Spring Symposium 2023
Join us for the 2023 Spring Symposium hosted by the Geophysical Society of Houston. We are very excited to honor Matt Hall this year in a new setting and in a new format.
The event will take place at the Houston Museum of Natural Sciences on April 19th and 20th. The technical program will occur on the 19th with a reception following in the Cullen Hall of Gems and Minerals. The second day of the Spring Symposium will be dedicated to a Hack-A-Thon led by our honoree, Matt Hall. The theme of this year’s technical session is Python & the Geosciences.
2023
Spring
Symposium
April 19 & 20, 2023
Python & the
Geosciences
@ the Houston Museum of Natural Science
Ticket prices
& Printable flyers
Matt Hall
Honoree
Scientist / programmer
Bergen, Vestland, Norway
Open source software, open data, and open access science make the world go round.
I'm a scientific Python programmer at Equinor, working on various open source tools for scientists and engineers. Check them out or contribute at https://github.com/equinor
I'm also involved in the Software Underground, a non-profit promoting productive conversation and collaboration among digital subsurface scientists. Join in at https://softwareunderground.org/join
Ticket prices
Ticket Type
Before 12th April
After 12th April
Member before April 12th | Non-Member before April 12th | Member after April 12th | Non-Member after April 12th | |
---|---|---|---|---|
2 Days (entire event) | $350 | $410* | $400 | $460* |
1st Day | $250 | $310* | $300 | $360* |
2nd Day | $150 | $210* | $200 | $260* |
Students (whole event) | $100 | $100 | $100 | $100 |
Evening Event | $50 | $60 | $60 | $70 |
* Ticket includes 1 year of GSH membership
Printable Symposium Flyers
Reception
Please join us for an evening of cocktails, appetizers and networking at the Houston Museum of Natural Science in the Cullen Hall of Gems and Minerals on April 19th from 6-8PM. There will be a cash bar along with an enticing lecture to entertain our scientifically-minded guests; parking is included with registration. Bring a spouse, partner or friends
...Everyone is welcome
April 19 - Day 1
Wednesday 6:00 - 8:00 PM
Artemis Lunar Exploration
presented by
David Kring, USRA
David King, USRA
Reception Speaker
-
Topic: Artemis Lunar Exploration
​​
-
Speaker: David Kring, USRA
​
-
Where: Cullen Hall of Gems and Minerals 2nd floor map.
-
When: April 19th, 6-8 PM
-
Food and Drink: Beer, Wine, and Liquor will be served at a cash-bar. Passed-appetizers will be served.
-
Cost of the Reception Event: Included with first day and entire event Symposium Attendees.
​​
-
Reception only tickets:
Member price:
$50 - Before 12th April
$60 - After 12th April
Non-Member price:
$60 - Before 12th April
$70 - After 12th April
Sponsored by
Hack-a-Thon
What is a
hack-a-thon?
-
A hack-a-thon is a collaboration event to design and build new tools for a specific problem via coding. It is usually framed as a competition.
-
Organizations such as the SEG and NASA have hosted hack-a-thons in the past.
-
Examples are linked below
Hack-a-Thon
Details
​​
-
For the hack-a-thon hosted during the GSH Spring Symposium participants
will be divided into different teams of
6 people each and presented with a geophysical question or problem.
-
Each team will then work together to determine the best method to solve the question and program the tool via Python.
-
The winning team will be awarded a prize; however, the real prize is the experience of learning with your teammates and the insight you will bring forward into your normal daily work.
Location
5555 Hermann Park Drive
Houston Texas 77030
The conference will be on the lower level in the W.T. and Louise J. Moran Lecture Hall.
If entering the museum from the Garage or the Caroline Street entrance,
-
Go to the second general exhibit entrance from the Grand Hall, towards the dinosaurs.
-
Go straight to the elevators and down to the lower level.
-
The Symposium is directly off the elevators.
There will NOT be GSH signage; however, all HMNS staff will be aware of our event and will be able to direct you to the Symposium location once in the museum.
Symposium attendees will have full access to the general exhibit to experience during breaks or prior to the evening reception after the first day.
Houston Museum of Natural Science
Getting There
​
​
-
Parking will be validated for Symposium Attendees in the HMNS garage. Entrance is located on Caroline St, and the garage is attached directly to the museum. Please make sure you park in the correct garage.
​​
-
HMNS is also conveniently located near the MetroRail stop, Fannin South (0.3 mile walk).
