Satellites, Software, and Polar Cyberinfrastructure: Creating a Data-to-Information Pipeline for Antarctic Science

High-resolution satellite imagery now allows us to map the presence and absence of Antarctic biology at unprecedented resolution (individual penguin nests) over unprecedented spatial extents (the Antarctic continent), largely solving the problem of mapping wildlife on the world's most remote and forbidding continents. However, the petabytes of data now available for the Antarctic creates new challenges of an entirely different sort: How do we automate the processing of high resolution satellite imagery, how do we design systems that integrate human interpretation and computer vision for maximally efficient workflows, and how do we re-package these raw data into information that is meaningful and accessible to the Antarctic policy and management community. This work sits at the interface of wildlife ecology, conservation, geography, computer vision, applied statistics, and high performance computing; in my seminar, I will discuss how far we've come and what lies ahead in constructing a seamless data-to-information pipeline for the Antarctic.