Том: 8th Russian Supercomputing Days, RuSCDays 2022. Lecture Notes in Computer Science (Moscow, Russia, September 26-27)
Том: 13708
, Год издания: 2022
Многотомное издание: 8th Russian Supercomputing Days, RuSCDays 2022. Lecture Notes in Computer Science (Moscow, Russia, September 26-27)
Издатель: Springer International Publishing
, Место издания: Berlin
Страницы: 237-249
Аннотация
We present a parallel algorithm for 3D seismic tomography that reconstructs the depth velocity model. The aim is to provide the parallelization of the developed method for processing 3D seismic data in the production mode. Time migration results are the input data for tomography. The tomography algorithm contains three blocks: data preparation for tomography from results of time processing, tomography matrix construction via ray tracing and tomography matrix inversion. Each block uses MPI technology. The paper provides numerical experiments performed using real data got from the field in Eastern Siberia.
