Reverse Geolocation of Images Taken from the International Space Station Utilizing Various Lightning Datasets

The crew aboard the International Space Station capture many images of the Earth's surface and atmosphere with handheld digital single-lens reflex (DSLR) cameras. These images can contain interesting lightning and meteorological phenomena, however contain little metadata past basic parameters; camera angles are generally unknown and timing is not known to high precision. As a result, the scientific potential of many of these images goes unrealized. Current lightning sensing technologies used to produce the datasets referenced for geolocation have high sample rates - in the millisecond range; but comparatively poor spatial resolution to standard high-resolution imagery attained by DSLR cameras. Cross-referencing Lightning Locating Systems (LLSs) sensor data and high-resolution imagery the orientation of the image can be determined, allowing for the calculation of latitude and longitude at each pixel. The process for stand-alone images was then extended to video of severe storms. Observations of the Chiba University Meteor Camera (METEOR) project containing high frame-rate video severe storms were segmented following standard image processing methods. Lightning flashes were identified with high accuracy. The video was orientated using ground control points; allowing for computationally inexpensive approximations. Accurate timing allowed for the matching of geometries between datasets and the subsequent conversion of individual frames to highly accurate latitude/longitude-coordinate-space. Lightning flashes were measured up to the pixel, or ~30-meter accuracy level. This geolocation effort now provides Marshall Space Flight Center's Earth Science Branch and Johnson Space Center's Earth Science and Remote Sensing Unit with tools for analysis of images and video from the International Space Station for scientific analysis of severe thunderstorms and fundamental processes associated with lightning propagation.