next up previous contents index
Next: High Spatial Resolution Up: Planetary Science and Previous: Hubble Space Telescope

Science with the Hubble Space Telescope -- II
Book Editors: P. Benvenuti, F. D. Macchetto, and E. J. Schreier
Electronic Editor: H. Payne

Characterization and Numerical Simulation of the Jovian UV Aurora Observed with the HST Cameras

D. Grodent, V. Dols and J.-C. Gérard
Laboratoire de Physique Atmosphérique et Planétaire, Institut d'Astrophysique, Université de Liège, Belgium

G.R. Gladstone
Southwest Research Institute, San Antonio, TX 78228-0510



A numerical model simulating a synthetic aurora observed from Earth orbit is able to reproduce the main features of the UV auroral images such as the limb brightening, the contribution from the back side of the planet and the signature of polar cap faint emissions. We describe the principles of this numerical model and we illustrate the results with a) basic geometry cases, b) a real image as observed by the WFPC-2 camera in the 1150--1600Å bandpass.

Keywords: Jupiter, aurora, ultraviolet


Our knowledge of the UV Jovian auroral morphology has increased tremendously since the FOC (Gérard et al. 1993, 1994a,b) and the later WFPC-2 cameras (Clarke et al. 1994) provided the first images of the H ultraviolet aurora. It is now possible to use the existing database to characterize the distribution and energetics of the emission generated by the interaction of auroral particles with the H-H Jovian atmosphere (Gérard et al. 1994b). The geometry of observation of the Jovian aurora from Earth orbit is complicated by 1) the inclination of Jupiter's magnetic field with respect to its rotation axis, 2) Jupiter's obliquity, 3) the flattening of Jupiter, 4) the fact that the auroral emission zone has a vertical distribution which is a complex function of altitude, and 5) the non-trivial geometry of the auroral emission zones. An intuitive approach fails to account for all these elements and leads to erroneous morphological interpretations. Instead, a numerical simulation model including each of these elements is a reliable and powerful tool to analyze the observed data.

Description of the Auroral Simulation Method

The purpose of these simulations is to visualize an arbitrary emission distribution on the planet as would be seen from Earth for various possible viewing geometries. The footprint of the auroral region on the Jovian surface can be roughly approximated by an oval which is characterized by the planetocentric longitude and latitude of its center, the length of the semi-major and semi-minor axes and the latitudinal width (Gaussian distribution across the oval). Once the sub-Earth latitude, the sub-Earth longitude and the planetary flattening have been set, a model Jovian disk image seen from Earth orbit is generated.

Figure: Principle of the numerical simulation method. The planetary disk is shown from the side with one particular line of sight parallel to the plane of the figure. A filled auroral oval is drawn with Chapman profiles for several latitudes.

By each point (x,y coordinates) of this image, a line of sight is traced perpendicular to the plane of the image (z coordinate), providing a 3-D coverage (x,y,z) of the auroral region. As illustrated in figure 1, a portion of a line of sight may intercept some auroral emission. In this case, the emission is sampled and summed along the line of sight (giving rise to the limb brightening effect). Also shown in figure 1 is the Chapman profile which is used as an approximation of the H auroral brightness distribution along the vertical. The Chapman profile is characterized by the H atmospheric scale height and by the altitude of the maximum of the profile.

Basic Geometry Cases

We have simulated a simple Northern Jovian auroral oval in three different configurations. In each case, the sub-Earth latitude is 3 S and the central meridian longitude is 180, the H atmospheric scale height is 100 km and the maximum of the Chapman profile is set to 400 km above the ammonia clouds top (1 bar level). This peak altitude is consistent with the conclusion based on IUE UV spectral observations that the peak of the auroral emission lies close to the homopause (Livengood et al. 1990). The parameters of the auroral oval are such that this reference oval fits the auroral emission observed in several WFPC-2 images taken in July 1994 in the 1150--1600Å bandpass.

Figure: Simulation of different cases of auroral basic geometry. A) simple oval 1 wide, B) inside of the oval only, C) both oval + inside of the oval. The sub-Earth latitude is so the jovigraphic north pole is hidden behind the planet. The scale height is set to 100 km and the altitude of the maximum of the Chapman profile is set to 400 km above the NH ice clouds (grid). The meridians are separated by 20 degrees and the parallels by 10 degrees. The central meridian longitude is 180.

First Configuration (Figure 2A)

A simple oval 1 wide is considered. Because of the observing geometry, the maxima of limb brightening appear close to the dawn and the dusk limbs, well inside the planetary disk. Indeed, only a fraction of the auroral emission originating from behind the planet exceeds the minimum altitude required to stand above the planetary limb. Actually, the location of the two maxima correspond to the two parts of the oval inside the planetary disk, where the tangent to the oval is nearly parallel to the lines of sight, that is where the auroral zones intercepted by the lines of sight are the most extended. The brightness ratio between the Earth facing section of the oval on the central meridian longitude (CML) and the maximum brightness is on the order of 5.

