Workshop 8

Elastic FWI for Model Building Monday, June 2nd |

Convenors

  • Bertrand Duquet (TotalEnergies)
  • Gilles Lambaré (Viridien)
  • René-Edouard Plessix (Shell)

Description

With the constant increase of HPC capabilities, elastic FWI has rapidly progressed over the last decade and is today used at an industrial scale. Elastic FWI, however, covers different levels of complexity: from modeling and inverting the P waves elastically with a mono-parameter inversion to modeling and inverting in a multi-parameter inversion scheme waveforms considered as coherent noise in classical acoustic processing flows (mode conversions, surface waves, …). The complexity level, and then computing cost, depend on the objective to achieve and the type of seismic acquisitions considered (land or marine acquisitions). During this workshop, we propose to focus on assessing the added value of elastic versus acoustic FWI in terms of velocity model building and imaging for on-shore and off-shore data.

Participant Profile

Researchers, practitioners, students, and decision makers interested in better understanding the value and challenges of elastic FWI.

Workshop Programme

Coming Soon!

TimeActivity
08:15Workshop Introduction
Topic 1: AI compatible Seismic Data Format; Augmentation...
08:25Leveraging OSDU cloud technologies to improve performances of seismic data – I. Frank Haeseler (Total)
08:45Human choices to define data structures to aid development of better AI applications – A. John Solum (Shell)
09:05GenAI for advanced seismic generation, data enrichment and instant seismic search – G. Anna Dubovik (Data Analysis Center)
09:25The Importance of Data Governance to Scale Up the Value in AI Solutions – Oddgeir Gramstad (AkerBP)
09:45 Round table discussion
10:00Coffee break
Topic 2: AI for Seismic Interpretation and Reservoir Properties Prediction
10:10Seismic Data Conditioning and it’s implications for AI seismic Interpretation – Ryan Williams (Geoteric)
10:30Generating horizon volume from seismic poststack images using artificial intelligence– Aria Abubakar (SLB)
10:50The role of AI in structural seismic interpretations – potential and pitfalls – Nicole Grobys (WintershallDEA)
11:10Some Challenges and Solutions in AI for Seismic Interpretation – S. Xinming Wu (University of Science and Technology of China)
11:30 Round table discussion
11:45Lunch break
12:35TerraSight and StrataFlux: new attributes for ML-assisted seismic interpretation. – Sergey Fomel (The University of Texas at Austin)
12:55But what about subtle faults and fractures? – Victor Aarre Madsen(Firda Geo)
13:15Assessing the Sufficiency of Training Datasets for Deep Neural Networks: Techniques and Insights – Tanya Colwell (Geosoftware)
13:35Property Prediction from Seismic with Neural Networks - Learnings from an 8-year Journey– Eirik Larsen ( Earth Science Analytics)
13:55Machine Learning-based seismic inversion methods - Arnaud Huck (dGB Earth Sciences)
14:15Round table discussion
14:30Coffee Break
Topic 5: AI in the New Energy Transition
14:40Leveraging Interactive Deep Learning for Rapid Detection of Quaternary Lithotypes in Ultra-High Resolution Seismic Data: A Case Study from Offshore Netherlands Scotty Salamoff (Bluware )
15:00AI Solutions to Cost-Effective CO2 Monitoring – G. Wenyi Hu (Slb)
15:20Workshop summary - feedbacks / next steps