Auroral Observations from the POLAR Ultraviolet Imager (UVI)

G. A. Germany, J. F. Spann, G. K. Parks, M. J. Brittnacher, R. Elsen, L. Chen, D. Lummerzheim, and M. H. Rees

AGU Monograph, "Encounter Between Global Observations and Models in the ISTP Era", Jim Horwitz, Dennis Gallagher, and Bill Peterson, editors, in press.

Abstract

Because of the importance of the auroral regions as a remote diagnostic of near-Earth plasma processes and magnetospheric structure, spacebased instrumentation for imaging the auroral regions have been designed and operated for the last twenty-five years. The latest generation of imagers, including those flown on the POLAR satellite, extends this quest for multispectral resolution by providing three separate imagers for the visible, ultraviolet, and X ray images of the aurora. The ability to observe extended regions allows imaging missions to significantly extend the observations available from in situ or groundbased instrumentation. The complementary nature of imaging and other observations is illustrated below using results from the GGS Ultraviolet Imager (UVI). Details of the requisite energy and intensity analysis are also presented.

INTRODUCTION

The principal science objective of UVI is to provide global information on the flow of energy between the Earth's magnetosphere and its ionosphere [Torr et al., 1995]. This is accomplished by using auroral images to estimate, on a per-pixel basis, incident energy flux and average energy of precipitating particles. The secondary objective of UVI is to provide a near-continuous monitor of the auroral morphology in the far ultraviolet for both sunlit and dark side conditions, subject only to the constraints of the orbital viewing geometry. Each of these objectives is to be achieved in the full context of the ISTP mission by using observations from the full suite of ISTP missions to provide a detailed understanding of the global nature of the integrated geospace system.

The UVI science goals directly address the GGS goals of measuring mass, momentum, and energy flow in the coupled solar wind/magnetospheric system; of studying geospace plasma processes; and of assessing their impact on the terrestrial environment [Acuna et al., 1995]. These goals, in turn, drove the design of the Ultraviolet Imager and, in part, the POLAR spacecraft. For example, the requirement to measure energy flow into the auroral regions required ultraviolet spectral resolution to a degree not previously obtained with filtered devices. The requirement that the instrument monitor the global behavior of the entire auroral region argued for a much wider field of view than typically available for such a compact camera. This requirement also called for a high apogee orbit both to allow the POLAR imagers to view extended fields of view and to allow full auroral viewing for extended periods of time. With an apogee distance of 9 earth radii, UVI is capable of viewing the entire auroral region continuously for almost 12 hours out each 18 hour orbit.

The Ultraviolet Imager is a small sophisticated camera with a wide field of view that uses a filter wheel to select one of five available far ultraviolet spectral regions for imaging. One of the major design accomplishments of UVI is the development of the high transmission, high resolution FUV filters used in the camera. Five filters are used to isolate emissions from OI 1304, OI 1356, N2 LBH (short wavelength), N2 LBH (long wavelength), and long wavelength scattered sunlight (Figure 1). These wavelength regimes were chosen to maximize the scientific return from the UVI images. Note that the OI 1304 and 1356 emissions are also resolved by two dedicated filter systems.

Extensive laboratory calibrations were conducted to determine absolute and relative photometric and spectral calibrations, as well as the pointing characteristics for each of the two detector systems. These studies have been augmented with data taken from orbit. The absolute photometric calibration allows the instrumental response to be directly related to the brightness of the observed emissions. This is the first requirement for quantitative analysis of image data. In addition, the relative response of the instrument across the image plane must also be determined so the brightness at different parts of the images may be directly compared with the rest of the image. The wide field of view of the Ultraviolet Imager made this part of the calibration especially challenging. Finally the spectral response of the camera for each filter selection was determined. This is especially critical for analysis of the LBH band system as well as for proper determination of out-of-band contamination of the 1356 emissions from the 1304 emission.

ANALYSIS TECHNIQUE

The technique for using FUV emissions as remote diagnostics has been discussed elsewhere [Strickland et al., 1983; Germany et. al., 1994a,b; 1990; Lummerzheim et al., 1991] and will only be summarized here. The principal emissions within the UVI spectral bandpass (120.0 - 200.0 nm) are atomic oxygen emissions and molecular N2 Lyman-Birge-Hopfield (LBH) emissions. Atomic nitrogen lines also appear. The only significant excitation mechanism is electron impact excitation. Thus, ignoring photoelectrons produced on the dayside, observed LBH intensities are directly proportional to the incident auroral flux provided there is no significant loss of intensity along the line of sight. Some of the LBH emissions will be absorbed by O2 along the imaged line of sight. The amount of absorption is determined by the O2 Schumann-Runge absorption continuum which peaks within the UVI bandpass. Since O2 absorption is significant only for altitudes below about 150 km, loss will only be observed for the more energetic electrons that cause emission from these altitudes. It is this altitude-dependent production and loss with mean auroral energy that allows the emissions to be used as remote diagnostics of mean energy.

