Lateral velocity heterogeneties modelling in seismic tomography by introducing different initial velocity models

Velocity model building is a crucial step for construction of seismic image of the subsurface in depth imaging. A wide variety of different velocity model building methods are available. Reflection tomography is one of those methods. One of the drawbacks of tomography method is that it requires picking reflection events in the prestack data. Picking procedure is extremely time consuming and can become difficult if the signal to noise ratio in the data is low. In this study, a new version of tomography called Normal Incidence Point (NIP) wave tomography is used for construction of velocity model. This technique makes use of traveltime information in the form of kinematic wavefield attributes. These attributes are coefficients of the second order traveltime approximations in the midpoint and offset coordinates and can be extracted from prestack seismic data by means of common reflection surface stack method. The required input data for NIP tomography inversion are taken from stacked results at number of pick locations, while these locations do not need to follow a continuous horizon in the section. The problem of building the velocity model by tomography method is solved in an iterative manner here. During iterations, difference of observed and modeled data is minimized and the model is updated. This procedure would continue until the misfit falls below a specified value. Modeling observed data for the first time requires an initial velocity model. Initial velocity model in normal incidence point tomography contains a constant near surface velocity which increases linearly with depth. In the present study, four different functions, introduced by different researches, used besides linear function to produce initial velocity model. In addition to these functions, the stacking velocity derived from kinematic wavefield attributes was used in NIP tomography, as initial velocity model. Accuracy and consistency of these velocity models were evaluated by application to a 1D and a 2D synthetic data. Result of these data showed that different initial velocity models due to different functions used in NIP tomography, have different effects on the final velocity model. In 1D data example, the result showed that the NIP tomography method with new velocity function introduced to the tomographic algorithm will gives accurate velocity model after little iteration with acceptable error and high consistency with the data. In case of 2D synthetic data example, five different velocity models obtained by normal incidence point tomography with four velocity function besides the stacking velocity as initial velocity model. Different final velocity models obtained here show different ability of functions in handling lateral heterogeneities. However, the velocity functions introduced in other studies showed that besides the importance of initial velocity model in normal incidence point tomography, they could not serve as a suitable initial velocity model. Although these models were consistent with the data, they were not able to separate close velocity anomalies. However, the velocity model obtained by stacking velocity as initial model in normal incidence point tomography shows higher accuracy and consistency with the data and could handle lateral velocity changes in tomographic procedure, too. These techniques were applied to a real dataset. This dataset contains geometric complexity and lithological complexes. Therefore it could clarify the ability of these velocity equations in producing acceptable migrated section. All of the equation used to make velocity model for this dataset are then used for post stack migration. By comparing the migrated section obtained, the linear function and stacking velocity function showed that they could perform better in the presence of lateral velocity heterogeneities. Later on, these two models were used to produce acceptable prestack depth migration sections. The stacking velocity used for the initial model for NIP tomography gave better result in presetack depth migration. This result was compared with the result of conventional prestack depth migration. The velocity model for the latter case was obtained by the migration velocity analysis technique. Results of both methods were comparable. However, the NIP tomography model was so simple and smooth and also obtained in so much less time compared to the complex migration velocity analysis model.