ing just over 1/15th of the spectral range 1 ^m to 5.2 ^m (in first order), and two Si detectors covering the submicron range 0.7 ^m to 1 ^m (in second order). The spectral resolution was 0.0125 ^m below 1 ^m, and 0.025 ^m above. Angular resolution was 0.5 x 0.5mrad, corresponding to approximately 500 x 500 km at perijove.
A spectrum was constructed by scanning the grating over a small angular range such that 17 subspectra were recorded by the individual detectors, which were then overlapped and combined. In addition to the grating scan, the secondary mirror of the telescope could also be scanned over 20 contiguous positions in the cross-dispersion direction. Hence, by scanning the grating for each mirror position, a single line of a spectral image could be constructed with 20 individual spectra of 408 wavelengths (24 grating steps x 17 detectors). To record an image, the whole instrument was simultaneously scanned by the remote-sensing platform in the grating dispersion direction, as is shown in Figure 7.40, building up an image with 20 rows (secondary mirror scan) x a variable number of columns (remote-sensing platform scan) x up to 408 wavelengths (grating scan). The format of the NIMS data were thus referred to as "cubes" composed of a number of two-dimensional images recorded at multiple wavelengths. In practice, the instrument could be operated with a variable number of grating steps and to reduce data volume, not all detectors were read during a particular observation. The instrument could thus very flexibly trade off between spatial and spectral coverage. This was particularly fortunate given Galileo's communications difficulties and meant that NIMS could adapt to best utilize whatever observation time/data storage was available.
Was this article helpful?