The Specialized Reading Competition 2020

The Specialized Reading Competition is a mentorship program that aims to bridge the experience gap and promote knowledge sharing. Young Professionals and students had the opportunity to work closely with mentors to guide them in a thorough reading program to publish papers in four themes: Machine Learning, Modeling, Seismic Imaging, and Climate Change

Dhahran Geoscience Society would like to thank the 40+ registered passionate readers and the 16 selected participants for their perseverance and enthusiasm to learn. However, we encourage all participants to continue reading and adopt reading as a habit. DGS would like also to extend special thanks to our mentors for sharing their valuable time and experience with the mentees.

Abdullah currently is a PhD student in Earth Science and Engineering at King Abdullah University of Science and Technology (KAUST), KSA. He received his B.S in Geophysics at King Fahd University of Petroleum and Minerals (KFUPM), KSA, and his M.S in Earth Science and Engineering at KAUST. In 2014, he enrolled in an international exchange program to study Geophysical Engineering at Colorado School of Mines, USA for one semester. In 2018, he was part of KAUST team in the SEG/DGS challenge bowl competition and won first place in the Middle East and second place in the final round held in the SEG annual meeting in Anaheim, California.

Abstract

Due to recent advances in computational resources and big data technology, machine learning (ML) algorithms have been widely embraced in the geoscience community. Several applications, including seismic processing and interpretation, have utilized several ML techniques. This paper focuses on one of the most crucial processing steps in seismic data processing, namely velocity modelling. Often, the first velocity model obtained from seismic data is the normal moveout (NMO) velocity. It is used to flatten the hyperbolic events in the data. Despite that NMO velocity is not accurate in complex media, it is often used as an initial guess for further velocity modelling methods such as traveltime tomography and full-waveform inversion (FWI). All the velocity modelling methods have certain limitations, which include extensive human intervention in form of picking semblances in NMO velocity analysis or converging to a local minimum as in FWI. In this paper, we critically review the studies that utilized the ML techniques to overcome the velocity modelling limitations. We also discuss ML applications that utilize an inversion-like velocity such as those obtained from tomography and FWI. The reviewed applications cover different types of learning: supervised, unsupervised, and semi-supervised. In view of contributing to the digital transformation agenda of the petroleum industry, this paper concludes with a set of recommendations to overcome the prevailing challenges and for the implementation of more advanced ML technologies. We hope that the recommendations will help to achieve complete automation of the NMO velocity and further enhance the performance of ML applications as used in the FWI framework.

Full paper

Ruba Afifi had joined Saudi Aramco College Degree Program for Non-Employees (CDPNE) in 2014. She graduated from the University of Arizona in 2019 with a BS degree majoring in Geophysics and minoring in Mathematics. During her time in college, she engaged in geochronological research to extract zircons from clastic and volcanic rocks and perform U-Pb dating to find their ages. She also contributed in a study with Dr. Jay Quade to construct a paleogeographical timeline for the evolution of river systems in the Afar Rift in Ethiopia, which opens the opportunity to trace the origins of the first hominin species through time and space. Since graduating, Ruba has been in training assignments as an Exploration Geoscientist in Exploration Resource Assessment Department. During her assignment, she applied the Play-based Exploration workflow in South Jaladi area. Ruba had excelled in her PDP training programs, UPOP and STEP. She had been selected as the female MC for the 25th cohort of STEP’s graduation ceremony. Ruba proudly identifies herself by her organization skills, fast learning abilities and her scientific mind.

Abstract

By the end of this century, the climate is likely to warm by 2 oC since pre-industrial times; model forecasts supported by observational and physical evidence of global warming suggest so (Stocker et al., 2013). On a regional scale, climate change is more evident in lower latitudes than in middle latitudes; warm dry regions are becoming warmer and dryer, and wet regions are becoming wetter. Ice sheets and glaciers are melting, causing together with the thermal expansion of water a notable rise in sea level. In desert regions such as the Middle East, livelihoods are affected, as seasonal mean temperatures are already near the tolerance level for humans and other species. As a result, productivity is decreasing and the pressure for migration is increasing (Hansen & Sato, 2016). The principle cause for climate change is the increase in greenhouse gas concentrations, fundamentally CO2, in the atmosphere. Reduction of CO2 emissions is proven to be a difficult and slow process.

There are current negotiations around developing and deploying technologies that aim to decrease global temperatures, and thus diminish the impact of climate change; these technologies include Carbon Dioxide Removal (CDR), and Solar Radiation Management (SRM). The former directly captures CO2 from the atmosphere, while the latter allows the Earth to absorb less solar radiation. This paper recites facts about each strategy’s mode of function, timescale over which it is effective, and the associated risks and uncertainties. Policymakers should invest in research and development for these technologies as soon as possible. SRM seems to be a less costly and fast-acting solution to counter the effects of climate change in the short-term and, if implemented, would buy time for more fundamental adjustments. On the other hand, CDR technologies are likely more effective to diminish future changes in the climate in the long term, with less environmental risks.

Full paper

Prof. Georgiy Stenchikov is a chair of the Earth Sciences and Engineering Program at King Abdullah University of Science and Technology since 2009. He began his research in the Russian Academy of Sciences, where he focused on numerical methods and radiation gas dynamics. He led a research group in the Computer Center of the Russian Academy of Sciences that studied anthropogenic impacts on the Earth’s climate and environmental systems.  Since 1992, Prof. Stenchikov worked in the US conducting interdisciplinary studies in the broad field of climate modeling, atmospheric physics, and environmental sciences. For his work on climate impact modeling, Prof. Stenchikov was awarded a Prize from the Council of Ministers of the Soviet Union, a Gold Medal from the Russian National Exhibition, outstanding scientific paper awards from the American Geophysical Union Journal and the National Atmospheric and Oceanic Administration. He also co-authored the Nobel Prize winning report from the Intergovernmental Panel on Climate Change AR4.

Fatai Anifowose is a researcher in the Geology Technology division of the EXPEC Advanced Research Center of Saudi Aramco. His career interweaves the artificial intelligence aspect of Computer Science with reservoir characterization aspect of Petroleum Geology. He is a member of the AI and Big Data technology team of EXPEC Advanced Research Center’s 4IR Committee. He has published over 45 journal papers and conference proceedings. In addition to DGS, he is also a member of SPE and EAGE.

Khaled is an experienced geophysicist with a demonstrated proficiency of working in Saudi Aramco for more than 15 years. He has a B.Sc. from the University of Tulsa, M.Sc. and Ph.D from the University of Calgary. He is skilled in seismic anisotropy, borehole seismic, seismic imaging, and velocity model building. Currently, he works in the Geophysical Imaging Department as the 3D Near Surface team leader. He was a member in DGS executive committee 3 times; last in 2019 as the technical Vice President.

Rainer Zuhlke research interests cover a wide range of reservoir characterization and modeling topics such as reservoir prediction, quantitative sequence stratigraphy, and forward depositional & diagenetic modeling. His work experience covers the Arabian basins, North Africa, Brazil, South and Central Europe. Rainer has his PhD from University of Heidelberg and his MSc from the University of Freiburg. He is currently a Research Geologist with Saudi Aramco in the Exploration and Petroleum Engineering Advanced Research Center.