Similarity Search uses self-supervised learning, a machine learning technique, to improve the efficiency of data search and collection. By automating the process of combing through satellite imagery archives to locate images or data of interest—for example, wildfires or dust storms—this tool dramatically reduces the time required to create relevant datasets.

Currently, Similarity Search supports Moderate Resolution Imaging Spectroradiometer (MODIS) datasets access through NASA’s Global Imagery Browse Services (GIBS).