Observing Ecology with CyVerse Atmosphere
With a continental-scale observation facility, the National Ecological Observatory Network (NEON), is collecting more ecological data than any other single ecological project in history.
“NEON data are incredibly high quality, thanks to their diligent staff and scientists,” said Tyson Swetnam, a CyVerse computational ecologist and GIS (Geographic Information System) mapping technology expert.
NEON, which is powered by science and technology
firm Battelle, is a National Science Foundation project charged with collecting data from across the North American continent that characterize how U.S. ecosystems are changing. However, Swetnam noted, NEON isn’t in charge of doing analyses on their processed data product. That job is up to the scientific community.
“CyVerse is in the unique position to provide the ecological science community with a cyberinfrastructure that can scale to NEON data,” he said.
This week in Boulder, Colorado, Swetnam is teaching NEON scientists and collaborators how to use the Python programming language to create reproducible workflows with the CyVerse Atmosphere cloud-computing platform.
The CyVerse training is part of the NEON Data Institute’s week-long annual workshop. This year the Institute is honing in on remote sensing, focusing on how researchers can harness computational tools and techniques to share data and analyses.
Swetnam’s presentation focuses on vertical scaling of remote sensing data to suit the data collection ambitions of NEON researchers who are now harnessing drone technology for landscape observation systems.
“CyVerse was a crucial component to the 2018 Data Institute in providing participants tools for scaling up their research workflows,” said Bridget Hass, a NEON science technician who works with remote sensing. “We received feedback from 2017 participants stating the need for moving beyond using a single flightline or tile, and CyVerse provides a powerful resource for large-scale remote sensing data analysis.”
With CyVerse’s Atmosphere platform, scientists can launch their own virtual machines on an easy-to-use web application, with access to compute resources including high-performance computing, software suites, and preconfigured analysis routines. Atmosphere was built for data-intensive bioinformatics challenges such as NEON’s continent’s worth of data.
Photos by: Tyson Swetnam/CyVerse