Introduction Background
Download 0.84 Mb.
|
- Navigate this page:
- Figure (4. 5)
- Figure (4. 6)
- Chapter Five
Modeling of coseismicIn this section, the co-seismic surface deformation data are inverted to determine the rupture's geometry and slip distribution by solving for a rectangular dislocation in a homogeneous elastic half-space with a Poisson's ratio of 14:0:25. The Green's function was used to simulate the relationship between unit slip and surface displacement (Okada, 1985). In the coseismic slip inversion, fault geometry includes the fault position, strike, dip, and rake angles. As a result, when the data quality and quantity are limited, they are frequently not well constrained, which could cause a bias in the slip distribution that results. This article assumes a single rectangular fault Model parameters in this inversion include the location of the fault centroid (eastward and northward shift (x, y) concerning the epicenter of the earthquake and the depth, length, width, strike, dip, and rake, as well as the slip magnitude of the dislocation. The inversion is implemented in a Bayesian framework to quantify the uncertainty of model parameters. A uniform prior distribution within a wide range has been assumed for each parameter in the model. The model space is sampled using a Matlab slice-sampling algorithm (Cervelli et al., 2001). It patches to the first invert for the fault geometry of the 2017 Sarpol Zahab earthquake. Most model parameters are tightly constrained because InSAR observations from various look directions have good coverage. However, the acceptable range of model parameters depends on the input data's error functions, which were estimated using data outside the deformation area under the simplified assumptions that the atmospheric noise is spatially homogeneous, isotropic, and exponentially decays with distance. The findings indicate the 2017 Sarpol-e Zahab earthquake rupture can be roughly described by a fault plane that is 30 km long, 30 km wide, and gently dips to the east (dip angle =11.2°), with an average slip of 345 cm and rake angle of 135°. The fault plane has a strike angle almost Northwest- Southeast (strike = 337.5ͦ). The slip centroid is discovered to be about 20-40km along the strike distance southwest of the USGS epicenter, at a depth of about 10-20 km. As expected, the slip magnitude and the fault's dimensions, particularly the width and depth, are traded off to some extent. Additionally, there is a slight trade-off between the strike and rake angles. The preferred strike of the 2017 Sarpol Zahab rupture, on the other hand, is similar to the overall orientation of the Mountain Frontal Flexure, which separates the high Lurestan arc to the east from the low Kirkuk embayment to the west near the epicenter area (Berberian, 1995). The preferred fault geometry and slip direction agree well with the focal mechanism determined by USGS and the moment tensor solution (Figure 4.5a) listed in Table 4.1. Overall, the surface displacements predicted by the preferred model of a single dislocation patch correspond well to the observations (Figure 4.5b). We assess the broad-scale slip distribution of the 2017 Mw 7.3 Sarpol Zahab earthquake using the fault geometry determined by the single dislocation inversion shown above. Figure (4. 5): The images show the superposition fault and slip (a) Inferred Fault (b) Coseismic slip model of the 2017 Sarpol‐e Zahab earthquake. Table `4.1: The best-fit faulting parameters of the 2017 Sarpol Zahab earthquake inverted from seismology and line-of-sight displacements 1USGS United States Geological Survey; 2Global CMT stands for Global Centroid Moment Tensor Project; and 3 The various model fault planes derived from co-seismic interferograms' elliptical fringe pattern; a Each fault patch determines the mean rake direction; b The longitude, latitude, and depth are defined as the fault plane's centroid. To ensure a uniform model resolution, the fault plane, which is 30 km long by 30 km wide, is divided into patches whose size gradually increases along the down-dip direction. The dextral-lateral slip components for each patch may be up to 8 m. Laplacian smoothing is applied between adjacent fault patches to prevent abrupt variations in slip. In addition, we have regularized the inversion problem by requiring no slip on fault edges, except at their up-dip edges. The model of a single dislocation patch-s exhibits oblique slip and nearly equal amounts of dextral and thrust slip. However, the distributed slip model predicts a more giant slip in the rupture's southern. Approximately 35 km by 15 km is the size of the prominent slip area (>1 m). The majority of the moment release occurs between 10 and 20 km below the surface, with a maximum slip of 5.3 m at 14 km, well below the estimated 6 to 15 km thickness of sedimentary cover in the Lurestan arc (Emami et al., 2010); (McQuarrie, 2004); (Vergés et al., 2011). The base of the seismogenic zone in this area is closely aligned with the bottom of the coseismic slip model (Karasözen et al., 2019). The total moment release is estimated to be 1.15 × 1020 Nm, corresponding to a moment magnitude of 7.31, which agrees with the seismic moment. The obtained geodetic moment was consistent with the seismic moment identified by the USGS. Surface displacements predicted by the preferred slip model are agreed well with observations (Figure 4.6). The residual was large signals identified from the north-south components (Figure 4.6 f, l). The east-west and up-down components were distinguished by fewer residual signals (Figure 4.6 d, e, j and k). Based on the result with a rectangular dislocation, the model with a variable slip distribution yields a better fit for the observations, particularly in the area south of the moment centroid, where slip is larger than average, as shown in Figure 4.5b. The interpretation of the study is consistent with (Barnhart et al. 2018).On the other hand, (Feng et al. 2018), propagated their source model towards the surface and suggested the Khanaqin fault as the source fault. Figure (4. 6): The images show the model prediction and residuals for 3D observations. The left-hand image shows the East-West component. The medial image shows the up-down componen. The right image shows the North-South component and 3D ALOS-2 model prediction (a–c) and the residual (d–f). Chapter Five Download 0.84 Mb. Share with your friends:
|