The N2 LBH emissions are thus divided into two regions: one at shorter wavelengths with significant losses due to O2 absorption (denoted LBHs and extending roughly from 140 to 160 nm) and longer wavelength emissions with less loss (LBHl, 160 to 180 nm), as shown in Figure 1. Modeling of expected emissions can be used to estimate incident energy flux from LBHl intensities and the average energy from the ratio of either OI 1356 or LBHs to LBHl. (The OI 1356 emission exhibits the same absorption losses as LBHs and can therefore also be used to estimate average energy.) The 1356 emission, however, is dependent on changes in the atomic oxygen density and can therefore exhibit significant variability. On the other hand, since the emissions in the LBHs and LBHl filter passbands originate from the same specie, their ratio is nearly independent of compositional changes with season or over a solar cycle [Germany et al., 1990]. For these reasons, the LBH ratio is used almost exclusively for average energy analysis for UVI images. The following discussion is therefore limited to the analysis of the LBHs/LBHl ratio.

Since absorption by O2 is the principal physical mechanism used to estimate average auroral energy, it can be expected that there will be variation with changes in O2relative to N2, similar to the variability seen in OI 1356 with changes in O. [Germany et al., 1990] examined this possibility and found that the LBH ratio is also stable with changes in composition from a variety of sources. Molecular oxygen dependence is generally not considered significant since the O2/N2 ratio changes little at the lower altitudes of significance here. This dependence can be included, however, as a special analysis though it is not part of normal UVI analysis.

In the ideal case, UVI average energy analysis would be performed with two LBH bands - one at the wavelength of peak absorption and the other at a longer wavelength where O2 absorption is negligible. The longer wavelength emission would be essentially independent of average energy and solely dependent on energy flux. This is the case modeled previously by Germany et. al. [1994a,b; 1990]. In practice, the UVI LBH filters must necessarily include multiple LBH bands and therefore contain a range of loss factors. Thus the LBHl emission shows a weak dependence with average energy that is not indicated in the previous work by Germany et. al. [1994a,b; 1990].

Figure 2 shows modeled vertical column brightnesses for LBH emissions convolved with the UVI instrumental bandpass. Thus each curve now contains contributions for all LBH bands and not just the idealized case of only two bands modeled previously. As noted above, the LBHl emission intensity produced by a fixed electron energy flux input is not entirely independent of mean energy and varies by about 10% between 1 and 10 keV. Thus the dependence on energy flux is, in turn, dependent on a knowledge of mean energy. This is illustrated in Figure 3 where vertical brightnesses as a function of incident energy flux is presented for a range of mean energies. In addition to this dependence on mean energy there will also be dependences on diurnal and seasonal changes as well as changes in magnetic activity, principally by changes in the neutral atmospheric composition. These variations were examined previously by Germany et. al. [1990] for the idealized cases of only two LBH bands. Table 1 shows sample variations for the case of the entire LBH band system convolved with the UVI instrumental response. The largest variation is the seasonal variation of about 6% which is slightly larger than the 3-4% reported by Germany et. al. [1990]. Since each of these effects is relatively small they are not included in the energy analysis shown below. Instead a fixed value for the brightness to energy flux conversion is used.

Two of the requisite three parameters, incident energy flux and average energy, are provided from the intensity of the LBHl images and from the ratio of LBHl and LBHs (or 1356) images. Germany et al. [1994b] studied the possibility that FUV auroral images could be used to specify the incident auroral energy distributions, i.e. the third parameter necessary to completely specify the energy of the incident auroral flux. They found that the sensitivity of this type of analysis is not enough to discriminate between different energy distributions. Nevertheless, the two derivable diagnostics - energy flux and average energy - are excellent probes of the observed auroral processes.

Since UVI images include auroral and airglow emissions, the images must have both instrumental and airglow backgrounds removed before analyzing. Airglow is approximated from the image by assuming pixels with similar values of solar zenith angle will have similar airglow emissions. The underlying airglow surface is estimated by binning image pixels by solar zenith angle, excluding auroral emissions.

