Velocity Inversion And Imaging Analysis Constrained By Acoustic Channel

Studies have shown that the velocity of sound waves in seawater varies continuously with depth and forms acoustic wave guides in deep seawater layers.The change that sound wave velocity according to depth is called the acoustic channel.The sound wave guide in the deep sea layer has been known as acoustic channel.The existence of the acoustic channel causes the seismic waves emitted by the seismic source to form reflections in the water layer,which is called water layer reflection.In addition,the existence of the deep-sea acoustic channel makes the propagation path and propagation time of seismic waves deviate from the situation of the uniform velocity layer,which indeed affects the imaging position of the underwater strata.Therefore,it is of great significance to accurately determine the seawater acoustic channel for migration imaging.For the sake of studying the construction of deep-sea acoustic channel using water layer and seabed reflections as well as discussing the influence of acoustic channel on seabed imaging positions,the vertical variation of water layer velocity is described by Munk formula,and on this basis,the stack velocity of water layer reflections is figured.By utilizing the event axis compensation method based on human-computer interaction,the water layer reflection event axis is effectively solved,and the real water layer reflection stacking velocity is obtained while retaining the original effective data information.On the premise of applying the event compensation method based on human-computer interaction to analyse velocity for seawater seismic records,it is find that the compensation method not only consistent with the theoretical model retaining the original valid data information as well,but also gives a correct velocity while calculating real water reflection data,which provides a more accurate initial velocity model for tomography;To simulate the kinematic response characteristics of waves in real data,the random velocity perturbation modeling method is exploited.Fourier transform and inverse transform have been used to establish a random process in frequency-wavenumber domain,in this way,the inverse Fourier transform is utilized to get the expression of the random velocity in spatial areas.The 4-th-order Runge-Kutta method can numerically solve the ray tracing equation to give an expression on the multi-traveltime problem caused by ray caustics in random media;By using Kirchhoff migration imaging and Gaussian beam migration imaging,we get the deviation of the seabed imaging position when the acoustic channel stochastic medium model exist;Combined with the stack velocity initial model obtained by the man-machine interactive velocity analysis method under the constraint of acoustic channel,on the basis of fast scanning method to solve the function equation,the adjoint tomography is exploited to invert the acoustic channel with the seabed reflection travel time as the constraint information.Using the high-precision analysis velocity as the initial iterative model effectively cut down the iteration times,as well as increasing the computation quantities.As a result,the velocity change of the acoustic channel is inverted as well as the channel axis position.It furnishes a new method for Munk formula parameter inversion.In the aspect of seismic exploration,it's of great significance to take these views into account in marine seismology inversion.