نشریه پژوهش های ژئوفیزیک کاربردی
پیاپی 2 (پاییز و زمستان 1394)

Summary For the present two-dimensional (2-D) lithospheric density-temperature distribution modeling, we have selected a north-south geo-transect cut in east of Iran from the Oman Sea to the Lut block with a length of about 800 km. As the structural domains, investigated in this paper, have mainly been created under the effect of the north-south subduction of the Arabian Oceanic Plate under the Eurasia, thus, most of these structures are approximately perpendicular to this direction. As a result, in order to avoid three-dimensional (3-D) structural effects in our 2D modeling, we have selected a profile perpendicular to the strike, which is roughly the north-south direction.
Introduction In order to avoid interpreting isolated anomaly features being located accidentally on the profile of the geoid, free air and topography data, we average the data within a width of 50 km to both sides of the profile and calculate the variance within this stripe using 5 km long steps along the profile. The resulting standard deviation is plotted as uncertainty bars. Since the lateral resolution of our data is of the order of several kilometers, therefore, we cannot reproduce small‐scale details. The generated model consists of a sedimentary layer, the upper, middle and lower crust and a lithospheric mantle layer. In order to constrain the various thicknesses, the results of previous works carried out in the study region have been used.
Methodology and Approaches Modeling carried out in this research is based on the assumption of local isostatic equilibrium and joint interpretation of gravity, geoid and topography data with temperature-dependent density. Bouguer anomaly data are mainly sensitive to crustal density distribution and decrease proportionally to the squared distance (r) from an object, whereas elevation is very sensitive to changes in density and thickness of the lithospheric mantle and the geoid undulations diminish proportionally to r−1. The data different sensitivity on distance from the object can be a good constraint on modeling lithospheric density-temperature distribution.
Results and Conclusions Because of subducting Arabian plate in Makran subduction zone, thickness of the lithosphere increase from the Oman Sea (90 km) to the Taftan-Bazman volcanic arc, with a maximum thickness (290 km) under the volcanic arc. Based on our modeling results, subduction of the oceanic crust of Arabian plate under the Makran belt starts with a very low dip angle and bends under the Taftan-Bazman volcanic arc with a relatively higher dip. There is a logical correlation between the subducted slab and the seismicity of Makran subduction zone. Because of the low dip subducting slab, there is a long distance between volcanic arc and forearc setting (~450 km). Further to the north, lithospheric thickness is reduced to the ~200 km under the Lut block which is in good agreement with earlier studies. Our integrated modelling approach reveals a crustal configuration with the Moho depth varying from ∼21 km beneath the Oman Sea, and ∼47 km beneath the Taftan-Bazman volcanic arc to about 37 km beneath the Lut block. Across the Makran subduction zone, Moho depth is increasing from the Oman seafloor and Makran forearc setting (~21 km) to the volcanic arc (∼45–50 km). Furthermore, we have found that the oceanic crust is a two layered plate and there is a thick sedimentation there to about 8-9 km.

Summary Radon transform is a useful tool in seismic data processing with lots of applications. However, incomplete information decreases its resolution and limits its applicability. Sparseness is a valid criterion to overcome this problem. In this paper, a forward-backward splitting algorithm is used to solve an l1-norm regularized Radon transform in order to obtain a high resolution linear Radon transform (LRT). Then, it is applied on down-up going wavefield separation in vertical seismic profiling (VSP) data and also seismic interference (SI) noise attenuation. It is also shown that a sparse LRT can be used to enhance the quality of teleseismic data.
