absorption,
three in regions of 
absorption, and two in atmospheric "windows."The VAS is an advanced version of the imaging VISSR flown on early geosynchronous satellites. The VAS retains the VISSR imaging function while expanding the
number of infrared channels that allow the three-dimensional structure of the atmospheric temperature and water-vapor distribution to be determined. Of the
12 infrared channels (Table 1), seven are situated in spectral regions of absorption,
three in regions of absorption, and two in atmospheric "windows."
The visible-channel detectors consist of an array of eight photomultiplier tubes each fed by a bundle of optical fibers. The fibers are arranged so that each
tube receives radiation from one-eighth of the north-south extent of the scan line. Thus, eight adjacent scans of visible data are sampled for each revolution
of the satellite. The instantaneous geometric field of view (IGFOV) of each detector is 21 x 25 
rads. The visible-channel data are digitized to a resolution
of only 6-bits (0-63).
The infrared channels are selected by a filter wheel which lies in front of the image plane containing the infrared detectors. The filter wheel contains 12
multilayer interference filters and can be rotated to bring any one of the filters into the infrared optical path during that part of the scan when the Earth
is not being viewed. There are six infrared detectors. Two are small mer-cury cadmium telluride detectors sensitive to radiation in the ~6.5 to ~15-m wavelength
interval and each with an IGFOV of 192 x 192 rads, corresponding to ~7 km at the Earth's surface at nadir. There are two more such detectors, but doubled
in size with an IGFOV of 384 x 384 rads corresponding to a surface resolution of ~14 km at nadir.
The final pair are indium antimonide detectors with the same geometrical characteristics as the other large pair, but sensitive to radiation in the ~3.5 to
~5-m wavelength interval. The infrared detectors are chilled to 94K by a radiative cooler. The data from the infrared channels are digitized to a resolution
of 10 bits (0-1023).
The radiation arriving at the radiometer aperture is directed into the optical telescope by a plane mirror and the spacecraft spin scans the field of view
across a zone on the Earth's disk from west to east. In the imaging mode, the position of the plane mirror is rotated about an axis lying in the Earth's equatorial
plane. The mirror is rotated 192 rads between successive revolutions. This increment corresponds to the field of view of the imaging infrared detectors and
that of the array of eight visible detectors, and thus images of the Earth's disk are constructed. There are 1821 north-to-south increments of the scan mirror
required to complete a full disk image. The process takes 18.21 minutes. The visible images consist of an array of 14,568 lines, each of 15,288 pixels, and
an infrared image of 1821 lines, each of 3822 pixels.
In the sounding mode, the larger infrared detectors are used and the position of the scan mirror is not incremented between spins. Instead the same line is scanned many times by the same detector through the same filter to improve the signal-to-noise ratio of the measurement. This is repeated for different filters before the scan mirror position is incremented. This mode of operation is called "dwell sounding," and the number of scans per channel is called the spin budget (Table 2). The spin budget is dependent on the noise characteristics of the individual detectors and is different, therefore, from channel to channel and instrument to instrument, and on the improvement in the signal-to-noise ratio needed to permit the correct interpretation of the measurements in terms of atmospheric profiles. Because of the time required in the dwell sounding mode, only part of the Earth's disk is sampled between the image modes, and because of weather forecasting operational requirements, there is frequently not enough time to accrue the full spin budget (Hayden and Schmit 1991.) Gibson (1984) outlines typical operational schedules.
The infrared channels are calibrated by using an internal blackbody target. The radiation from the target is directed onto the infrared detectors by inserting a reflecting shutter into the optical path at each spin of the satellite.
The detectors measure a segment, Ri, of the electromagnetic spectrum, R(l), weighted by the transfer function, 
i(
),
of Channel i :
If the radiation is emanating from a blackbody source, R() is given by the Planck function. This is the radiation at the height of the satellite, which is
often referred to as being at the top of the atmosphere (TOA).
Montgomery and Endres (1985) and Rao et al. (1990) give more detailed instrument descriptions.
