AAS WorldWide Telescope Project Awarded NSF Grant

I’m delighted to announce that the AAS WorldWide Telescope (WWT) team was awarded a three-year grant from the National Science Foundation's (NSF) Office for Advanced Cyberinfrastructure through its Cyberinfrastructure for Sustained Scientific Innovation (CSSI) program. This grant will support the development of new features in WWT to enable the interactive visualization of the enormous images that emerge from cutting-edge telescopes like the Rubin Observatory. Along with this software development effort, the team will create an online learning environment in the Khan Academy model that will teach astronomers some of the core skills for handling these kinds of large, modern datasets — the Little Big Data University (LBDU). 

What motivated this grant proposal? For many (but not all!) astronomers, a bread-and-butter part of their workflow is to look at images of the sky. Ironically, as we enter the Large Synoptic Survey Telescope (LSST) era in which the sky is imaged more comprehensively than ever, we are often losing our ability to do this basic activity. While traditional image viewers load up files from their hard drive into memory for display, images from modern telescopes are so large and numerous that it is impractical or impossible to download them from the observatory to the computer or even to load them up into memory at once. 

While it’s true that a lot of science can be done with image cutouts or other forms of data subsetting, we believe that powerful exploratory data visualization is essential to getting the most science out of the new generation of imaging telescopes. And with the data volumes associated with these telescopes, this visualization has to work over the web — which is where WWT comes in. Thanks to its “tiled image pyramid” capability, the WWT rendering engine can interactively display gigapixel-class astronomical images in your browser — but right now it can only display colorized RGB images (e.g., JPEGs) in this fashion. The NSF grant will support the FastTract project to expand this capability to include tiled pyramids of floating-point data (e.g., FITS files), unlocking web-based exploratory visualization of raw data like full LSST tracts or DASCH plate scans. Importantly, the grant will support work to build out tools and a community of practice and empower as many astronomers as possible to start using this technology. 

The problem addressed by this project is one example of the broader challenge of the “data deluge”: astrophysicists and other domain scientists will have to learn how to work with ever-larger data sets, even though the vast majority of them are never going to become experts in computer engineering nor would become their primary fields. Given that reality, what are the key lessons that domain scientists need to learn so that they can actually make the most of the data that their instruments will produce? The Little Big Data University project will attempt to answer this question, designing and releasing a data curriculum for scientists guided by the insights gained during the FastTract effort. LBDU will adopt the empirically successful Khan Academy model, offering a series of bite-sized lessons compatible with the busy schedule of a typical working scientist. 

The AAS is now putting together the team that will work on these projects for the next three years. If you would like to be involved, please reach out to the WWT Director Peter Williams for more information about project plans and upcoming opportunities.