10.1117/2.1201104.003589 A low-cost thermal IR hyperspectral imager for Earth observation Sarah T. Crites, Paul Lucey, Robert Wright, Harold Garbeil, and Keith Horton A compact and efficient sensor designed for small satellites uses an uncooled, commercial microbolometer array to measure interferometric IR radiation. A novel niche for Earth observations exploiting new tech- nologies in focused, short-duration missions is opening with the growth of the market for small satellites. To demonstrate the ways in which a university’s scientific and instrument development programs can innovate under this new model, we are building a low-cost thermal IR spectral sensor. The sensor is a low-mass, power-efficient thermal hyperspectral imager (THI) with electronics contained in a pressure vessel to enable the use of commercial-off-the-shelf (COTS) electronics (see Figure 1). Since the early 1990s, we have been developing spatial Fourier transform thermal hyperspectral imagers for Earth-based and airborne remote sensing as well as laboratory thermal IR microscopy. 1–4 The space-qualified THI is an imple- mentation of this approach. It will allow us to assess our data quality, map out improvements to the design, and quantify cost savings possible through a COTS-based approach. Current space-based thermal spectral sensors for science of Earth’s surface are multispectral, collecting measurements of IR radiation in discrete bands. True hyperspectral data has been collected for atmospheric science by, for example, NASA’s Atmospheric Infrared Sounder instrument, which collects data in 2378 channels between 3.7 and 15:4./m, 5 but only at low spatial resolution. Earth surface science can greatly benefit from hyperspectral imaging—gathering reflectance or emittance data in tens or hundreds of narrow, contiguous wavebands— especially to derive laboratory-quality spectra from space. Although NASA’s Hyperion instrument acquired hyperspec- tral data in the 0.4–2:5./m region, 6 there is no equivalent to Hyperion in the thermal IR wavelengths (8–14./m). This is in Figure 1. Schematic of the thermal hyperspectral imager (THI) show- ing the calibration system, interferometer cube, and location of the cam- era and electronics within a pressure vessel. part because traditional thermal IR technology requires cooled optics that have large cost, mass, and power penalties. Our inno- vative design overcomes the traditional barriers to hyperspectral thermal IR data acquisition, with a signal-to-noise ratio of up to 1000 and a compact, low-power, low-mass configuration using uncooled microbolometer detectors. 1–3 Our sensor is based on a Sagnac interferometer (see Figures 2 and 3) and uses an FLIR R  Photon uncooled 320  256 micro- bolometer array similar to the focal planes used by NASA’s recent LCROSS lunar-impact experiment. 7 The sensor will collect calibrated hyperspectral radiance data at thermal IR wavelengths in 230m pixels with 20-wavenumber spectral res- olution from a hypothetical 400km Earth orbit. THI is designed to be compatible in terms of mass, volume, power, and software- hardware interfaces with small-satellite-platform concepts. The Continued on next page