Title: Gradient inversion of marine seismic reflection data: Parameterization and geometrical spreading
Abstract: The inverse problem of finding the elastic properties of the Earth’s crust given a set of seismic reflection data can be considered as formed by two distinct problems with very different characteristics. First, we have to find the longwavelength velocity variations of the seismic waves, that best match the arrival times of at least the most prominent signals in the data. This tomographic part of the problem is highly nonlinear. To properly solve it, global search methods should be applied to find a. pool of likely models. Due to the large dimensions of the model space this is too costly for problems of realistic size. Common practice is standard velocity analysis or visual inspection. Once one or several good velocity models have been found, the problem of finding the location and strength of the reflectors in the medium can be attacked. This problem is quasi-linear and well posed in the framework of gradient algorithms, where the function to be minimized describes the misfit between measured data and calculated synthetic seismograms. The success for this second part strongly depends on the parametrization of the model space, and the correct modeling of the seismic source used in the experiment as well as the three-dimensional effects in the data (geometrical spreading). Otherwise there is no hope that the inversion results are reliable in particular when not only properties of P-waves are sought but also shear-waves or attenuation are considered. Here we address the problems of parametrization and geometrical spreading.
Publication Year: 1993
Publication Date: 1993-01-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
Cited By Count: 6
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot