Seismic velocity is a very useful tool for pore pressure prediction prior to drilling a well. This chapter identifies various sources of velocity data - checkshot, VSP, well logs, laboratory, and seismic measurements. The goal is to obtain velocity variations in 3D that not only reflect the subsurface structures in depth but also convey the expected range of velocity variations that is compliant with rock physics principles, structural geology, and stratigraphy of formations and geopressure. This chapter discusses various ways to obtain velocity data for pore pressure analysis and points out how to establish a link between seismic traces that are recorded in "space and time" and the "space and depth" that are required by the drilling community. It stresses that the seismic model building step must deal with separating imaging velocity from the velocity that is close to rock velocity.

This chapter conjectures on some of the upcoming technologies and how they might impact pore pressure prediction in the future. It gives a schematic of a workflow that integrates basin modeling derived pore pressure to rock physics and seismic anisotropic velocity modeling and imaging, executed in a grand loop. Some recent applications of machine learning applied to pore pressure prediction are discussed. The chapter concludes that the future is almost here, and it is bright indeed!

This chapter discusses technologies that yield geologically plausible and physically possible interval velocities from surface seismic data. The workflow is termed RPGVM, rock physics guided velocity modeling. The approach can be used on any algorithm for interval velocity computations – be it conventional or based on inversions such as tomography and FWI. The goal is to define the parameter base associated with a particular inversion approach so that the inferred velocity model is constrained by rock physics and bounds of pore pressure. Applications from the Gulf of Mexico and offshore India are described. The chapter shows the value of rock physics templates for deriving velocity models with anisotropic tomography.