Introduction Overlapping wavefields in time-space domain can be separated in Radon domain, however, different reasons cause resolution problems and aliasing in Radon transform. Obtaining a sparse Radon panel enables us to separate coefficients of each event without harming the others. Using sparsity promoting methods results in a high resolution Radon domain with minimum non-zero elements. An l2-l1 norm cost function is constructed and a forward-backward splitting algorithm is employed to obtain a sparse model, which can be used for different purposes among which VSP wavefield separation, SI noise attenuation and wavefield enhancement are dealt with here. Both down- and upgoing wavefields of VSP data contains information of subsurface but they need to be separated. Mapping their coefficients to separable regions using a sparse linear Radon transform is an efficient solution. One other application could be SI noise attenuation. This kind of marine noise is originated by other surveys in the same area and overpowers reflections. According to their linear characteristics and the use of sparse LRT, they can be subtracted from the data. Moreover, it is shown that injecting sparsity to Radon domain eliminates random noise and also non-linear events. Furthermore, low quality traces can be replaced by interpolated new reconstructed traces applying a mask on misfit term in the cost function. This can be used to enhance the quality of teleseismic data with linear wavefields.
Methodology and Approaches The ability of separating coefficients in Radon domain relies on the resolution of Radon transform. Least square regularization results in a smooth solution, which causes the application limitation. Here, a sparse LRT is developed by applying an l1-norm constraint on Radon model. This transformation can be applied on VSP data to map upgoing and downgoing wavefields to negative and positive slowness regions, respectively. Then, each of them can be reconstructed separately using inverse Radon operator. High amplitude seismic interference noise is harmful to many processing steps and they need to be attenuated beforehand. Their linear coherency in shot gathers makes LRT an appropriate tool to eliminate them. Using a sparse LRT converges SI plane waves energy, however, attenuates reflections coefficients. Thus, only SI noise can be reconstructed, and then, subtracted from marine shot gathers. Sparse data acquisition in teleseismic data, and also, the presence of random noise reduce the quality of the data and make arrival time picking difficult. Sparsity in Radon panel and interpolation increases the resolution of the reconstructed data so that the P- and S-wave arrival times can be determined with more accuracy.
Results and Conclusions Synthetic and real numerical examples demonstrated the efficiency of applying a forward-backward splitting algorithm to solve an l2-l1 cost function for obtaining a sparse Radon panel in which even linear wavefields with very close slopes can be identified. This method is a feasible and effective way to separate overlapping upgoing and downgoing VSP wavefields, and to attenuate the seismic interference noise and to enhance teleseismic wavefield.

Summary In recent years, due to increasing the size of three-dimensional (3-D) seismic data and the number of seismic attributes, it is difficult for interpreters to examine every seismic line and time slice. Pattern recognition techniques as first-hand interpretation tools are used to both address the problem of large size of seismic data and to provide an initial guidance when working on a new seismic data where previous studies and data are limited. Different types of unsupervised learning techniques have recently been used for seismic facies clustering and object detection in seismic data. Among unsupervised learning techniques k-means, self-organizing maps (SOM), generative topographic mapping, and principal component analysis (PCA) are used for facies analysis. In this study, we have applied k-means, SOM, and PCA on a 3-D seismic data volume acquired over the Strait of Hormuz to detect buried channels. Not surprisingly, the most important parameter in this study is the choice of appropriate seismic attributes. Although the PCA is not a clustering technique, it can detect channels in 3-D seismic data more efficiently than the k-means and SOM. According to the dip of the structure, the detected channels are prolonged from the west to the east and the southeast where there is a mini basin within the Mishan Formation.
Introduction One important class of machine learning tasks is the unsupervised learning. In the unsupervised learning, no labels are given to the learning algorithm, leaving it on its own to find structure in its input. The main task of this learning method is data clustering, but some different tasks such as dimensionality reduction and density estimation are belonged to this category. PCA is a dimensionality reduction technique, which can be used for better visualization of data. After explaining the geology of the study area, we discuss the learning methods, and their workflows. In the next step, we present the chosen attributes, and the learning algorithms applied to the data.
