A community oriented approach to enabling open science with Earth science data at scale

To enable wider participation in open science with geospatial data at scale, we need to reduce the effort and custom approaches required for setting up scalable scientific data analysis environments and computing workflows. We have made great strides in this pursuit by evolving and promoting community-developed open source frameworks, tools, and libraries for cloud-native data access and analysis, making them the default for scientists on the public cloud and local systems.

Many of our achievements have been supported by the NASA Visualization, Exploration, and Data Analysis (VEDA) project which seeks to proliferate cloud-native approaches for open science on Earth science data from NASA’s rich archives and many other providers. Our presentation highlights how we have engaged with communities like Pangeo, OpenScapes, Earth Science Information Partners, and the Cloud Native Geospatial Forum to build joint initiatives, target development, and ensure uptake of new solutions. We present key results from working groups, community showcases, and hackdays and hackweeks organized by VEDA team members, as well as specific contributions to the open source ecosystem, including the eoAPI platform for quickly and easily deploying an open-source Earth Observation stack, JupyterHub fancy profiles (with BinderHub) for seamless environment building, and Lonboard for fast, interactive vector visualization.