Additional corrections are also made for brightness variations with changing spacecraft view angle (the angle from the local zenith to the spacecraft). These are illustrated in Figure 4 which shows the increased slant path brightness (relative to the vertical brightness) for a range of spacecraft angles. For the LBHl images the enhancement is nearly the cosine enhancement that would be expected from purely geometric concerns. The departure from pure cosine behavior is due to O2 absorption from some of the shorter wavelength LBH bands which are observed by the LBHl filter. These shorter bands are partially absorbed by O2, introducing an additional, nongeometric, factor into the line of sight analysis. The enhanced absorption along the slant path is competing with the increased emission resulting in a brightness less than expected from the line of sight geometry. This is a small effect for the LBHl filter since the absorbed bands are only a small fraction of the total observed bands. For LBHs, however, this effect has significant consequences as seen in the figure. Not only is the slant path enhancement countered by the increased absorption along the view path, but the absorption is a function of the neutral composition of O2 which can vary significantly across the UVI field of view due to local time effects. The result is a surface dependent on both solar zenith angle and spacecraft look angle. Note that this behavior is significant only for higher energy LBHs auroral emissions. LBHs observed dayglow, originating from higher in the thermosphere, will not be as strongly affected by O2 absorption and will have behavior more similar to the LBHl data shown in Figure 4.

For average energy calculations, the image data must be registered onto a regular magnetic latitude-MLT grid before image ratios can be calculated. This ensures the pixels used in the image ratios correspond to the same location even though the two images are not temporally coincident. The fact that the two ratioed images are not coincident can lead to analysis problems in cases where the auroral morphology is significantly changing between the two observations. In a companion paper, Germany et al. [1997] analyze a dynamic auroral event showing how energy characteristics can be extracted under such conditions.

For the work presented here typical uncertainties can be estimated. Image processing (binning, smoothing, ratioing, etc.) introduces about 3% uncertainty and the Poisson uncertainty is about 5%, dependent on the brightness of the observed features. Assuming an accuracy of 25% for the instrumental calibration, 22% for the LBH cross sections [Ajello and Shemansky, 1985], and 25% for all other model uncertainties provides a total uncertainty of about 45%.

There is also a well-known wobble in the POLAR despun platform that affects the imagers on that platform. The wobble is seen in UVI images as a vertical smearing of about 0.5 degree, or about 10 pixels or so. (There is no distortion in the transverse direction.) The effect is clearly seen for fixed bright sources such as stars but is not always present in auroral displays. We interpret this as evidence that the observed auroral emissions are changing over the course of the image integration and don't result in a smeared signature.

OBSERVATIONS AND RESULTS

In the discussion below UVI images are used to illustrate how global imaging can be used to augment in situ and ground based observations, including the capability of observing sunlit aurora. This is then followed with two case studies based on UVI images. The first study is an examination of energy influx during the onset of the magnetic storm of January 10, 1997. The second study is a coordinated ISTP study involving multiple spacecraft and groundbased observations.

One of the principal uses of spacebased imagery is to provide a larger context for measurements that are often limited in temporal and spatial extent. For example, in Figure 5, taken from Doe et al. [1997], a one hour time history of incident auroral flux is shown corresponding to the location of the Sondrestrom radar. Also shown are energy flux values derived from four radar measurements during the same period. (The mean and maximum curves are used to investigate the relative influence of the binned 0.5 x 1 degree bins.) The principal feature to note from the figure is that the energy flux data derived from the image data shows much more temporal detail than is available from the ground based observations.

Another example of this is given in Figure 6 from Lummerzheim et al. [1997] which compares hemispheric power input derived from UVI images with the same parameters derived from energetic electron and ion precipitation measured during passes of the NOAA/TIROS satellites over the aurora [Evans, 1987; Fuller-Rowell and Evans, 1987]. The power derived from the satellites is obtained by fitting the in situ observations to a statistical database derived from the entire auroral oval. Possible shortcomings of such an approach are seen in the figure at 19:20 and 22:45 UT where there are significant disagreements between UVI and NOAA/TIROS power estimates. Examination of UVI data for these periods reveals that the satellite pass at these times sampled locally brighter areas of the oval, thus leading to an overestimate of the total power.