Methodology and Approaches We have used OpendTect for attribute measurements. After preparing the data, we have applied three unsupervised learning techniques of k-means, SOM, and PCA, on attributes of the 3-D seismic data volume acquired over the Strait of Hormuz. The chosen attributes in this study are spectral decomposition, curvedness, and gray-level cooccurrence matrix (GLCM) homogeneity. First, we have applied the PCA to reduce the dimension of attribute data to 2-D, and then, the k-means and SOM are applied on the data. Next, we have presented the two first principal components of attributes to the RGB color system, and consequently, we found that this method is superior than the k-means and SOM in the illumination of the channels.
Results and Conclusions Although the PCA method is not a clustering technique, it can detect channels in 3-D seismic data more efficiently than the k-means and SOM clustering methods. According to the dip of structure, these channels are prolonged from the west to the east and to the southeast where there is a mini basin within the Mishan Formation.

Summary Reflection seismology is one the exploration geophysical methods which uses the principles of seismology to estimate the properties of the earth subsurface from reflected seismic waves. It is a type of geophysical imaging technique used to image the subsurface of the earth and understand its geology. Seismic data acquisition always is accompanied by variety of noises. Therefore, improvement of the signal-to-noise ratio (SNR) of seismic data is a sticky thread in the processing of seismic data for high quality imaging. Many efforts have been made to remove this type of noise from seismic data and several authors have suggested different methods for this purpose. Each of these methods has its own advantages and disadvantages. In this paper, we have used non-linear diffusion filter to attenuate the reflection seismic random noise. The results pointed out in this paper show significant improvement of SNR through enhanced reflector continuity for a better interpretation. They show that the diffusion filter produces acceptable results compared to the well-known f-x deconvolution and median filter.
Introduction Seismic data are affected by various types of noises such as ground rolls, multiples and etc. Generally, seismic noises are divided into two categories: coherent and incoherent. Unlike the coherent noise, incoherent noise known as random noise, is not correlated from trace to trace. Random noise resulted from random oscillation during acquiring data is one of the most important and harmful noises that exist in seismic data over entire time and frequency. Therefore, improvement of the signal-to-noise ratio (SNR) of seismic data is a sticky thread in the processing of seismic data for high quality imaging. Many efforts have been made to remove this type of noise from seismic data and several authors have suggested different methods, such as f-x filtering, polynomial fitting, and singular value decomposition, to remove random noise from seismic data. Each of these methods has its own advantages and disadvantages. A significant part of the seismic random noise can be attenuated by stacking procedure, but part of random noise also remains after stacking. Attenuation of remaining part of noise requires an advanced method. Smoothing filter is one of the most efficient and traditional type of random noise suppression in signal and images. Diffusion filter is one of the most important smoothing filters, which is based on a physical concept named as diffusion.
Methodology and Approaches The diffusion defines a physical procedure for balancing concentration changes without creating or destroying mass. The analogy with image processing can be drawn if we consider the concentration as the intensity of the image. The diffusion coefficient or diffusivity function, which adaptively controls the smoothing amount is one of the most effective parameters in diffusion process. If the diffusion coefficient is selected as a constant value, the diffusion process will be linear. In other words, the smoothing procedure will be performed as homogeneous and isotropic and causes the output image to be blurry. If the diffusion coefficient is a function of image gradient, the diffusion process will be scalar. In these circumstances, the smoothing procedure will be performed as inhomogeneous and isotropic. Scalar diffusion filter has better efficiency than the linear form; however, it has not good performance in the edge position due to its isotropic property. Anisotropic non-linear diffusion filter is a proper tool to filter images while preserving details and even enhancing edges. The diffusion coefficient of an anisotropic non-linear diffusion filter is defined as tensor. The basic idea of the anisotropic non-linear diffusion filter is to preserve edges. This is attained by orienting the diffusion process parallel to the edge.
