SCINet Scientific Computing
Geospatial Research Working Group
This working group was formed as the result of a September 2019 SCINet-funded workshop hosted by Deb Peters, the acting ARS Chief Science Information Officer. The focus of the group is to provide continued input on the development of SCINet, to determine the computational needs of ARS geospatial researchers, and to develop collaborative research projects. The group will also learn from each other on topics related to high-performance computing. For example, how to use Ceres (the ARS high-performance computer), optimizing code for parallel processing, getting software onto Ceres, etc. Monthly working group teleconference calls and working sessions are beginning soon.
If you would like to join the group, please contact Kerrie Geil to be added to the group’s Basecamp.
| Working Group Leadership Team | Title | Location |
|---|---|---|
| Kerrie Geil | SCINet Postdoc | Las Cruces, NM |
| Rowan Gaffney | Physical Scientist | Ft Collins, CO |
Workshops
August/September 2020: SCINet Geospatial Workshop 2020
Aug 25 - Sep 1, 2020 · remotely on Zoom · lead: Kerrie Geil, Rowan Gaffney
Visit the workshop website for more information
The working group held their annual workshop over 6 separate Zoom sessions which were attended by over 60 scientists, post-docs, research leaders, and data managers. The sessions included an annual meeting of the working group, four tutorials on high-performance computing including 1 tutorial on the use of machine learning, and a symposium on the use of AI techniques in agricultural research. Detailed information on all the sessions, including recordings (to be posted soon), can be found on the session pages of the workshop website:
Session 1: Annual Meeting of the SCINet Geospatial Research Working Group
Session 3 Tutorial: Intro to Distributed Computing on the Ceres HPC System Using Python and Dask
Session 5 Tutorial: Distributed Machine Learning: Using Gradient Boosting to Predict NDVI Dynamics
September 2019: Artificial Intelligence and the Accessibility and Analysis of Geospatial Data: A SCINet Workshop
09/10/2019 - 09/11/2019 · Las Cruces, NM · lead: Deb Peters
A group of over 36 geospatial scientists, post-docs, research leaders, and data managers gathered at the ARS Jornada Rangeland Research Unit in Las Cruces, NM on September 10-11, 2019 to discuss high-performance computing (HPC) issues and artificial intelligence research methods applied to geospatial problems. The group identified computational issues with accessing and using SCINet and the ARS HPC system for geospatial research and also gained exposure to relevant machine learning and deep learning research methods. The workshop resulted in the creation of a SCINet Geospatial Research Working Group to continue identifying and addressing complex computational problems as well as to collaborate on new geospatial research projects.
Workshop Agenda (opens pdf in this browser window)
Participant List with Poster/Presentation Titles (opens pdf in this browser window)
Workshop Notes (downloads Microsoft Word document)
Visit the SCINet-funded Workshops Basecamp for posters, presentation slides, and recordings of the presentations. ARS employees can contact Kerrie Geil for access to the Basecamp project.