Both of these examples examine total hemispheric power input and necessarily ignore smaller scale structures in the auroral zone. As discussed above, however, UVI images can be used to estimate the magnitude of the incident energy flux over the entire auroral zone on a per pixel basis. Plate 1 shows a map of incident energy flux derived from a UVI LBHl image from May 19, 1996. The UVI image data have been rebinned and registered in a mlt-magnetic latitude coordinate system using apex coordinates. Superimposed on the energy map is the ground track of a DMSP satellite as it passed through the region. The inferred energy fluxes generally agree in magnitude and morphology with selected DMSP overflights (Figure 7). This is a more challenging comparison due to differences in spatial and temporal resolution between UVI and DMSP as well as the fact that this is a dynamic event in which the poleward boundary was changing rapidly on a time scale of less than a minute. The fluxes inferred from UVI images do not exhibit the same spatial or temporal resolution as in situ measurements but offer a global perspective unavailable from single satellite passes. This event is studied in more detail by Germany et al., [1997].

In summary, when spacebased imagery is available from a high altitude the information content provided is significantly greater than can be supplied by either in situ or groundbased observations alone. However, as seen in Figure 7, in situ observations are often available with a resolution that exceeds that of the imager. (UVI spatial resolution varies from 4 km at perigee to 40 km at apogee.) hence the two observation techniques must be viewed as complementary.

A significant consequence of the ultraviolet wavelength discrimination capabilities of UVI is its ability to monitor dayside aurora. When coupled with the extended field of view of the Ultraviolet Imager this allows simultaneous observations of auroral activations throughout the auroral zone on both dark and sunlit conditions. Plate 2 illustrates this with two views of northern auroral activity observed on April 9, 1996 at 14:35 and 14:45 UT.

The dayside is to the right in both images. The auroral oval straddles the FUV terminator with dayside intensifications seen in both images. The circular field of view of UVI is seen outlined on the map. The period from 13:45 to 14:00 UT, prior to the images in Plate 2, represents the recovery phase of a minor substorm on the predawn nightside over Russia and Alaska. By 14:00 UT the nightside intensifications had faded away. However, throughout this period the dayside auroral zone showed continuous activity covering local time from prenoon to almost dusk. From 14:00 to 14:30 UT the auroral activity is confined totally to the dayside. The situation at 14:34 UT is shown in the first image of Plate 2. At 14:39 UT a substorm began on the nightside, roughly premidnight. Over the next 40 minutes this activity will expand to encompass the full nightside oval. Observations over the 1.5 hours monitored here clearly show that on this day the dayside auroral activity is much less variable than is the nightside activity, a fact also noted by Brittnacher et al., [1997]. In that study the total power deposition was estimated in the dayside and nightside halves of the auroral region (divided along the 0600 to 1800 MLT boundary) for two quiet days in spring (April 7, 1996) and summer (July 23, 1996). The results are reproduced here in Figure 8 showing that the power deposited on the dayside aurora varied significantly less than did that on the nightside.

In the image shown in Plate 2 the dayside and nightside activities are essentially decoupled with continuous dayside activity coincident with distinct quiet and substorm periods on the nightside. In other cases, however, the two regions appear to be coupled with simultaneous night and dayside activations or with one type of activity leading to the other. Brittnacher et al., [1996], for example, observed dayside precipitation that subsequently propagated eastward toward local midnight. This particular event is especially interesting in that it accompanied an unusually strong dipolarization as observed by ISTP spacecraft yet did not proceed to a classical substorm development.

One principal advantage of having a continuously operating camera is the ability to observe unexpected or rapidly changing auroral morphologies. An example of this capability occurred during the preparation of this manuscript with the magnetospheric response of the solar coronal mass ejection event observed at Earth on January 10, 1997. A shock front was observed around 01:00 UT with a magnetic cloud following around 03:40 UT. UVI images from the period show extremely distorted ovals with multiple, distorted transpolar arcs evident (Plate 3). The principal storm is typically identified as beginning near 03:40 UT. However, analysis of power deposition derived from UVI images (Figure 9) for the 3 hours preceding this time clearly shows significant activity beginning before 01:10 UT. The surface in Figure 9 is shifted so that local midnight occurs in the center of the axis with dawn to the right and dusk to the left. Near 01:10 UT there is significant enhancement in both the dawn and dusk sectors. The dawn enhancement persists in some form until the main storm onset. At 1:30 UT there is a very localized, intense brightening at midnight. This brightening then expands into the poleward arcs shown in Plate 3. Near 02:50 UT there is a brightening throughout most of the dusk sector that expands into the dawn sector beginning near 03:15 UT. Storm onset is indicated at 03:35 with an initially localized intense brightening at midnight that quickly broadens into the full storm expansion by 04:00 UT..

