Interferometric imaging of geostationary satellites

Even the longest geosatellite, at 40 m, subtends only 0.2 arcsec (1 microradian). Determining structure and orientation with 10 cm resolution requires a 90 m telescope at visual wavelengths, or an interferometer. We de- scribe the application of optical interferometry to observations of complex extended targets such as geosatellites, and discuss some of its challenges. We brie y describe our Navy Optical Interferometer (NOI) group's eorts toward interferometric observations of geosatellites, including the rst interferometric detection of a geosatellite. The NOI observes in 16 spectral channels (550{850 nm) using up to six 12-cm apertures, with baselines (separa- tions between apertures) of 16 to 79 m. We detected the geosatellite DirecTV-9S during glint seasons in March 2008 and March 2009, using a single 16 m baseline (resolution 1:6 m). Fringes on a longer baseline were too weak because the large-scale structure was over-resolved. The fringe strengths are consistent with a combination of two size scales, 1:3 m and & 3:5 m. Our near term NOI work is directed toward observing geosatellites with three or more 10 to 15 m baselines, using closure phase measurements to remove atmospheric turbulence eects and coherent data averaging to increase the SNR. Beyond the two- to three-year time frame, we plan to install larger apertures (1.4 and 1.8 m), allowing observations outside glint season, and to develop baseline bootstrap- ping, building long baselines from chains of short baselines, to avoid over-resolution while increasing maximum resolution. Our ultimate goal is to develop the design parameters for dedicated satellite imaging interferometry.