How would you define an absorbing region (at the walls of a finite numerical simulation) in physics and how would you implement it? Can it be simulated by a high potential, low negative potential at the boundaries, how?
Or an imaginary (damping) potential? It has to do with this sentence if found in arXiv:083.0507 (last paragraph of page 5):
...but an absorbing region far from the horizon is required to prevent the onset of spurious dynamical instabilities due to excitations circulating around the integration box.