Figure 10 shows the instantaneous power input and the total energy deposition during this period. It shows the initial energy onset near 01:00 UT, the precursor buildup from 03:00 to 03:30 UT followed by the main storm onset between 03:30 and 04:00 UT.

The hallmark of the ISTP program is the ability to simultaneously observe disparate parts of geospace and analyze the data from this unique perspective. One of the first opportunities for this type of coordinated analysis, using UVI images, occurred early in the operational lifetime of the POLAR spacecraft on March 27, 1996. This corresponded to a perigee pass of the WIND spacecraft which placed both WIND and GEOTAIL in the earth's magnetotail. The two spacecraft were in the dusk flank near the Z=0 plane. Both WIND and GEOTAIL were within 700 km of each other in the Z direction. At the same time IMP8 and INTERBALL-TAIL were upstream of the bow shock sampling the solar wind and the interplanetary magnetic field. POLAR was operational with auroral imaging available. Added to these platforms were NOAA, DMSP, and GOES spacecraft as well as ground observations. Taken all together this has served as an excellent opportunity to characterize substorm events from multiple perspectives [e.g. Angelopoulos et al, 1997; 1996; Germany et al., 1996].

Ground magnetometer data from the CANOPUS Magnetometer and Riometer Array (MARIA) [Rostoker et. al., 1995] indicate auroral activity near 10:00 UT and near 13:00 UT over central and western Canada. Smaller events near 06:00 and 12:00 UT are also indicated. UVI image data shown in Plate 4 illustrates an auroral substorm beginning at 09:45 UT over northern Alaska with the auroral arc extending over the active magnetometer stations which had indicated the substorm activity. Airglow is seen in the images in the bottom right hand portion of the image over Greenland and northern Europe. Some dayside auroral activity is also seen near the dusk terminator. The temporal development of the 09:45 UT substorm is shown in Plate 5 as a mosaic of six images from 09:44 to 10:34 UT. (The image times for each image, from top left to bottom right, are 09:44, 09:52, 10:06, 10:18, 10:24, and 10:34 UT.) The substorm begins with an intensification near 22:00 magnetic local time. This is then followed by a classic expansion eastward and poleward until near 10:34 UT when the auroral is in recovery leaving a boundary arc on the poleward boundary of the expansion. The substorm advances westward in a series of discrete intensifications with each new brightening taking place westward of the previous one. The edge of the UVI field of view is seen near 06:00 LT

IMF and magnetometer data are shown in Figure 11 on a common plot with auroral activity shown qualitatively by horizontal bars. The bars indicate the beginning and duration of a specific auroral intensification. Where necessary, the horizontal bars are displaced vertically to avoid overlap. Some auroral activity is clearly correlated with the magnetometer data. However, at 11:00 UT UVI images reveal auroral intensifications that do not clearly correspond to enhanced magnetic activity in the magnetometer data underneath the aurora. Conversely, at 12:00 UT the magnetometer data indicate magnetic activity that does not correlate with auroral activity in the images, at least not with the intensity of the other activations. Correlating auroral activity with the IMF is more difficult because of the processes in the intervening magnetosphere and the associated time delay. The IMF data indicate general activity through the period but it is difficult to correlate individual features to specific auroral events in the images without detailed modeling.

This type of multi-instrument coordinated analysis has been one of the principal modes of study since the POLAR launch with UVI images being used to provide a reference for multiple magnetospheric and groundbased measurements [e.g. Brittnacher et al., 1996; Elsen et al., 1996; Horwitz et al., 1996; Nagai et. al., 1996].

FUTURE DIRECTIONS

Each of the examples discussed above were basically event driven. That is, a narrow period of time was selected and then the image data was processed with significant interactive support. In particular, the location of the oval is critical for much of the analysis and must either be estimated based on statistical principals or be specified in an interactive fashion. This significantly slows down the image analysis effort. Furthermore, it restricts analysis to event driven studies since there is no way to automatically search available image data for specific events.

