only take into account the larger gradient value, ignoring its direction (Zhang Yingzhen et al., 1992). Specific calculation steps are as follows:

(1) Using the Kriging interpolation method, we interpolate the vertical movement velocity field calculated from leveling lines and GPS survey stations in Fig. 1, and get a velocity value of 10′×10′ grid points;

(2) Using velocity value of grid points to calculate the horizontal gradient of velocity by the following method. According to the scalar function z=(x, y), the horizontal gradient is defined as follows:

 $\text{grad}Z=\frac{\partial Z}{\partial x}\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\rightharpoonup}$}} {i}+\frac{\partial Z}{\partial y}\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\rightharpoonup}$}} {j}$ (1)

Where, i, j denote the unit vector on the x, y axes respectively, so the module of horizontal gradient is

 $\text{ }\!\!|\!\!\text{ grad}Z|=\text{sqrt}\left[ {{\left(\frac{\partial Z}{\partial x} \right)}^{2}}+{{\left(\frac{\partial Z}{\partial y} \right)}^{2}} \right]$ (2)

Partial differential of function Z can be approximately expressed as:

 ${{\left(\frac{\partial Z}{\partial x} \right)}_{ij}}=({{Z}_{i+1, j}}-{{Z}_{i-1, j}})/2\Delta x$ (3)
 ${{\left(\frac{\partial Z}{\partial y} \right)}_{ij}}=({{Z}_{i+1, j}}-{{Z}_{i-1, j}})/2\Delta y$ (4)

That is, the gradient of grid point pij in x direction, ($\partial Z/\partial x$)ij, is calculated using the field value Zi-1, j and Zi+1, j of two adjacent grid points pi-1, j and pi+1, j in x direction. Similarly, we get the gradient of grid point pij in the y direction.

1.3 The Results of Vertical Deformation Velocity Gradient

 Fig. 2 The North-Sorth Seismic Belt vertical deformation velocity gradient (leveling)

 Fig. 3 The North-Sorth Seismic Belt vertical deformation velocity gradient (GPS)

Table 1 Faults located in the high gradient zone

The value of vertical deformation velocity gradient reflects the intensity of crustal tectonic activities. Slow vertical deformation velocity gradient zones are mostly located inside the blocks, reflecting weak relative tectonic movements inside the blocks. From the view of distribution of high gradient zones, the intersection of the Longmenshan fault zone, the south segment of Xianshuihe fault zone and the Anninghe-Zemuhe fault zone has the most intense tectonic movement along the whole North-South Seismic Belt, with a maximum deformation velocity gradient of 97×10-9/a, which is consistent with the results achieved by Guo Liangqian et al. (2007). Secondly, the junction of the Zhuanglanghe fault zone near the mid-east segment of Qilian Mountain and the Lenglongling fault zone has the most intense tectonic movement on the northeastern margin, with the maximum deformation velocity gradient of 60×10-9/a. The recent Menyuan MS6.4 earthquake on January 21, 2016 took place in the high vertical deformation gradient zone based on leveling data and on the margin of the high gradient zone based on GPS data.

2 THE RELATIONSHIP BETWEEN THE VERTICAL DEFORMATION VELOCITY GRADIENT AND STRONG EARTHQUAKE ACTIVITIES

Previous research shows that the high vertical deformation velocity gradient zone might be a potential strong earthquake risk zone (Guo Liangqian et al., 2007;Zhang Yingzhen et al., 1992;Zhang Zushenget al., 1996;Guo Liangqian, Bo Wanju et al., 2001). Although there has been leveling data since 1970, considering the differences of data accumulating time and monitoring range in time and domain, in the analysis of correlation between vertical deformation velocity gradient and earthquake activities, we do not analyze earthquakes from 1970-1985 with leveling data, and do not analyze earthquakes occurred before 2000 with GPS data.

Strike-slip earthquakes, such as the Kangding MS6.3 earthquake which occurred along the Xianshuihe fault zone in 2014, were located in the sub-high gradient zone based on both leveling and GPS data. The Yushu MS7.1 earthquake in 2010 occurred on the margin of a sub-high gradient zone based on leveling data. The Yingjiang MS6.1 earthquake in 2014 took place on the margin of a high vertical deformation gradient zone based on GPS data, but its epicenter was located outside the leveling monitoring area. The Yongsheng MS6.0 earthquake in 2001 occurred in a sub-high gradient zone based on both leveling and GPS data. From 2000-2009, earthquakes with magnitude of 6.0-6.4 occurring near Yaoan and Dayao did not respond to both leveling and the GPS vertical deformation velocity gradient, and the Wuding MS6.5 earthquake in 1995 also did not occur inside or on the margin of a high or sub-high gradient zone based on leveling or GPS data.

It can be seen that since 1980, almost all MS≥7.0 earthquakes have been related to spatial distribution of a high or sub-high vertical deformation velocity gradient zone, but were also associated with the duration of data accumulating time and spatial distribution of survey lines or survey stations. Earthquakes near survey lines or survey stations with longer data accumulating times basically took place in the high gradient zone, such as the Wenchuan earthquake and the Lushan earthquake. Earthquakes near the margin of leveling networks with short data accumulating time or sparse survey points occurred in the margin of high or sub-high gradient zones, such as Langcang-Gengma earthquake, Lijiang earthquake and the Gonghe earthquake. Except for the above-analyzed MS≥7.0 earthquakes, regardless of whether there were strike-slip earthquakes or thrust or normal earthquakes, of the 15 earthquakes with magnitude of 6.0-6.9 along the North-South Seismic Belt we analyzed, 11 took place inside or in the margin of a high or sub-high deformation velocity gradient zone. However, not all MS≥6.0 earthquakes are related to the high vertical deformation gradient, such as the Yaoan earthquake, the Dayao earthquake, the Ninglang earthquake and the Wuding earthquake, of which, except for the Ninglang earthquake, which is normal type, the other three are all strike-slip earthquakes.

3 CONCLUSIONS

Vertical deformation velocity gradients along the North-South Seismic Belt calculated by using two different measuring methods, regional leveling and GPS surveying, are basically consistent. By comparison with MS≥6.0 earthquakes, it is believed that spatial distribution of MS≥6.0 earthquakes along the North-South Seismic Belt is in good coincidence with the regional high deformation velocity gradient zone in space, especially MS≥7.0 earthquakes, as according to the distribution of survey points and survey lines, most earthquakes occurred inside or on the margin of the high gradient zone, and for MS≥6.0 earthquakes, the vertical deformation velocity gradient still shows good anomaly features before earthquakes.