Introduction

The first global images of Earth's aurora were acquired by three scanning photometers, one each for emissions from N₂⁺ 391.4 nm, OI 557.7 nm, and OI 630.0 nm, on board the low-altitude spacecraft ISIS 2 [Anger et al., 1973; Shepherd et al., 1973]. Although these images were obtained over the Northern Hemisphere only once per orbit because orbital motion provided one of the dimensions for the pixel array of an image, the potential value of global images for organizing and interpreting in situ observations of particles and fields was clearly demonstrated. Indeed the ability to obtain global images provided considerable excitement in the scientific community from the visualization of the overall morphology of the auroral emissions [Anger et al., 1974]. The ISIS2 observations were soon followed by the availability of broadband images from similarly scanning radiometers, 400--1130 nm, from the Defense Meteorological Satellite Program (DMSP) spacecraft [Rogers et al., 1974]. Because the ISIS2 and DMSP satellites orbited at low-altitudes, in the range of 1000 km, typically only a portion of the auroral oval was viewed during a single polar pass. These spacecraft were later followed by the launch of Kyokko into an orbit with apogee of about 4000 km. Kyokko was equipped with an imagememory tube and obtained global auroral images at farultraviolet wavelengths, in this case within a broad passband at about 120 to 140 nm, with unprecedented temporal resolution [Kaneda et al., 1977]. The frame repetition period was 2 minutes. In the 1980s auroral imagers were launched with several spacecraft. These spacecraft were Dynamics Explorer 1 [Frank et al., 1981], HILAT [Meng and Huffman, 1984], Viking [Anger et al., 1987], Polar BEAR [Meng et al., 1987], and Akebono [Kaneda and Yamamoto, 1991]. HILAT and Polar BEAR were launched into low-altitude orbits and thus yielded an auroral image once per orbital period, about 100 minutes. However, their imaging equipment included spectrometers with sufficient resolution to separate the OI 130.4 nm and 135.6 nm emissions. Such spectroscopy is important in quantitative evaluation of the spectrum of precipitating electrons [Strickland et al., 1983]. The scanning spectrometer on Polar BEAR also provided images of the nighttime auroral oval at visible wavelengths. The apogee of the Viking orbit was located at higher altitudes, about 13,500 km. The two cameras, one for the OI emissions at 130.4 and 135.6 nm with passband extending into the longer wavelengths of N₂⁺ Lyman-Birge-Hopfield (LBH) emissions and the other for the LBH emissions, were capable of frame repetition periods of 20 s. These Viking images were very useful in studies of the temporal evolution of the spatial distribution of features at auroral and polar cap latitudes [Murphree et al., 1987].

The three spin-scan photometers on Dynamics Explorer 1 (DE-1) provided hundreds of thousands of global auroral images during the nine years following their initial operational turn-on in fall of 1981 [Frank and Craven, 1988]. Two of these imagers viewed the aurora at visible wavelengths in narrow passbands as selected with a filter wheel. The third imager was equipped with broadband filters for farultraviolet wavelengths in the range of 120 to 175 nm. The orbit was sufficiently high, an altitude of about 22,000 km, that the viewing time of the entire auroral oval during a single orbit was 2 to 3 hours. Such viewing times allowed continuous global viewing of the development of entire auroral substorms, i.e., through growth, onset, expansion and recovery phases. The imagers for visible wavelengths are the first, and at present only, optical systems that have successfully viewed the dim auroral emissions in the nighttime atmosphere with the intense emissions from sunlit Earth in the field-of-view. These imagers were usually operated in a mode such that an image frame with a field-of-view 30° × 30° divided into 14,400 pixels was acquired once each 720 s.

The Visible Imaging Instrumentation (VIS) for the Polar spacecraft is designed to achieve high-time and -spatial resolution images of the nighttime polar and auroral emissions at visible wavelengths. There is an ancillary camera for far-ultraviolet wavelengths within a broad passband, 124--149 nm. This camera can provide full images of Earth from radial distances and is used to verify the proper pointing of a two-axis targeting mirror for the two primary cameras for visible wavelengths. The optics for the visible cameras is based upon the off-axis catoptric design with super-polished surfaces that was successfully used for the DE-1 spin-scan imagers. Because the VIS is mounted on a despun platform and can stare at Earth its performance in terms of angular resolution and frame rate can be greatly improved relative to that for the serial single-pixel sampling on the rotating DE-1. For example, consider the viewing of the nighttime auroral zone from a Polar spacecraft altitude of 7.4 R_e and high latitude. The targeting mirror for the cameras for visible wavelengths can be used to position the fields-of-view of these cameras such that viewing of the auroras is optimized. For the low-resolution camera its field-of-view is 5.6° × 6.3° and sufficient to usually include the entire nighttime auroral oval. Within this field-of-view a frame of 65,500 pixels can be telemetered every 12 s. For a DE-1 image taken at this altitude the corresponding frame of 576 pixels could be telemetered every 144 s. The counts/pixel for a given auroral brightness are similar for the two images.

