Gulf of Mexico Field Data

Next, I test the filtered GDM on a marine seismic data set recorded in the Gulf of Mexico. This data set is known as the Mississippi Canyon line and was released by Westerngeco as a benchmark test for the 1997 SEG workshop on multiple attenuation (Dragoset, 1999). Figure a shows the migration velocity model, where a shallow salt body with a high velocity (around $4.7~km/s$ ) is embedded in a background sediment model with a smoothly varying velocity distribution. The water bottom is at a depth of about $1.5~km$ .

Figure b displays a raw RTM image migrated with the velocity shown in Figure a. Due to the strong velocity contrast in the model, especially around the top of the salt, strong amplitude, low-frequency artifacts are present in the image, which are undesired and obscure the image of sediment layers above the salt.

Figure: The migration velocity and the standard RTM image of the Gulf of Mexico data set.

Figure: The GDM image in comparison with the filtered RTM image. The white dashed box highlights the differences of the two images.

A high-pass filter is applied to this RTM image to reduce these artifacts and the result is shown in Figure a. As we can see, this simple filtering is effective in reducing the artifacts, but large residuals still remain. To compute an artifact-free image, I apply the filtered GDM algorithm associated with equation , and only take the last term in equation  to get the clean image shown in Figure b. The artifacts are successfully eliminated in Figure b and as a result, the geology of the shallow sediments is much easier to interpret compared to that in Figure b.