Houston 2010

From Madagascar
Jump to navigation Jump to search

 

Madagascar School on Reproducible Computational Geophysics and Hands-On Workshop
Sponsored by PTTC Texas/SE New Mexico Region
Photo courtesy of Joe Dellinger

 

Program[edit]

Day 1: Friday, July 23
9:00-10:30 Introduction (Sergey Fomel)

The Madagascar project has been in public existence for four years. Madagascar provides a complete environment for organizing one's research, from new software development to running computational experiments to publishing the experimental results in papers and reports, archiving them for future usage, and sharing them with colleagues and sponsors. The introductory presentation will describe the history of the project, the Madagascar components and design principles, and the future development goals.

Introduction to Madagascar (1.2M)

10:30-10:45 break
10:45-12:15 Seismic finite-difference modeling and migration example (Paul Sava)

The theoretical part of this module provides an overview of reverse-time imaging methodology applied to wavefield seismic data. The main technique discussed is reverse-time migration with emphasis on modern imaging conditions which enable migration velocity analysis and amplitude-versus-angle analysis. The applied part demonstrates this technique on a complex geologic model using Madagascar codes in a fully reproducible setup.

Seismic imaging tutorial (744K)
SConstruct

12:15-1:15 Lunch
1:15-2:45 Workflows in SCons and automatic testing (Jim Jennings)

The rich and well-documented Python syntax used in SConstruct files provides great flexibility when coding Madagascar SCons workflows. In the first part of this module a few geostatistical workflows will be presented to illustrate some useful techniques.

Data processing flows with SCons (7.5M)
SConstruct

Two important components of the Madagascar design goals are reproducibility and regression testing. In the second part of this module our progress towards these goals will be discussed and some of the built-in tools for automatic testing will be presented.

Automatic testing in Madagascar (148K)

2:45-3:00 break
3:00-4:30 Discussion

Open Q&A session and discussions on the future development of Madagascar

5:30-8:00 Dinner and Madagascar 1.0 celebration

 

Day 2: Saturday, July 24
9:00-9:30 Learning Madagascar (Tariq Alkhalifah)

Why Madagascar was easy for me? The art of the Madagascar template! In a year, I managed to write 7-8 papers using Madagascar, and the papers are still coming. Considering that I have used SU (Seismic Unix) all my life and taking into account my not-so-young age, I think Madagascar was good to me. In the presentation, I will share my experience (no Python background knowledge is needed!) and share some insights on how to use Madagascar efficiently.

Learning Madagascar (444K)
SConstruct

9:30-10:30 Programming with Madagascar (Jeff Godwin)

There are many programs already built into the Madagascar project, but if you use Madagascar long enough you will eventually run into a problem that you cannot solve using only provided codes. Fortunately, Madagascar has a variety of programming language APIs already built, that allow you to: design, code, and integrate your programs into the Madagascar framework. This session will discuss an overview of the Madagascar APIs, and then focus in particular on the C and Python APIs. By the end of the session, you should have a good starting point for developing your own codes, and adding them to the growing library of open-source software available in Madagascar.

Programming in Madagascar (148K)
SConstruct

10:30-10:45 break
10:45-11:15 Vplot graphics language - past, present, and future (Joe Dellinger)

Back in the 1980's we had a lot of different hardware graphics devices to support. Now we are back to the future... the challenge today is to support a multitude of different software graphics devices. The same structures that were useful then are useful now. I will talk about the philosophy of vplot and how it is as useful today as it was when it was devised. I will also show some vplot tricks you can use to astound your friends.

Vplot (3.6M)

11:15-12:15 Plotting and high-performance computing with Madagascar (Vladimir Bashkardin)

The first part of this module will cover visualization and preparation of data plots with Madagascar tools. Major available styles of plots will be explained as long as the mechanism of creating figures in reproducible fashion for papers and presentations. The second part will be a tutorial for using Madagascar in a high-performance environment. Different types of such environments will be described and principal approaches to handling parallel problems with Madagascar will be unfolded. This module will specifically address how to run data-parallel tasks and how to create Madagascar programs with MPI and GPU(CUDA) routines.

Plotting in Madagascar (2.5M)
SConstruct

High-performance processing and development with Madagascar (236K)

12:15-1:15 Lunch
1:15-2:45 Seismic field data processing example (Yang Liu)

Field data processing is always a tough problem. We will use a 2-D public dataset (From FreeUSP) to illustrate how Madagascar can implement common seismic data processing tasks.