To address this problem, UVI researchers are developing techniques to automatically identify the oval location in image data. An existing neural network algorithm with proven utility in medical and defense image analysis applications is being adapted for use with image data from the Ultraviolet Imager. The goal is to use the algorithm not only to locate the auroral oval but also to enable automated recognition of auroral forms in UVI image data. Once developed for use with images from the Ultraviolet Imager the algorithm can be used with image data from multiple auroral imagers. The purpose of this study is to develop methods to process large volumes of auroral image data from past, current, and future image programs.

ACKNOWLEDGEMENTS

We are grateful to G. Rostoker for providing CANOPUS magnetometer data and for several helpful comments. The CANOPUS instrument array was constructed and is maintained and operated by the Canadian Space Agency for the Canadian scientific community. A. Skalsky kindly supplied the INTERBALL magnetometer data from the MIF instrument (S. Romanov PI). We are also grateful to R. Lepping for sharing IMP8 magnetometer data. This work was supported, in part, under U. Washington contract 256730 to the University of Alabama in Hunstville. The authors gratefully acknowledge the helpful comments of the reviewers.

REFERENCES

Acuna, M. H., K. W. Ogilivie, D. N. Baker, S. A. Curtis, D. H. Fairfield, and W. H. Mish, The Global Geospace Science program and its investigations, Space Sci. Rev. 71, 5, 1995.

Angelopoulos, B., T. D. Phan, D. E. Larson, F. S. Mozer, R. P. Lin, K. Tsuruda, H. Hayakawa, T. Mukai, S. Kokubun, T. Yamamoto, d. J. Williams, R. W. McEntire, R. P. Lepping, G. K. Parks, M. Brittnacher, G. Germany, J. Spann, H. J. Singer, and K. Yumoto, Magnetotail flow bursts: association to global magnetospheric circulation, relationship to ionospheric activity and direct evidence for localization, Geophys. Res. Lett., 24, 2271, 1997.

Angelopoulos, V., T. Phan, D. Larson, F. S. Mozer, R. P. Lin, M. Moyer, T. Mukai, K. Tsuruda, R. W. McEntire, D. J. Williams, G. Parks, R. Lepping, K. Yumoto, and E. Friis-Christensen, Correlative IMP8-WIND-GEOTAIL-POLAR measurements of magnetotail substorms, EOS Trans. AGU, 77(46), Fall Meet. Suppl., F609, 1996.

Brittnacher, M. J., G. K. Parks, L. J. Chen, R. Elsen, J. Spann, and G. Germany, Dipolarization of the plasma sheet observed without a classical substorm injection, EOS Trans. AGU, 77(46), Fall Meet. Suppl., F633, 1996.

Brittnacher, M., R. Elsen, G. Parks, L. Chen, G. Germany, and J. Spann, A dayside auroral energy deposition case study using the Polar Ultraviolet Imager, Geophys. Res. Lett., 24, 991, 1997.

Doe, R. A., J. D. Kelly, D. Lummerzheim, G. K. Parks, M. J. Brittnacher, G. Germany, and J. Spann, Initial comparison of POLAR UVI and Sondrestrom IS radar estimates for auroral electron energy flux, Geophys. Rev. Lett., 24, 999, 1997.

Elsen, R. K., G. K. Parks, M. J. Brittnacher, L. J. Chen, S. M. Petrinec, R. M. Skoug, R. M. Winglee, G. A. Germany, J. F. Spann, Mapping UVI images into the magnetosphere with empirical and global MHD magnetospheric models, EOS Trans. AGU, 77(46), Fall Meet. Suppl., F633, 1996.

Germany, G. A, G. K. Parks, M. J. Brittnacher, J. F. Spann, J. Cumnock, D. Lummerzheim, F. Rich, and P. G. Richards, Global Auroral Remote Sensing Using GGS UVI Images, AGU Monograph "Encounter Between Global Observations and Models in the ISTP Era", Jim Horwitz, Dennis Gallagher, and Bill Peterson, ed., 1997.

Germany, G. A., G. K. Parks, M. Brittnacher, L. Chen, R. Elsen, J. F. Spann, J. Cumnock, P. G. Richards, and F. Rich, Characterization of an auroral intensification using multiple spacecraft observations, EOS Trans. AGU, 77(46), Fall Meet. Suppl., F623, 1996.

Germany, G. A., M. R. Torr, D. G. Torr, and P. G. Richards, Use of FUV auroral emissions as diagnostic indicators, J. Geophys. Res., 99, 383, 1994a.

Germany, G. A., D. G. Torr, P. G. Richards, M. R. Torr, and S. John, Determination of ionospheric conductivities from FUV auroral emissions, J. Geophys. Res., 99, 23297, 1994b.

