CATALOGING ASTRONAUT PHOTOGRAPHY FROM THE INTERNATIONAL SPACE STATION (ISS)

The Gateway to Astronaut Photography of Earth (https://eol.jsc.nasa.gov/) hosts photographs of Earth taken by astronauts on the International Space Station (ISS). The ISS serves as a unique remote sensing platform, and the images obtained of the Earth’s surface show changes over time associated with human impacts, such as urbanization, and dynamic natural processes such as volcanic eruptions and hurricanes. We are one of three cataloging teams in Texas who have been contracted through Jacobs Engineering to catalog photographs for NASA with precise center coordinates, estimates of cloud cover, and labels for visible geographic, geomorphic, and anthropogenic features. The cataloging is done using Catalite, a NASA software application, in conjunction with Google Earth Pro. The metadata associated with each cataloged photograph is used as search parameters to help users locate photos in the database. With over 4 million photos currently in the database, only ~300,000 have been manually cataloged. Since 2019, teams of UTEP undergraduates have cataloged over 9,000 photographs. Our current efforts at UTEP have two objectives: (1) cataloging an additional set of 2,000 photographs; and (2) generating object label training data using the online software tool Make Sense (https://www.makesense.ai/). The training dataset is generated for a new initiative that will enable the Earth Science Remote Sensing (ESRS) team at NASA Johnson Space Center to train machine learning algorithms to automatically "Geo-tag" features that are recognized within millions of uncatalogued photos. To do this, we are identifying a small set of geomorphic features (volcanoes, rivers, glaciers, lakes, islands, and bays), anthropogenic features (agricultural fields and cities), and indications of active processes (smoke plumes) in a set of 1,000 photographs that were previously cataloged by UTEP. This activity improves the searchability of the database, allowing researchers, K-12 students, and average citizens worldwide to have easier access to this free and unique scientific dataset.