- Parent Category: outreach
- Published on Monday, 23 November 2020 09:03
2017–2018 Webinar Schedule
October 12. Taras Gerya, ETH Zurich. Geodynamic modeling with staggered finite differences and marker in cell: theory, teaching and examples
November 16. Max Rudolph, University of California, Davis. Tools and approaches for teaching computation and modeling: geodynamics and beyond
February 8. Gabriele Morra, University of Louisiana at Lafayette. Pythonic Geodynamics
March 8. Eric Mittelstaedt, University of Idaho. Where have all the dimensions gone? Hands on methods for introducing students to non-dimensional numbers in laboratory and numerical modeling
April 17. Lion Krischer, Simon Staehler, and Martin van Driel, ETH Zurich; Tarje Nissen-Meyer, Oxford University. New developments in AxiSEM/Instaseis for seismic wave propagation on local scales
May 10. Rene Gassmöller, Juliane Dannberg and John Naliboff, UC Davis; and Menno Fraters, Utrecht University. ASPECT 2.0: Improved architecture, new features
THURSDAY, OCTOBER 12
Numerical modeling of geodynamic processes is an essential approach in both science and industry with ever- growing demand and high efficiency/cost ratio. Current trend in geodynamic modeling is to develop universal approaches with potentially unlimited number of applications.
One simple and flexible method is based on staggered finite differences and marker in cell techniques (SFD-MIC), which demonstrated superior performance in several branches of modern quantitative Earth sciences. It is suitable for modeling various long-term and short-term thermomechanical processes involving large 3D deformation of rheologically complex materials. Recently, potential applicability of this method to technological processes (material science) and natural processes of industrial significance (geo-hydro-mechanics, waste deposits) has also been demonstrated.
This webinar gives a short theory of the SFD-MIC method, discuses Matlab-based teaching approach and presents modeling examples of natural and technological significance. [slides]
THURSDAY, NOVEMBER 16
I will share experiences teaching computation and modeling using a variety of tools and techniques. First, I will present my philosophy and approach to teaching modeling at the undergraduate level using jupyter notebooks and python. I will describe a set of exercises designed to familiarize students with the conservation laws and the properties of partial differential equations relevant to modeling geologic processes as well as simple numerical approaches based on low-order finite differences. I will my share experiences and provide a brief demonstration of the functionality of the jupyterhub environment with nbgrader, which provides an integrated environment for assignment creation, distribution, and automated grading of computer laboratory assignments. Second, I will discuss approaches used in a graduate level geodynamic modeling class. I will provide examples of curricular materials and student projects that use the ASPECT mantle convection code. I will demonstrate the integration of ASPECT with jupyter notebooks to reproduce classic results associated with the onset of convection and mantle mixing processes.
THURSDAY, FEBRUARY 8
Students and young researchers who want to learn to use computational tools for geodynamic modeling have the option to choose among a wide range of numerical tools. I will show how Python and its libraries represent an easy-to-use platform for self-learning, with performance close to compiled codes. I will present (1) how to visualize and run vectorial calculations, (2) examples from classical Mechanics like particles trajectories in 2D-3D, (3) a detailed description of how to write Lagrangian, Eulerian and Particles in Cell codes for solving linear and non-linear continuum mechanics problems and (4) advanced techniques like tree-codes, Boundary Elements, Lattice Boltzmann Method, as well as use Jupyter Notebooks for creating and distributing content. The goal is to encourage professional and students to learn by experimenting and experiencing, like children who learn by playing. [slides][Jupyter Notebooks]
THURSDAY, MARCH 8
Experienced modelers are familiar with how non-dimensionalizing mathematical systems can help improve numerical stability, reduce the number of free variables needed to explain a physical system, capture the essential driving processes of a problem of interest, and scale laboratory experiments to the Earth. However, when first introduced to non-dimensional numbers, students often have difficulty understanding how the mantle can have a depth of 1, or how numbers such as the Nusselt number or the Rayleigh number are derived. In this webinar, I will discuss a hands-on, in-class experiment involving a simple oscillator (mass on a spring) that I have used to introduce students to non-dimensionalizing equations, deriving non-dimensional numbers, and scaling experimental results. The mass-spring system is familiar to many students from their introductory physics classes and the mathematical system is simpler than many problems of interest in geodynamics. The combination of familiarity, basic mathematics, and a hands-on experiment facilitate student comprehension and future application of non-dimensional numbers. These methods are aimed at an introductory graduate course or senior level undergraduate course on modeling.
TUESDAY, APRIL 17
Instaseis (http://instaseis.net) is a Python tool to quickly extract high-frequency seismograms for any source-receiver geometry from databases generated with AxiSEM (http://axisem.info/). So far it has been limited to global datasets, which in turn limited the maximum frequencies due to exploding storage requirements; a number of global databases are hosted by the IRIS DMC (http://ds.iris.edu/ds/products/syngine/). We recently expanded the AxiSEM/Instaseis combination to additionally handle local scale simulations as well as databases, which enables much higher frequencies.
In this webinar, we present a short overview of the theory behind AxiSEM/Instaseis, highlight the new extensions for local scale simulations and databases, and show a short practical tutorial on how to use it.