Journal of the Earth and Space Physics
Year:2005 | Volume:31 | Issue:1|Papers Count:12

Investigation of the step variable numerical integration methods is the main purpose of this paper. Also it is shown how the one-dimensional numerical integration method of solving an ordinary differential equation can be employed to solve a three-dimensional differential equation numerically. In order to integrate the orbit, the step variable methods of Runge-Kutta-Fehlberg and Adams are employed and their accuracies are investigated.

In recent years, several methods for calculating the seismic wave’s traveltime have been presented (Moser 1989, Vidale 1988). Among them, the finite difference ray tracing has given reliable results. In this study in order to calculate the traveltime in a cell model, Vidal’s finite difference eikonal solver was used. To evaluate, different velocity models were tested and in all cases the results were compatible with the theoretical existing basis. Then by the damped least square method and conjugate gradient algorithm, which was introduced by Scales (1987), the inversion of the traveltime residuals for producing a velocity section of the subsurface media was performed. In order to qualify the final section, statistics parameters such as resolution matrix, covariance matrix, uncertainty matrix and rms, were used. This method has been applied on the real data from a weighted concrete dam. At the bottom of the final tomogram a low velocity zone is detected which was attributed to poor quality of the existing rock in that zone. The back projection method was used on the real data to evaluate the method that has been applied in this article. The result obtained from the back projection method compared to the conjugate gradient method was not satisfactory.

To record local earthquakes, the telemetric-digital seismic network of the tabriz comprising of eight three component seismic stations was installed in the north-west of Iran. This array started operation at the end of 1995. Investigation of archived data in the Tabriz seismic network between 1999 to September 2004 shows that magnitude values in the database have not been determined using a single formula or a specific method. Comparison of magnitude values in the database with those values that have been published in the ISC/NEIC bulletin shows that magnitude values calculated in the Tabriz network are underestimated for events occurred in greater distances. By using magnitude values in the database and corresponding mb ISC/NEIC (mb ISC or mbNEIC) values, a scale as M=log (v/4π)+2.6log (Δ)-2.2 is derived for determination of the magnitude in distance range of 170 to 1000 km, where v is peak-to-peak amplitude in micrometer/second and Δ is epicentral distance in kilometers. This formula gives a better estimate for the magnitude of events in comparison with the formula that is presently used in the Tabriz seismic network.

Wind data is usually used to obtain wave statistics in most coastal engineering projects. Here we present the results of wind waves predictions. Wave and wind data have been recorded by a buoy in the Bushehr region in 1995, 1996 and 1999. About 3900 recorded events of this data have been used to study suitability of existing parametric methods for prediction of wind generated waves in the region. Various important parameters such as wave height, wave period, wind speed, wind duration, wind direction and fetch length have been considered for this purpose. This study provides estimates of wave conditions (fetch- limited, duration limited and fully developed sea) and atmospheric stability that influences wind wave growth. Using this data we estimate the non-dimensional energy and non-dimensional peak frequencies as a function of the non-dimensional fetch and Bulk Richardson numbers for this region. These results agree well with similar results obtain by others. The interesting and new aspect of the present study is finding the relationship between the non dimensional energy (ε) and peak frequencies (v) and the bulk Richardson number (Rb ) of the atmospheric surface layer which are: ε=0.091 Rb0.88, v=0.056 Rb-0.3, (0.005< Rb<1) . Such relations could be incorporated in wind wave prediction models for greater accuracy.

This study deals with modeling of gravity anomalies by Compact Inversion Method of Last and Kubic (1983) and Lewi et al (1994). The principle used is to minimize the volume, to maximize of the causative body. The anomalous density distribution is obtained using an iterative technique which is numerically stable and rapidly convergent. This principle can also be adapted to include modeling of gravity anomalies by single density sources. The advantage of this method is that desirable geological characteristics are automatically incorporated into the model with minimum subjective judgments on the part of interpreter. The treatment of noise in data fits naturally into formulations of the inversion procedure. The practical effectiveness of this method is assessed by inversion of synthetic and real acquired over chromite ore in Hormozgan province of Iran.