METRORail | Red Line | Green Line | Purple Line | Houston, Texas (ridemetro.org)
Program Schedule
April 19 - Day 1 Wednesday
April 20 - Day 2 Thursday
8:00 am - 5:00 pm
Morning & Afternoon Sessions
​
6:00 pm - 8:00 pm
Evening Networking Event,
Cullen Hall of Gem & Mineral
8:00 am - 5:00 pm
Hackathon & SEG Challenge Bowl
Speakers
Abstract & Bios
The first day of the symposium will include contributions from 8 technical speakers, time to speak with the Symposium vendors, and an evening reception. Information on the speakers and their topics can be found below.
Speakers: Bios & Abstracts
Day 1 Speakers
Brendon Hall
8:45-9:30 AM
Python and Geoscience: How Open Tools Have Changed the Way We Work
Bio: Brendon has more than 15 years of industry experience using scientific computing to solve problems in oil and gas exploration. He began his career at ExxonMobil Upstream Research and ION Geophysical. Recently he has been focused on building software tools that allow geoscientists and engineers to augment their work with Artificial Intelligence. Brendon holds a Ph.D. in mechanical engineering from the University of California Santa Barbara, and a B.Eng. in mechanical engineering and a B.Sc. in computer science from Western University in Ontario, Canada.
Abstract: The scientific Python ecosystem has made a profound impact on both academic and industrial geoscience workflows. This talk will explore various applications of Python in geoscience, highlighting the compounding benefit of open and reproducible projects. We will briefly discuss where this technology came from, the current landscape, and future possibilities. Given Python’s popularity for data intensive applications, machine learning and AI will continue to have a growing impact on geoscience. As computational tools become more important the importance of reproducibility will be emphasized and the challenge for all practitioners clearly stated.
Rafael Pinto
9:30 - 10:15 AM
Re-ranking Seismic Uncertainty Analysis Surfaces with Python
Bio: Rafael holds a M.S. in geophysics applied to oil and gas development from the Colorado School of Mines and a B.S. in geophysical engineering from the Universidad Central de Venezuela. He is passionate about developing innovative applications that assist subject matter experts in unlocking value, such as web apps that make it easier to find insights in large data sets and adapting solutions from interdisciplinary fields. Before joining Enthought, Rafael worked for a decade as a geophysicist at Noble Energy, where he characterized oil reservoirs and matured prospects through seismic interpretation and inversion. During this time, he also learned to use Python for automation and reproducibility, which bridged his transition into data analytics applied to resource plays. He then worked as a data analyst at EPAM Systems, where he transformed the vision of their energy customers into actionable software requirements. In his spare time, Rafael enjoys honing his coding skills, discovering new swimming holes, and exploring the local food scene.
Abstract: Geoscientists often face digital challenges that cannot be solved with commercially available software built for executing typical exploration and development workflows. This can be due to issues with the data and parsing algorithms within the software, or because the practitioners have found a creative way to use their data but lack the coding skills to bring these ideas to fruition. Python, with its beginner-friendly learning curve and wide range of open-source packages for multidimensional array manipulation and visualization, has become an essential tool in applied sciences. This study showcases how adding foundational software development skills to the modern geophysicist toolbox can accelerate innovation. Specifically, I present a method for ranking equiprobable surfaces to assess prognosed depth uncertainty. This work is inspired by a seismic uncertainty analysis performed on proprietary data to understand the distribution of resources in an exploration prospect. To demonstrate the method, I simulate the results of the analysis using the Top of Rotliegend Group from the Groningen gas field open dataset and rank the resulting surfaces based on the target's depth.
Cody MacDonald
10:30 - 11:15 AM
Open-Source Python tools for Deep Learning Seismic Interpretation
and Challenges
Bio: Cody MacDonald holds a B.Sc. and M.Sc. in geology from Dalhousie University, Halifax, Nova Scotia, Canada. For just over a decade, he was employed at ExxonMobil, where he had roles in exploration, technology, applied geophysics, and research & development. Previous experience includes leading machine learning and deep learning projects and programs at ExxonMobil. Following interests in AI and deep learning at scale, Cody left ExxonMobil and joined SambaNova Systems in 2022, where he currently works as a customer engineer, with a heavy focus on computer vision.
