The CIG Webinar Series draws from a pool of experts from mathematicians, to computer scientists, and to geoscientists, among others to bring together a cross-cutting community of faculty, students and researchers to both inform and disseminate knowledge on the tools and methodologies employed to further the study of problems in geodynamics.
The 2016-17 series aims to expand upon our previous code tutorials by targeting select code features widely used by the community and expanding our knowledge about best practices and tools in software development.
The one hour webinars will be held the 2nd Thursday of each month October through May. Webinars will be recorded for later viewing. Reminders and details will be sent out through the cig-all mailing list.
To connect to a meeting, click on the meeting link, then select Enter as Guest, type your on screen Name for the meeting, and then click on Enter Room.
Detailed instructions are available here (PDF)
Click here to Connect to Meeting
October 13 - Anshu Dubey, Argonne National Laboratory, Software Practices in Computational Science Communities – an Overview
November 10 – Rene Gassmoeller, CSU Fort Collins, Intricacies of particle-in-cell methods in convection models with adaptive meshes: Using ASPECT's particle implementation
February 9 – Ian Rose, UC Berkeley, Project Jupyter for the geosciences
March 9 – Hom Nath Gharti, Princeton University, Introduction to the spectral-infinite-element method
April 13 – Brad Aagaard, USGS
May 11 - Julianne Dannberg, CSU Fort Collins
Do you have a suggestion for or have heard a talk recently you think may interest the CIG community? Let us know by contacting firstname.lastname@example.org
Introduction to the spectral-infinite-element method
Hom Nath Gharti, Princeton University
Software Practices in Computational Science Communities – an Overview
Anshu Dubey, Argonne National Laboratory
Scientific code developers typically adopt software processes derived from the mainstream (non-scientific) community when continuing without them becomes impractical. However, many software best practices need modification and/or customization, partly because the codes are used for research and exploration, and partly because of the combined funding and sociological challenges. This webinar will describe the lifecycle of scientific software and important ways in which it differs from other software development. We will provide a compilation of software engineering best practices that have generally been found to be useful by science communities, and how they are evolving as the needs of their communities grow. [slides] [no recording]
CSU Fort Collins, Intricacies of particle-in-cell methods in convection models with adaptive meshes: Using ASPECT's particle implementation
Rene Gassmoeller, CSU Fort Collins
Particle-in-cell methods have a long history in modeling of mantle convection, lithospheric deformation and crustal dynamics. However, their efficient parallel implementation and application in models - in particular combined with adaptive meshes - is involved due to the complex reassignment of particles to cells and frequent parallel communication.
In this webinar, I present the implementation of a flexible, scalable and efficient particle-in-cell method for the massively parallel finite-element code ASPECT. I discuss the complexity of the implemented algorithms, present scaling tests and discuss load-balancing strategies like balanced repartitioning for particles in adaptive meshes with their strengths and weaknesses. Additionally, I will show an application tutorial on how to convert a model with compositional fields into one using the particle advection scheme, and which consequences follow from this conversion for model runtime and accuracy.
Project Jupyter for the geosciences
Ian Rose, UC Berkeley
As the results of scientific computing become more central in the geosciences, we have been confronted with a series of challenges that were not necessarily obvious from the outset. How does one ensure that the process by which the geoscientist arrives at a result is transparent and reproducible? What are the most effective ways to communicate a computational result? How does one teach computational geoscience to students?
Project Jupyter is a suite of tools for scientists and educators which helps to address these questions. Here I demonstrate the Jupyter (formerly IPython) notebook, which allows one to combine prose, equations, code, and the outputs of code, all in the same computational environment. Jupyter notebooks run in a web browser, which allows it to leverage the rich rendering capabilities of the modern web environment, including images, tables, audio, and video. It also has the ability to instrument code with interactive user interface elements like sliders and buttons, allowing for more user friendly interaction with scientific software.
I also discuss some more advanced usage, including JupyterHub deployments for use in the classroom. The web-based nature of the notebook allows teachers to serve notebooks to students over the internet. This allows students to learn programming in a custom environment without having to install or configure anything. This ability to serve computing environments could also be leveraged by an organization such as CIG to disseminate and advocate for software in the geosciences. [slides on github]