Seismic field data processing example (12M)
SConstruct

2:45-3:00 break
3:00-4:30 Discussion

Open Q&A session and discussions on the future development of Madagascar

Location[edit]

The University of Texas at Austin
Bureau of Economic Geology
Houston Research Center

Address


Dinner Location[edit]

Los Candiles Banquet Hall
6518 Brittmoore Rd
Houston, TX 77041-5241
(713) 896-0235


Speaker biographies[edit]

  • Tariq Alkhalifah is currently a Professor of Geophysics at KAUST in Saudi Arabia. He graduated with a PhD from Colorado School of Mines, Golden, Colorado, in 1996, and served afterwards as a Post Doc at Stanford University for 2 years sharing an office with Sergey Fomel. I used to be a devote SU Unix follower for most of my research carrier even as a Post Doc at Stanford (SEPlib people), but I have recently seen the light and converted to Madagascar. https://sites.google.com/a/kaust.edu.sa/tariq/
  • Vladimir Bashkardin is currently a PhD student in geophysics at The University of Texas at Austin. Before joining the research group of Dr. Sergey Fomel at UT Austin, he worked as a software engineer for Paradigm (former Paradigm Geophysical) with specialization in seismic data visualization and interpretation. He also was a part-time lecturer at Gubkin Oil and Gas University (Moscow, Russia), an industry-oriented school from which he holds a degree in exploration geophysics.
  • Joe Dellinger graduated with a PhD in Geophysics from the Stanford Exploration Project in 1991 and currently works for BP in Houston, specializing in anisotropy and multicomponent seismology. Joe has often provided advice to the SEG (much of it unsolicited) on how they should best advance into the brave new online/digital world, for which he was awarded Life Membership in 2001. Joe currently is the editor of the Software and Algorithms section of GEOPHYSICS, and maintains the accompanying software and data website. http://software.seg.org
  • Sergey Fomel has been working at the Bureau of Economic Geology at the University of Texas at Austin since 2002 and currently has an Associate Professor appointment, jointly with the Department of Geological Sciences. He received a Ph.D. in Geophysics from Stanford University in 2001 and worked previously at the Institute of Geophysics in Novosibirsk, Russia, and the Lawrence Berkeley National Laboratory. Sergey started work on Madagascar (at that time named RSF for Regularly Sampled Format) in 2003. http://www.beg.utexas.edu/fomel/
  • Jeff Godwin is currently a MS student in geophysics in the Center for Wave Phenomena (CWP) at the Colorado School of Mines (CSM). He received his undergraduate degree in Geophysics (2009) from CSM as well, and has interned with Hess, Marathon Oil, and Landmark Graphics in various positions. His research interests include: high performance computing, inverse methods, and seismic imaging.
  • Jim Jennings currently works for Shell International Exploration and Production in Houston Texas as a company consultant on carbonate reservoir modeling, but he contributes to Madagascar as a hobby and will be participating in the workshop on his own time. Before joining Shell in 2007 he worked for 23 years at the Bureau of Economic Geology, Arco, and BP. Jim has a PhD in Petroleum Engineering from Texas A&M University (1983), was chairman for an SPE reprint volume on Advances in Reservoir Characterization (2006), and was a Distinguished Lecturer for the AAPG (2008-2009). http://www.aapg.org/education/dist_lect/jennings.cfm
  • Yang Liu is currently a Postdoc at Bureau of Economic Geology, The University of Texas at Austin. He received a Ph.D. in Geophysics from Jilin University (China) in 2006. His research focuses mainly on seismic data processing. http://www.beg.utexas.edu/personnel_ext.php?id=50
  • Paul Sava is an Assistant Professor of Geophysics and a member of the Center for Wave Phenomena at Colorado School of Mines. He holds an Engineering degree in Geophysics (1995) from the University of Bucharest, an M.Sc. (1998) and a Ph.D. (2004) in Geophysics from Stanford University where he was a member of the Stanford Exploration Project. His research interests are in wavefield seismic imaging, stochastic imaging and inversion, computational methods for wave propagation, numeric optimization and high performance computing. http://newton.mines.edu/paul/home.php