The scientific objectives for observations with VIS can be grouped into five primary categories: (1) quantitative assessment of the dissipation of magnetospheric energy into the auroral and polar ionospheres, (2) an instantaneous reference system for the in situ measurements with the ISTP spacecraft, (3) development of a substantial model for energy flow within the magnetosphere, (4) investigation of the topology of the magnetosphere, and (5) delineation of the responses of the magnetosphere to substorms and variable solar wind conditions. It should be realized that these general objectives cannot be achieved without the in situ observations from the various ISTP spacecraft. Much has been learned from previous studies as to the specific investigations that will contribute to these general objectives. Because this paper is devoted to a description of the instrumentation we limit our discussion of the objectives to an illustrative example for each category.

In order to achieve (1) above inter-leaved sequences of images of the emissions from N₂⁺ at 391.4 nm and for OI at 630.0 nm are acquired. The 391.4nm emission is a good measure of the electron energy flux into the atmosphere and the ratio of the two intensities is a measure of the electron energy spectral index. The complication for determination of the energy fluxes and electron spectra in this manner is caused by the reflectance of Earth's surface and, if present, clouds. This is basically a tractable radiative transfer problem [Rees et al., 1988]. The reflectance of Earth and clouds and their contributions to the observed intensities are to be evaluated in part with images at filter wavelengths that are offset from auroral emission lines. The determination of electron energy spectra with the visible emission lines is complementary to that achieved at farultraviolet wavelengths and can be applied for lower electron energies than the latter measurement [Rees et al., 1988; Strickland et al., 1983]. Of course, the visible observations can be only taken for the nighttime aurora where most of the precipitating charged particle energy fluxes occur whereas the farultraviolet measurements are possible for the sunlit atmosphere.

The provision of an instantaneous coordinate system, category (2), is obvious. The global auroral images place the in situ observations in the context of auroral substorm phase or other activity and of geographical location of the imprint of charged particle precipitation. For example, measurements of particles and fields in the distant polar magnetosphere during periods that a theta aurora [Frank et al., 1986] is observed can resolve the controversy as to whether the transpolar arc of this auroral configuration is the footprint of bifurcation of the magnetospheric lobes [Frank, 1988] or large-scale spatial distortion of the plasma regimes in the magnetotail [Akasofu and Roederer, 1984; Lyons, 1985].

Category (3) studies of the gross flow of energy within the magnetosphere extend over a broad range, including the inference of the total magnetic energy in the magnetotail from the area poleward of the auroral oval and the relative motions of the ion and electron plasmas in the vicinity of the inner edge of the electron plasma sheet. In order to obtain the footprint of protons precipitating into the ionosphere the VIS is equipped with a narrow-band filter for HI 656.3 nm emissions. With interlaced images of OI 557.7 nm emissions the large-scale inter-relationship between precipitation of electrons and protons from the near-Earth plasma sheet into the atmosphere can be studied.

The mapping of plasma boundaries into features of the auroral luminosities is important for extending in situ observations of these boundaries with a single spacecraft into a visualization of their geometries and temporal evolutions. Such studies are included in category (4). It is clear that such identification of these boundaries can contribute significantly to our knowledge of the magnetic field topology of the magnetosphere. Only limited studies of this type have been reported. One of the notable examples is the identification of poleward discrete arcs in the auroral oval with the plasma sheet boundary layer from simultaneous observations with the DE-1 imager and an ISEE2 plasma analyzer [Frank and Craven, 1988] and with magnetometers on both spacecraft [Elphic et al., 1988]. The comparison of images from the Polar spacecraft and in situ fields and particles observations with both the Polar and Geotail spacecraft should substantially increase our knowledge of the magnetic topology of the magnetosphere and its relationship to major plasma regions.

Analysis of DE-1 image sequences for small, isolated substorms has revealed that the polar cap area, i.e., that area enclosed by the poleward edge of the auroral oval, responds to the southward turning of the interplanetary magnetic field [Frank and Craven, 1988; Frank, 1988]. Although previous studies with low-altitude observations indicated that this response occurs [Meng and Makita, 1986] the DE-1 images provided determination of the entire polar cap boundary with sufficient temporal resolution, 12 minutes, to clearly identify this effect. These studies are part of the general topic (5) above. The DE-1 results showed that the polar cap area expands when the interplanetary field turns southward and increases until a substorm onset occurs. During the expansion phase the polar cap area decreases. The expansion and subsequent contraction of polar cap area can be interpreted as the storage and release, respectively, of the total magnetic energy in the magnetotail lobes. This energy can be quantitatively estimated with simple models of the magnetotail magnetic fields [Coroniti and Kennel, 1972]. The question remains as to the precise connection between the polar cap area and open magnetic field lines in the lobes. Detailed analysis of auroral images and simultaneous fields and particles measurements with the Polar spacecraft should resolve this issue and refine the estimates of the transport of solar wind energy into the magnetotail and its explosive release during substorms.

The opportunity to construct and launch a state-of-the-art camera does not occur very often. With the addition of a few filters into the instrumentation the objectives can address several targets of opportunity. The two filters at 317.3 and 360.1 nm provide high spatial resolution for the total columnar ozone in Earth's sunlit atmosphere. One of these filters can be used to acquire global monitoring of the occurrence of lightning. A narrowband filter at 589.0 nm is also included for surveys of the Moon's Na cloud [Mendillo et al., 1991] and Na emissions in Earth's nighttime atmosphere. The filter for OH emissions at 308.5 nm can be used to pursue the topic of atmospheric holes by a search for clouds of OH above and in the upper atmosphere [Frank and Sigwarth, 1993].