Dynamic Plane-wave LSRTM

To suppress the noise in Figure (b), the dynamic encoding approach (Krebs et al., 2009; Schuster et al., 2011) is employed. The plane-wave LSRTM is still applied to just one sugergather at each iteration, but the encoding function (in this case, the ray parameter) is changed at every iteration. This time, 31 iterations are used to fully cover the angle range, and the result is shown in Figure (c). In this example, the ray parameter $p$ ranges from -333 $\mu$ s/m to 333 $\mu$ s/m sequentially with a unique $p$ for each iteration. With dynamic encoding, the image has a much higher signal-to-noise ratio (S/N) compared to Figure (b), and in the deep part, most of the structures are well imaged. The LSRTM with dynamic encoding is very efficient and has the potential to produce high quality images, but there are no image gathers available for quality control or migration velocity analysis.