Results and Conclusions The anisotropic non-linear diffusion filter is a powerful method for image de-noising and edge preserving. The efficiency of the diffusion filter was tested on both real and synthetic seismic data. The method effectively recovered seismic signals on noisy synthetic and field seismic data. The obtained results were compared with the results of the well-known f-x deconvolution and median filter. After analyzing all of the results, the anisotropic non-linear diffusion filter proved to be an alternative algorithm for seismic data de-noising and had the best performance among the other three filter methods.

Summary The main objective of this research work is to investigate the subsurface structures in Sabalan geothermal field using two-dimensional (2-D) nonlinear conjugate gradient and smoothness-constrained least square inversion methods of magnetotelluric (MT) data. To achieve the goal, 8 MT sites in the Sabalan area have been selected along one profile. Dimensionality analysis using the WALDIM code and impedance polar diagrams indicated that the subsurface structures of the selected sites are One-dimensional (1-D) or 2-D at shallow depths, whereas they are mainly three-dimensional (3-D) at middle and lower depths. Tipper and induction arrows have been used to determine the geoelectrical strike. As a result, NE-SW direction has been identified as the major strike direction of subsurface structures in the area. Spatial averaging method has also been used to remove the galvanic distortions. After setting up the modeling parameters and appropriate mesh design, 2-D inversion of the MT data has been performed using nonlinear conjugate gradient and smoothness-constrained least square methods. The obtained 2-D models have illustrated very well the geothermal structures including cap rock, reservoir and the heat source. Moreover, there is a close correspondence between these models and past investigations in the area.
Introduction The Sabalan geothermal field is located on the western slopes of Sabalan mountain in the Moil valley in Ardabil province. Warm and hot springs w are found within the Moil valley in the area. The four major units, originally identified and used for the original geologic study and mapping at northwest of Sabalan, are in order of increasing age: Quaternary alluvium, fan and terrace deposits; Pleistocene post-caldera tracheyandesite flows, domes and lahars; Pleistocene syn-caldera tracheydacitic to tracheyandesitic domes, flows and lahars; Pleistocene pre-caldera tracheyandesitic lavas, tuffs and pyroclastics. Fractures and faults can play an important role in geothermal fields, as fluid mostly flows through them in the source rocks. Structural study confirms two major types of structural setting: a set of linear faults and several ring-faults. Sabalan area has been under studies including geology, hydrology, geochemistry and resistivity survey since 1998. Three deep exploratory wells were consequently drilled from 2002 to 2004 within the moil valley. Measured temperature at wells drilled in the area varies between 160 to 250ºC. In 2006, a hydrological model of the region has been proposed. In 2007 and 2009, considering the proposed hydrological model, MT survey was conducted to determine the possible center of the geothermal system and identify the drilling targets. In this research work, the subsurface structures in Sabalan geothermal field is investigated using 2-D nonlinear conjugate gradient and smoothness-constrained least square inversion methods of the acquired MT data from 8 MT sites in the area. Based on the inversion results, the geoelectrical model or specifications of subsurface structures in the area will be estimated.
Methodology and Approaches To investigate the subsurface structures in the Sabalan area using 2-D nonlinear conjugate gradient and smoothnessconstrained least square inversion methods of MT data, 8 MT sites were selected from 78 MT sites measured in 2007 and 2009 MT survey, and then, the MT data of these 8 sites were used in the inversion process. The survey utilized a Phoenix MTU-5A data acquisition system, which is a standard five-channel system recording both two electric and three magnetic field components. The raw time series data were processed using SSMT2000 software. The processed data were then used to obtain the impedance tensors from which the apparent resistivity and phase data in a frequency range of 0.0011-320 Hz were calculated. WALDIM code and Polar diagrams were used for dimensionality analysis of MT data. Static shift correction based on the similarity between the MT curves of adjacent sites was utilized to remove the effects of galvanic distortions. After dimensionality analysis and static shift correction, and then, setting up the modeling parameters and appropriate mesh design, we performed 2-D inversion of the MT data by employing nonlinear conjugate gradient and smoothness-constrained least square methods in MT2InvMatlab code and WinGLink software. Because the fitting of the transverse magnetic (TM) mode data in the 2-D inversion can be more easily achieved even for a 3-D structure, often the TM mode data is only used for 2-D inversion of MT data in geothermal fields. Therefore, we have only shown the inversion results of TM mode data and interpretation has also made for the obtained model of the TM mode data.