Abstract: Deep learning for geophysical analysis, such as seismic interpretation has been an active field of R&D for several years (Waldeland, et al., 2018). Despite encouraging beginnings and early software development, uptake has been relatively slow. Some of this lag in adoption is due to the complex, scientific, and process driven nature of petroleum exploration. However, other factors that are frequently attributed to a lack of progress, include: a lack of available labels for supervised learning, limited easy-to-use software to create such labels, a readily available source of datasets (even within large organizations), computationally sufficient resources for AI at the scale of geoscientific datasets, and the requisite combination of geoscience, data science, machine learning engineering expertise required to productionize these methodologies. Although self-supervised approaches, such as contrastive learning or masked image modeling, may provide a means to pre-train models and help alleviate label requirements, labels of a wide variety of features will be needed to fine tune models for specific interpretation tasks. Leading industry software from SLB, Bluware, GeoTeric, Enthought, and others are beginning to provide tools for both labeling and active learning, but these packages are expensive for enthusiast use. Luckily, with Python, there are many free basic options available, from packaged interfaces such as LabelMe to customizable dashboard frameworks, such as Plotly Dash and Streamlit. This presentation aims to give an overview of tools and Python utilities that make it easier to get started with deep learning for seismic interpretation. Covered materials will include: dealing with typical seismic data formats, creating labels with free tools, and how to digest these data and labels into a machine learning framework. Computational challenges in scaling to high quality results will be touched on, but not the focus of this talk. References: Anders U. Waldeland, Are Charles Jensen, Leiv-J. Gelius, and Anne H. Schistad Solberg, (2018), "Convolutional neural networks for automated seismic interpretation," The Leading Edge 37: 529–537.
Niven Shumaker
11:15 AM - 12:00 PM
Domain Data Science Insights Through Animations Using Two Case Studies
Bio: Niven Shumaker is currently Data Scientist and Innovation Advisor on the Global Innovation Factori Team at SLB. He has previously consulted for an AI-based exploration start-up and Sandia National Labs. Niven spent 15 years at Noble Energy in various positions including Tamar Asset Manager, Sr. Finance Advisor, Geophysical Advisor and was a key contributor to high impact deep water discoveries in the US Gulf of Mexico, West Africa, and the Eastern Mediterranean.
Abstract: The democratization of data science with Python open source libraries has revolutionized the field of data analysis and machine learning. Online communities like Stack Overflow and Medium have lowered the expertise barrier for citizen data scientists to learn and implement key components of data manipulation, machine learning, and statistical analysis. While Python open source is a universal enabler, it is not a substitute for domain expertise. Domain expertise enables the generation of novel insights by providing a deep understanding of the subject matter and the ability to recognize patterns and trends that may be overlooked by those without such expertise. When data scientists possess a thorough understanding of the domain they are analyzing, they can contextualize the data and identify relevant variables, engineer useful features, and build the relevant pipeline for the challenge at hand. This presentation covers two challenging domain-related case studies: production forecasting of a mature tight gas field and induced seismicity from saltwater disposal in the Midland Basin. Both case studies benefit from the use of animations generated from Python open-source libraries to highlight changes over time, which is particularly useful for dynamic feature engineering and predictive model conceptualization. Moreover, animated dashboards facilitate model explainability and convey complex dynamic processes in a compact form.
Chiran Ranganathan
13:30 PM - 14:15 PM
Permeability Predictions in Tight Sands from Core Data Leveraging Python and Machine Learning
Bio: Chiranjith Ranganathan works as a Software Architect/ Project Lead with Geosoftware with over 15 years of experience in the oil and gas industry. He has a Masters Degree in Computer Science from the University of Houston Main Campus. In the past few years he has worked closely with clients and Consulting services in solving Petrophysical problems using Machine Learning. His areas of interests are in Machine Learning, Compiler Design and Distributed Computing.
Abstract: Permeability is essential in reservoir modeling and frac design. Predicting permeability in low porosity formations from core data measurements often has challenges using empirical methods. In this presentation we leverage the extensive visualization and machine learning capabilities in python to build a model that trains with a feature vector augmented from basic curve measurements and learns to predict core permeability measurements across the entire well and in wells that don’t have core data measurements in the Montney formation. The steps involve cleaning up the data, feature selection, distributed training / hyperparameter tuning leveraging cloud capabilities to evaluate several hundreds of models and selecting the one that best fits the dataset. The end results are compared to recorded measurements in blind wells that weren’t used for training, the correlation of modeled permeability to bulk density and modeled permeability to recorded core permeability values are presented.
Venkatesh Anantharamu (Venki)
14:15 PM - 15:00 PM
Python to Prototype Machine Learning and Geophysical Applications for Wider Geoscience Communities
Bio: Venkatesh Anantharamu (Venki) has an undergraduate degree in Mining Engineering, from NITK India and a Masters in Geophysics from the University of Houston. He joined Ikon Science in 2014. He has worked on several conventional and unconventional projects as a Quantitative Interpretation (QI) specialist focusing on rock physics and seismic inversion. Currently, he is working as a Product Development Geoscientist to implement new Machine Learning technologies, modernized rock physics modeling tools, and 3D/4D reservoir characterization workflows into their software.
