Modeling the Near-Surface Using High-Resolution Seismic Data

Abstract

In the Arabian Peninsula, the near-surface represents a major challenge in seismic exploration. For accurate deep subsurface reservoir imaging, an accurate near-surface velocity model is required. In this dissertation, I review the regional geology and geomorphology and the existing methods used for velocity modeling. A new method is developed, which depends on acquiring high-resolution shallow seismic data. The data is processed to obtain a near-surface velocity. The numerical modeling shows that the accuracy required for accurate imaging can be obtained through this new method. Alternative existing methods either lack the required accuracy or are very expensive to use. Three real field data cases are presented. In each case a high-resolution velocity model is obtained and used to process the conventional data. To make the new method more practical, I recommend using the CMP mode and a land streamer for data acquisition. I show that some artifacts in the data acquired using a land streamer can be overcome by using optimally-designed plate-mounded geophones. The two main factors in the design are the plate weight and area. From the experimental data, I conclude that land-streamer data over sand dune can be more coherent than conventional data, especially the first-arrival events.