Modeling

1) Modeling

Details



As it is at the heart of imaging and Full waveform inversion our objective in this matter are:
  a. Get the physics right: Enhance our ability to model in complex anisotropic media taking attenuation into account. This will include a focus on Orthorhombic elastic and acoustic (Alkhalifah, 2013; Song and Alkhalifah, 2013; Sindi and Alkhalifah, 2014).
 
  b. Get it to be stable and clean: We will look into breaking the barrier of instability whether it is with regard to the CFL condition, or the boundary representation, or even the acoustic anisotropic approximation. This has included our utilization and devotion to spectral methods, which lends itself to natural integration of anisotropy (Wu and Alkhalifah, 2013; Alkhalifah, 2014).
 
  c. Get it to be efficient: This is a big one. However, by breaking the CFL and Newmann conditions of stability we are capable of utilizing bigger time steps and with spectral methods cleaner coarsely sampled wavefields. We utilize residual wavefield extrapolations over velocity. We, also, use GPU implementations for better computational speed per cost (Alkhalifah, 2013; Wu and Alkhalifah, 2013; Wu and Alkhalifah, 2014).
 
  d. More exotic modeling: We have heavily and still are invested in direct prestack modeling using the double square equation in what we refer to as prestack exploding reflector modeling and migration (PERM). This modeling starts from reflectivity to offer cheap prestack wavefields that include the extended domain with potential applications in MVA and migration (Alkhalifah and Fomel 2010 and 2011; Alkhalifah, Fomel, and Wu, 2014; Wu and Alkhalifah, 2013; Alkhalifah, 2013; Wang and Alkhalifah 2014). We already have anisotropic implementations of it, which lends itself to the approach naturally.