Second Configuration (Figure 2B)

The inside of the auroral oval (excluding the oval itself) is uniformly filled in with auroral emission. In this case the maximum limb brightening effect appears close to the night limb. This stems from the fact that nearly all the auroral emission inside the planetary disk contributes to the limb brightening. Even though the initial brightness of the inside emission is set three times lower than the initial oval brightness, this effect exceeds the maximum limb brightening observed in the previous case by one order of magnitude. In this simulation, the brightness ratio between the dimmest and the brightest auroral region is on the order of 20.

Third Configuration (Figure 2C)

Both the inside and the auroral oval are considered. As expected from the sum of the two previous cases, the maximum of limb brightening appears at the night limb and is mainly due to the contribution of the inside of the oval. The brightness ratio between the Earth facing section of the oval on the CML and the brightest region is on the order of 8.

Real Image

In July 1994 (during the Jupiter/comet Shoemaker-Levy 9 encounter period), a set of 22 images of the UV Jovian aurora was obtained by the WFPC-2 cameras using the F160WB and F160WB+F130LP filter set (Clarke et al. 1994). We describe one particular image (Figure 3B) acquired in the F160WB filter set on July 20. The exposure time was 400 s and the CML was 187 (system III longitude) which gives the best view of the Northern aurora. This image clearly illustrates the complexity of the auroral morphology. It cannot be described by a single auroral oval as previously discussed. Instead we have to define a set of two kinds of auroral modules : a) partial ovals, all derived from the reference oval already used in section 3 but having different semi-axis lengths, and b) uniformly filled extended zones following the reference oval and restricted in longitude.

Figure: Simulation of a UV image obtained by the WFPC-2 camera. A) simulated image (same parameters as those used in figure 2), B) observed image (the single red spots are due to cosmic rays). The meridians are separated by 20 degrees and the parallels by 10 degrees. The central meridian longitude is 190.

The WFPC-2 image shown in figure 3B is satisfactorily simulated (along with several other images from the set) using three partial ovals and one uniformly filled zone (figure 3A). Figure 4 shows an equidistant polar projection in system III (S) coordinates of this set of auroral modules. The partial oval ranging from S longitude 215 to 90 (green arc) simulates the bright dawn limb ansa and the faint emission at the night limb (as described in point 3.2.), the two other arcs (red and yellow) follow smaller ovals and are well inside the planetary disk, the smallest one ranges from S longitudes 170 to 185 and the largest ranges from S longitudes 155 to 175, a brightness variation along these two modules is needed in order to simulate the corresponding bright arcs in the WFPC-2 image. This intrinsic brightness variation is clearly independent of the geometrical effects. The uniformly filled zone (blue) is restricted by the reference oval and by S longitudes of 90 and 175(at the level of the reference oval). It gives rise to 1) the brightness enhancement appearing at the late evening limb, 2) the inside emission around the two bright arcs (red and yellow), 3) the low latitude faint emission (Grodent et al. 1996) appearing in the afternoon region.

Figure: Illustration of the auroral modules used in the simulation of the real case. The parallels are separated by 10. System III longitudes increase by 20 degrees steps clockwise from the reference meridian (0) which is oriented toward the top. The 180 meridian is toward the bottom. The black square indicates the CML (190).


By using a set of auroral modules, the numerical simulation model is able to reproduce the very complex Jovian auroral structures observed with the WFPC-2 cameras. It allows to discriminate the geometrical effects from the intrinsic brightness variations for an Earth orbit observation complicated by the viewing geometry. The application of the model to the images from the WFPC-2 database may reveal auroral structures that statistically appear to be fixed in local time or intrinsically linked to the planet. It will provide a better understanding of the origin of the energetic particles and their acceleration processes in the Jovian magnetosphere.


This study is based on observations with the NASA-ESA HST that were obtained at the Space Telescope Science Institute, which is operated by AURA, Inc., for NASA under contract NAS5-26555. JCG acknowledges support from the Belgian National Fund for Scientific Research (FNRS). Funding for this research was provided by the PRODEX program of the Belgian Federal Office for Scientific, Technical and Cultural Affairs.


Clarke, J.T., R. Prangé, G.E. Ballester, J. Trauger, D. Rego, R. Evans, K. Stapelfeldt, W. Ip, F. Paresce, J.C. Gérard, H. Hammel, M. Ballav, L. Ben Jaffel, J.L. Bertaux, D. Crisp, C. Emerich, W. Harris, M. Horanyi, S. Miller, A. Storrs & H. Weaver 1995, Science, 267, 1302

Gérard, J.C., V. Dols, F. Paresce, & R. Prangé 1993, J. Geophys. Res., 98, 18793

Gérard, J.C., V. Dols, R. Prangé, & F. Paresce 1994a, Planet. Space Sci., 42, 905

Gérard, J.C., D. Grodent, R. Prangé, J.H. Waite, G.R. Gladstone, V. Dols, F. Paresce, A. Storrs, L. Ben Jaffel, & K.A. Franke 1994b, Science, 266, 1675

Grodent, D., V. Dols, & J.C. Gérard 1996, J. Geophys. Res., in press.

Livengood, T.A., Strobel D.F., & H.W. Moos 1990, J. Geophys. Res., 95 ,10375

next up previous contents index
Next: High Spatial Resolution Up: Planetary Science and Previous: Hubble Space Telescope