An inversion method, based on a genetic algorithm, is proposed for interpreting jointly various kinds of stress data in order to overcome the limitation, in number and quality, of each data set. The method has been applied to results from hydraulic tests in boreholes and to focal mechanisms of induced seismicity observed within the same depth interval. The regional stress field is described by two symmetrical tensors. The first one represents the stress at a given depth and the second one the vertical stress gradient. Results indicate that one of the principal directions is vertical. They are consistent with all but one of the hydraulic tests considered in the inversion and with about 70% of the focal mechanisms. This inversion confirms previous results suggesting that, for the scale of these induced microseismic events, the regional stress field cannot be determined solely from an inversion of the fault plane solutions.

a b s t r a c tThis paper presents a modified method of discontinuous deformation analysis (DDA). In the presented method, open-close iteration may not be needed, small penetration is permitted among blocks, and springs are added between contacting block pairs only when a penetration takes place. The three contact patterns (i.e. sliding, locking and opening) in original DDA method are not involved, and the recognition of these contact patterns and treatment of transformation among patterns are not required either, significantly saving the computing time. In a convex to concave contact, there are two candidate entrance edges which may cause uncertainty. In this case, we propose the angle bisector criterion to determine the entrance edge. The spring stiffness is much larger than Young's modulus in the original DDA, however we find that the correct results can still be obtained when it is much smaller than Young's modulus. Finally, the penetrations by using penalty method and augmented Lagrangian method are compared. Penetration of the latter is 1/4 of the former. The range of spring stiffness for the latter is wider than the former, being 0.01e1 of the former. Both methods can lead to correct contact forces.

The movement and deformation characteristics of an active fault zone are an important prerequisite for the stability analysis of cross-fault project sites and breaking resistance design research. Based on the global positioning system (GPS) data of the Crustal Movement Observation Network of China, a GPS velocity clustering algorithm to avoid the uncertainties of block division results from geological data is proposed and applied for velocity field analysis of Yunnan province. Using the data of active faults in the central Yunnan water diversion project area and the clustering results along with a block rotation and uniform strain model, the current slip rates of the main fault were obtained by using the difference between the speeds of the same point at the boundary of neighboring blocks. Furthermore, using the 1999–2007, 2009–2013, and 2013–2015 GPS velocity field data as the constraints, the relation between the spatial points and velocities was established based on the least squares collocation theory. The distribution of strain rates in the central Yunnan water diversion project area were resolved by the spherical strain calculation method, and the dynamic variation characteristics were analyzed using regional historical earthquakes and fault properties. The results of this study show that Xiaojinhe–Lijiang and Xiaojiang faults currently have noticeable strike–slip motion. The southern Xiaojinhe–Lijiang and Xiaojiang faults have always been two strong deformation zones, and the extensional deformations in Longpan–Qiaohou, Lijiang–Jianchuan, and Chenghai–Binchuan regions are being enhanced. Xiaojiang fault is the zone with the maximum shear strain rate. These results indicate that the deformation distribution characteristics are closely related to the characteristics of the activity features of the major faults in the region.

No abstract