Germany, G. A., M. R. Torr, P. G. Richards, and D. G. Torr, The dependence of modeled OI 1356 and N₂ LBH auroral emissions on the neutral atmosphere, J. Geophys. Res., 95, 7725, 1990.

Hirahara, M., J. L. Horwitz, T. E. Moore, J. F. Spann, G. Germany, W. Peterson, E. Shelley, M. Chandler, G. Giles, P. Craven, C. Pollock, J. Scudder, D. Gurnett, J. Pickett, A. Persoon, N. Maynard, F. Mozer, M. Brittnacher, and T. Nagai, Correlation of ionospheric ion outflows with auroras, waves, and convection observed with the POLAR satellite: Case studies during active and quiet intervals, in press, J. Geophys. Res., 1997.

Hirahara, M., J. L. Horwitz, G. Germany, T. E. Moore, J. F. Spann, M. O. Chandler, and B. L. Giles, Properties of upflowing ionospheric ion conics and magnetosheath proton precipitation at 5000 km altitude over cusp/cleft auroral forms: Initial observations from the TIDE and UVI instruments on POLAR, Huntsville 96 Workshop 'Encounter Between Global Observations and Models in the ISTP Era' (Guntersville, AL) September 15-20, 1996.

Horwitz, J. L., G. Kunin, M. Hirahara, G. Germany, D. G. Brown, J. F. Spann, T. Nagai, and R. Lepping, Auroral and magnetospheric activity during intervals of steady southward interplanetary magnetic field from ISTP POLAR and WIND observations, EOS Trans. AGU, 77(46), Fall Meet. Suppl., F613, November 1996.

Lummerzheim, D., M. Brittnacher, D. Evans, G. A. Germany, G. K. Parks, M. H. Rees, and J. F. Spann, High time resolution study of the hemispheric energy flux carried by energetic electrons into the ionosphere during the May 19/20 auroral activity, Geophys. Rev. Lett., 24, 987, 1997.

Lummerzheim, D., M. H. Rees, J. D. Craven, and L. A. Frank, Ionospheric conductances derived from DE-1 auroral images, J. Atmos. Terr. Phys., 53, 281, 1991.

Moore, T. E., C. R. Chappell, M. O. Chandler, S. A. Fields, C. J. Pollock, D. L. Reasoner, D. T. Young, J. L. Burch, N. Eaker, J. H. Waite, Jr., D. J. McComas, J. E. Nordholdt, M. F. Thomsen, J. J. Berthelier, and R. Robson, The Thermal Ion Dynamics Experiment and Plasma Source Instrument, Space Sci. Rev., 71, 409, 1995.

Nagai, T., Y. Saito, T. Yamamoto, T. Mukai, A. Nishida, S. Kokubun, M. Hirahara, G. Germany, J. L. Horwitz, J. Spann, and M. Brittnacher, GEOTAIL-POLAR simultaneous observations of substorm onsets, EOS Trans. AGU, 77(46), Fall Meet. Suppl., F631, 1996.

Richards, P. G., and D. G. Torr, Auroral modeling of the 3371 A emission rate: Dependence on characteristic electron energy, J. Geophys. Res., 95, 10337, 1990.

Rostoker, G., H.C. Samson, F. Creutzberg, T.J. Hughes, D.R. McDiarmid, A.G. McNamara, A. Vallance Jones, D.D. Wallis and L.L. Cogger, CANOPUS - a ground-based instrument array for remote sensing the high latitude ionosphere during the ISTP/GGS prgram, Space. Sci. Rev., 71 , 743 , 1995.

Strickland, D. J., J. R. Jasperse, and J. A. Whalen, Dependence of auroral FUV emissions on the incident electron spectrum and neutral atmosphere, J. Geophys. Res., 88, 8051, 1983.

Torr, M. R., D. G. Torr, M. Zukic, R. B. Johnson, J. Ajello, P. Banks, K. Clark, K. Cole, C. Keffer, G. Parks, B. Tsurutani, and J. Spann, A far ultraviolet imager for the International Solar-Terrestrial physics mission, Space Sci. Rev., 71, 329, 1995.

FIGURE CAPTIONS

Figure 1. UVI spectral sensitivity as a function of wavelength. The curves are shown on a relative scale and the diagnostic solar filter centered longward of 180 nm is not shown.