Abstract: Python prototyping allows us to rapidly compare state-of-the-art technologies to find the best model and application of models for the geoscientific challenge at hand. For example, how does a geo-minded non-data scientist begin to optimize a Random Forest training dataset with numerous inexplicable parameters for the task of log prediction? We use python prototyping as an effective method to design simple software that incorporates data science expertise to hard-code constants and develop relevant validation and QC plots to rapidly deliver solutions to geoscience teams. Feedback can be used to refine complex functions with the goal to simplify workflows and appropriately reduce user inputs. Ultimately, a decision is made to either deploy simple python interfaces or refine and commercialize technologies with the lessons learned. We present examples of the development process for a stochastic AVO modeling tool, automated rock physics and ML log and volume prediction tools.
Tim Brown
15:15 PM - 16:00 PM
Machine Learning Fault Segmentation on 3D seismic: Incorporating Synthetic Images and Sparse Labels
Bio: Tim Brown is a geophysicist at OXY, currently supporting development activities in the Delaware Basin. Tim received a Master's in Geophysics from the University of Houston in 2012. Before joining OXY in 2018, he began his professional career in 2012 at Southwestern Energy as a QI geophysicist and seismic interpreter. Tim specializes in subsurface characterization of unconventional reservoirs and has experience in several U.S. resource plays.
Abstract: Machine Learning Fault Segmentation (MLFS) on 3D seismic data has become a valuable tool across the E&P life-cycle, informing everything from exploration risk analyses to reservoir modeling. We describe a robust MLFS workflow that includes training a 3D Convolutional Neural Network (CNN) on synthetic images and fine-tuning using sparse labels from actual data to address domain shift. Our synthetic image generation and augmentation process is highly flexible, accounting for different faulting styles, noise patterns, and wavelet spectra. Furthermore, we use multiple network architectures with varying input sizes and field-of-view scaling to balance fault detail and continuity. Finally, we use a fine-tuning strategy that adapts the base feature recognition to faulting geometry defined by sparse user labels. The inference code is scalable to multiple GPUs and can handle input volume sizes upward of 500 GB. Multiple case studies are discussed.
Cable Warren
16:00 PM - 16:45 PM
Automating Salt Interpretation in the Gulf of Mexico with AI: A Case Study on Using Open Source Python Libraries MDIO, MLFlow, and PyTorch
Bio: Cable Warren holds both Bachelor's and Master's degrees in Geophysics from Virginia Tech and currently serves as a Geophysical Advisor for Data and Analytics at TGS. In this capacity, he leverages AI techniques to address real-world geophysical challenges that industry professionals encounter daily. Before his tenure at TGS, Cable spent eight years at BP's Houston location. He explored numerous areas and basins, such as offshore South America, Africa, Mexico, and GoM. His career at BP commenced with work on the Thunder Horse and Nakika fields in GoM development.
Abstract: The talk will focus on the importance of open source python libraries to this effort, including mdio.dev, which was recently open sourced by TGS to the benefit of the geophysical community. MDIO is a significant enabling technology, it’s a Python native library and builds on several modern Python libraries including ZArr and Dask. Outside of traditional energy, mdio has also been finding applications in new energy, powering data and analytics in the domains of wind and solar.
Jacqueline Ming
Committee Chair
Jackie Ming has a dual role as both the EVP of Sales at SOJA Efficiency Consultants and the VP of Business Development with Tech Limit, a geomechanics firm based in Australia, currently bringing their software tools to market in North America. Previously Jackie spent over ten years in business development in the upstream oil & gas sector with both IKON Science and CGGVeritas/Hampson Russell. Her primary responsibilities include building technology awareness and networking within larger operators such as ConocoPhillips, Chevron, Apache and Noble Energy. Jackie holds an Executive MBA from the University of Houston with a focus on Global Leadership, and a Bachelor of Business Administration in Management, also from the University of Houston. At the GSH, Jackie has been actively involved serving on the Board and various committees (’07-’16). She is excited to now serve as the Chair for the GSH 2023 Spring Symposium.
