报 告 人: Dr. Sanzong Zhang

单    位: OPERA, Pau, France

报告题目：High-frequency full waveform inversion and imaging

报告时间: 2021年7月9日（周五）下午 15：00-16：30

地    点: 腾讯会议(会议ID：707 364 797）

报告人简介：

Dr. Zhang is now a research geophysicist at OPERA, Pau, France, which is a research center on seismic image and inversion affiliated with the French oil company, TOTAL. His research mainly focuses on full waveform inversion, least-square reverse time migration, and their application on the real data. Before joining OPERA, Dr. Zhang is a research scientist at IBM Watson research center located in Cambridge and New York, USA. His research work at IBM includes high performance computing and optimization, GPU computing, deep learning and their application in seismic simulation, tomography, and inversion. He was leading several projects supported by the department of energy and oil companies (bp, ExxonMobil, chevron, and Saudi Aramco). Dr Zhang got PhD in geophysics in 2014 from KAUST under the supervision of Professor Jerry Schuster. He also holds the bachelor’s degree and the master’s degree in applied geophysics from China University of Geosciences (Wuhan) in 2002 and 2005.

报告摘要：

The seismic data, including transmitted waves, primary reflections, and multiples, has been widely applied to estimate velocity model and migration image. However, reverse time migration uses only the primary reflections to image the geological structure in the subsurface, and it also introduces multiples as crosstalk noises into migration images. Full waveform inversion makes full use of all information in the seismic data to  invert  for  a  high resolution  of  velocity  model.  Therefore,  full  waveform  inversion  provides  a  more balanced illumination to image the subsurface than reverse time migration, especially for the complex geobodies, such as salt and subsalt. This talk would introduce a new technology to invert for the subsurface imagedirectly from full waveform inversion, which avoids the conventional time-consuming preprocessing, velocity model building, and migration. Compared with reverse time migration, the image directly from FWI often provides  additional  structural  information with  better  illumination  and  higher signal-to-noise  ratio. Some  real examples demonstrate its advantage and benefit in the industrial application.