Figure 2. Modeled vertical column brightnesses for LBH emissions convolved with the UVI instrumental bandpass. The convolution is over all bands of the LBH band system. Emissions are modeled for a Gaussian energy distribution with an incident energy flux of 1 mW/m2.

Figure 3. Modeled vertical column brightnesses using the UVI LBHl filter for Gaussian energy distributions with mean energies of 0.5, 1, 5, and 10 keV.

Figure 4. Slant path brightness enhancements for a range of spacecraft angles, where the spacecraft look angle is the angle between the local zenith at point of emission and the direction to the spacecraft. (Look angle equals zero degrees when viewing directly down toward the center of the Earth and equals 90 degrees when viewing the limb.) Brightness is shown relative the nadir direction (look angle equal zero degrees). Both the LBHl and LBHs emissions are modeled for auroral emissions with a mean energy of 10 keV. Each modeled point represents an observation corresponding to a single pixel of the UVI field of view. For clarity, only a fraction of all points are shown on the plot. The line labeled 'Cosine' is the enhancement expected from purely geometric considerations.

Figure 5. Time history of energy flux derived from UVI and Sondrestrom observations. The line labeled 'maximum' corresponds to the brightest pixel within a bin size of 0.5 (lat) x 1 (lon) degree; the line labeled 'mean' is the mean of all pixels within the binned area. From Doe et al. [1997].

Figure 6. Total hemispheric energy flux derived from UVI images on May 19/20, 1997. Thick horizontal bars show the hemispheric power derived from NOAA/TIROS satellites. From Lummerzheim et al. [1997].

Figure 7. Comparisons between incident energy flux determined from coincident DMSP measurements and UVI image analysis. In situ observations of incident energy flux are shown as solid lines. Boxed crosses represent values from the UVI energy map corresponding to the DMSP ground track. The vertical line corresponds to the beginning of the UVI LBHl integration of 37 seconds. Dashed lines represent an estimated 45% uncertainty level.

Figure 8. Dayside and nightside incident power for (a) April 7, 1996 and (b) July 23, 1996. The diamonds show the times at which the images were acquired. Of note is that the nightside precipitation shows much more variation than does the dayside precipitation. From Brittnacher et al. [1997].

Figure 9. Incident power as a function of local time at the onset of the magnetic storm on January 10, 1997. The local time scale has been shifted so midnight is at the center of the axis.

Figure 10. Total deposition for the onset of the magnetic storm on January 10, 1997. The bottom plot shows the total power for a given image frame in Watts. The top plot shows the time integration of this quantity in Joules.

Figure 11. Interplanetary magnetic field and ground magnetometer data with auroral activity shown qualitatively by horizontal bars. The bars indicate the beginning and duration of a specific auroral intensification. Where necessary, the horizontal bars are displaced vertically to avoid overlap.

Plate 1. Map of incident energy flux derived from a UVI LBHl image. The UVI image data has been rebinned and registered in a mlt-magnetic latitude coordinate system using apex coordinates. The image is from 21:44 UT May 19, 1996 and is coincident with a DMSP F12 overflight as shown by the colored track near 21:00 MLT.

Plate 2. Observations of auroral activations throughout the auroral zone on both dark and sunlit conditions on April 9, 1996 at 14:34 and 14:44 UT. The dayside is to the right in both images. The dayside and nightside activities are essentially decoupled with continuous dayside activity coincident with distinct quiet and substorm periods on the nightside. The scale is in units of photons cm-2 s-1.

Plate 3. Multiple distorted transpolar arcs seen in response to the solar coronal mass ejection event observed at Earth on January 10, 1997. The image color table has been modified to emphasize the structure of these faint arcs. Peak intensity is about 10 photons cm-2 s-1 for both images.

Plate 4. An auroral substorm beginning at 09:45 UT as seen by UVI using the LBHl filter. The substorm is located over northern Alaska with the auroral arc extending over CANOPUS magnetometer stations which indicated substorm activity. Airglow is seen in the images in the bottom right hand portion of the image over Greenland and northern Europe.

Plate 5. Temporal development of the 09:45 UT substorm as a mosaic of six images from 09:44 to 10:34 UT. The image times for each image, from top left to bottom right, are 09:44, 09:52, 10:06, 10:18, 10:24, and 10:34 UT. Each image is independently scaled in brightness to emphasize the relative morphological changes with time.

TABLES

Table 1. Sample variations of the LBH band system convolved with the UVI instrumental response. Each of the test cases represents a perturbation from the reference case.