Results and Conclusions As a result of the inversion of the MT data, 2-D models of subsurface geoelectrical structures have been obtained, and consequently, the cap rock, reservoir and heat source of Sabalan geothermal field have been identified. Moreover, the ACB algorithm introduced by Yi et al. (2003) has been used to stabilize the inversion process and provide a highresolution optimal earth model in the MT2InvMatlab code. Constructing the forward and inversion mesh in the MT2InvMatlab code is also made automatically.

Summary The semi-automatic techniques play an important role in interpretation of potential field data. There are many semi-automatic interpretation techniques available for interpretation of gravity data, such as Werner deconvolution (Hartman et al., 1971) and Euler deconvolution (Thompson, 1982; Reid et al., 1990). Some methods (Hsu et al., 1998; Keating and Pilkington, 2004; Salem, 2005) use the analytic signal amplitude. Other methods involve use of the gradient ratio of the potential field data and often give the edges of the bodies. In this paper, the gradient ratio method has been applied on gridded gravity map data for simultaneous estimation of the location and depth of the subsurface bodies. This method has been applied on synthetic gravity data as well on high resolution real gravity data acquired from Shavaz Iron Ore in southwest of Yazd province.
Introduction Circular structures in gravity data may correspond to kimberlite pipes or to meteorite impact structures. They can be modelled in different ways such as inversion. The horizontal gradient of a potential field dataset in a given direction enhances linear features which trend at 90° to that direction, while diminishing those that lie parallel to that direction. Such directional derivatives are commonly applied to potential field data to enhance or diminish features such as dykes, faults, and trends. These derivatives are easily computed in the space or frequency domains. One of the filters, in which gradient ratio factor is used, is tilt angle (Miller and Singh 1994). Tilt angle is an attribute of potential field data and is becoming increasingly popular among geoscientists for interpreting potential field, especially, the magnetic data. Salem et al. (2007) obtained tilt angle by substituting the analytic expressions for horizontal and vertical derivatives of the magnetic field over a vertical contact. This paper generalizes this idea and applies it to gravity data related to vertical cylinder and buried sphere models.
Methodology and Approaches The interpretation methods existed for circular features are restricted to edge detection only. This paper generalizes tilt-depth method for the magnetic vertical contact model to gravity data of vertical cylinder and buried sphere models. For simultaneous depth and edge estimation of such structures, we can use from gradient ratio method. The meaning of the gradient ratio is the ratio between vertical and horizontal derivatives. The method computes the ratio of the vertical derivative to the total horizontal derivative of data, and then, identifies circular contours within it. Having the radius of the contour and the contour value itself, we can determine the depth to the source. In this regard, a second order analytical equation is developed in which solving it leads to estimation of the depths of circular features. This method can be classified into model-dependent methods. In this paper, vertical cylinder and sphere have been used as synthetic models.
Results and Conclusions The following results and conclusions can briefly be drawn from the study made in this paper: 1. The ratio between vertical and horizontal gradients of the gravity anomaly related to cylinders can be used to estimate depth and location of the subsurface bodies. 2. The problem of noise enhancement in the computation of the gradients can be solved via the following two approaches: - acquiring the vertical and horizontal gradients in the field by gravity gradiometer or - calculating the vertical derivatives (more capable in noise amplifying) by means of Hilbert transform. The Hilbert transform is more stable than fast Fourier transform (FFT) and the results are more consistent. 3. The method has been tested on synthetic gravity data and on real gravity data from Shavaz Iron ore in southwest of Yazd province.