Simon Voisey
1st VP GSH
Simon Voisey is GSH's 1st VP and a staff geophysicist specializing in quantitative interpretation (QI) at Apache Corporation based in Houston with 18 years of industry experience. He holds an MSc in Petroleum Geoscience from the University of Aberdeen and a BSc in Geophysics from UCL (University College London). He has played active roles in both the SEG and GSH. Simon has served in 2023, 2022, 2021 & 2019 GSH spring symposium planning committees. Simon is on the SEG DISC committee, reviews abstracts, and chaired the 2019 AVOSI technical sessions for the SEG annual conference. 2004 Simon started his geophysical career in London at Scott Pickford. In 2005, he joined Hampson-Russell software and moved to Houston in 2009. 2014 Simon joined Apache Corporation and continues to work in their Applied Geoscience Technology (AGT) group, focusing mainly on Egypt and North Sea assets. In Simon's free time, he is a keen hiker and has trekked many of America's National Parks. In addition, he is a season ticket holder for the Houston Dynamo soccer team and the Houston SaberCats rugby team.
Norbert Van De Coevering
Technical Chair
He graduated from Utrecht University in the Netherlands with a Masters in Geophysics and started his career almost 26 years ago with CGG in London. Before joining Oxy in Houston in 2016 he worked at Murphy. Currently he is Manager Geophysics Delaware Basin for Oxy after returning from Oman and served as AVOSI Session Chair for IMAGE21 and SEG 2020 Conferences. His main interests are in Seismic Data Conditioning and all aspects of Quantitative Interpretation and data integration. He has (co-) authored various publications.
Gabino Castillo
Technical Chair
Subject Matter Expert - Quantitative Interpretation at APA Corporation. Graduated from IFP School, France in 1999, Gabino holds a Master in Geophysics and in Computer Science. The main areas of interest are rock physics driven pre-stack seismic data interpretation, seismic inversion technologies from pre-stack simultaneous inversion to azimuthal, 4D and stochastic approaches. He also enjoys working in integrated projects and to use machine learning and data analytics techniques. Python enthusiast, Python has been fundamental component of his work, in a typical Python way, it spreads virally throughout different G&G workflows. Before joining APA, he was Seismic Reservoir Characterization Regional Manager at CGG and Consulting Services Director at Sharp Reflections.
Lillian Jones
1st VP Elect GSH
Lillian Jones is the GSH 1st VP Elect. She has worked at Apache Corporation for almost 10 years as a geophysicist; she currently works on the Suriname Appraisal Team. Lillian received her MSc in Geophysics from Colorado School of Mines in 2013. She has been an active GSH member and volunteer for many years, serving as the GSH Secretary in 2017. Lillian spends her free time with her husband and daughter. She is an avid reader, cook, and gardener.
Sponsorship
Requesting Committed Sponsors at This Time
The Geophysical Society of Houston is 501(c)(3) Nonprofit organization. Please support the society and our premiere annual event.
Sponsorship
Levels
FULL EVENT - April 19 & 20
Sponsorship Level
Bronze
Silver
Gold
Platinum
Diamond
Amount
$500
$1000
$2500
$5000
$10,000
EVENING EVENT- April 19
Sponsorship Level
Jade
Ruby
Sapphire
Amount
$500
$1000
$3000
PARTICIPATION
Item
Session Sponsor
(company’s name by
session’s publications)
Booth
(includes: (i) 2 x entire event tickets (ii) Table, 2 chairs, space for floor standing banner all within Lecture Hall
​
Challenge Bowl Sponsor
Program Advert
(full page)
Program Advert
(1/2 page)
Part sponsor for food
(company name by food throughout the event)
Amount
$1500
$1000
$1000
$500
$500
Complimentary Registration(s)
0
1
2
3
6
Complimentary Guests​
2
4
6
$2500
Sponsors
Gold level
Diamond level
Silver level
Bronze level
Evening Event
Challenge Bowl
The Geophysical Society of Houston has hosted the Gulf Coast Regional SEG Challenge bowl since 2007. The prize for this year’s winners will be $1000 or $500/player as well as the opportunity to represent the Gulf Coast in this year’s international finals. Last year’s Finals had 15 teams from 11 countries.
April 20 - Day 2
Thursday 12:00 - 1:00 PM
Call for Teams
-
Teams: ​Teams are composed of two students, either from the undergraduate or graduate program, including those graduating this Spring. Only competitors in last year’s international finals are barred from this year’s Bowl.
-
Registration: We welcome teams from all Gulf Coast schools. However, the GSH cannot pay for travel/accommodations. Players have a reduced registration. Players can also volunteer for the entire event to receive free registration.
​
-
More Info: Please visit the SEG Challenge bowl website for more info: https://seg.org/Education/Student/SEG-Challenge-Bowl