One of the main problems in geophysics data analysis is the presence of noise. This problem in seismic survey is more obvious than in the other branches. In this study the random noise suppression is presented by a filter which is called eigenimage filter and operates on stacked 3D seismic data in frequency domain. Our tools in suppression of random noises were SVD and Lanczos. The Lanczos method works much faster than SVD, specially when we have a sparse matrix. The special feature of the Lanczos method is its high performance. The F-xy filter has its own abilities such as fantastic signal preservation and well noise suppression, and can be used even before stacking. In this study presented performance of the filter on staked synthetic seismic data.

Rayleigh waves are surface waves which are dispersive in layered media. In order to study Rayleigh wave dispersion, there are several methods. Among these methods the application of matrix methods for elastic layered media is preferred to other methods. In the matrix method for a layered medium, boundary conditions for different interfaces are satisfied to obtain a relation for calculating Rayleigh wave dispersion which should be resolved by numerical methods. A program for the computing of Rayleigh waves dispersion was developed. The resulted dispersion curves from this program werecompared to similar methods which showed good agreement. According to the resulted graphs, wherever the earth structure is complex the dispersion curves also will be complex and different models will be moved oscillatory. Phase and group velocity curves tend to lower values when there is a surface sedimentary layer in the earth model. Thus, synthesis of these curves leads to clarification of different seismic phases, history of waves propagation from source to receiver and consequently imaging of earth structure.

Temperature and wind fields over Tehran were simulated using the MM5 modeling system. To do this a nested domain with three nests with grid spacing of 27, 9 and 3 km were chosen. The model was run for two different days, one with a continental polar air mass (CPA) and the other one with a continental tropical air mass (CTA) over the region. The model outputs for the innermost nest were studied for the first 24 hours of the forecast. The model results were compared with observed 2m temperature and 10m wind speed at Mehrabad, Aghdasieh and Chitgar stations. The results show a fairly good agreement with the corresponding observations. The results also show that during the time of CPA, the Tehran heat island is well formed and has a relatively high intensity in the northern part of the city while during the time of CTA the heat island does not grow during the night time and has a lower intensity as well. Examining the model results also shows that the model is capable of capturing the micro and mesoscale characteristics features of the heat island and catabatic and anabatic winds over the city. Also, the results show that the mountain cooling intensities the urban heat island circulations at night and these circulations reduce heat island intensity.

The study presented here is an artificial neural network modeling for prediction of permeability in one of the limestone reservoirs in the south of Iran. The permeability ANN is a three layer network. We have used two separate ANNs to model permeability. Initially, density, gamma ray, neutron and sonic logs were applied for input. Then depth data related to these data were added to the input; the results of the two networks have then been compared. This comparison has shown that in the second state the accuracy of the model has been improved significantly.

In this investigation long-period records of earthquake data were used for separation of surface waves and calculation of phase and group velocity dispersion curves. Kazeroun main shock (3 May, 1995) recorded by Iranian Long Period Array (ILPA) at station 1, 2, and 7; Sant- Hellana main shock (27 April, 1986) recorded by two very long-period European stations (KON in Spain and TOL in Norway) have been processed. First, we have corrected all records by special techniques for the arm and tilting of instrumentation. Then, we processed the earthquake data for separation of surface waves (Love and Rayleigh waves) by two dimensional frequency-time algorithms. After separation of surface waves from main shocks full automotive intelligent programs have been used for detection and computation of phase and group velocity dispersion curves. Results show that phase velocities calculated from three-components of ILPA data at stations 1, 2, and 7 are less than the group velocities for the same stations (inverse dispersion). Vertical components of the three stations show that the phase velocity with respect to frequency is increased, but for the horizontal components (N-S and E-W directions) the phase velocities with respect to frequency are decreased. This behavior shows that the phase velocity of shear waves is reduced with depth of penetration. Besides, the higher modes of surface waves are mimicked as they pass through the Zagros belt from the earthquake epicenter to ILPA recording stations. Similar computations of phase and group velocities from the TOL and the KON data stations show that the phase velocities are greater than the group velocities (forward dispersion).

Near-field data obtained from the main shock of May 28, 2004 Baladeh-Kojour earthquake, with Mw = 6.2 is used to determine the frequency dependence of direct shear wave quality factor Qβ, in the range of 1 to 32 Hz. Based on the regression analysis, the following equation is proposed: Qβ= 90f0.74 High attenuation factor and frequency dependence of quality factor for shear waves imply the high seismicity of the region and agrees well with similar studies in other seismically active regions such as Greece, Japan, India, etc.