Magnetospheric mapping of the dayside UV auroral oval at Saturn using simultaneous HST images, Cassini IMF data, and a global magnetic field model

We determine the field-aligned mapping of Saturn’s auroras into the magnetosphere by combining UV images of the southern dayside oval obtained by the Hubble Space Telescope (HST) with a global model of the magnetospheric magnetic field. The model is tailored to simulate prevailing conditions in the interplanetary medium, corresponding to high solar wind dynamic pressure and variable interplanetary magnetic field (IMF) strength and direction determined from suitably lagged field data observed just upstream of Saturn’s dayside bow shock by the Cassini spacecraft. Two out of four images obtained in February 2008 when such simultaneous data are available are examined in detail, exemplifying conditions for northward and southward IMF. The model field structure in the outer magnetosphere and tail is found to be very different in these cases. Nevertheless, the dayside UV oval is found to have a consistent location relative to the field structure in each case. The poleward boundary of the oval is located close to the open-closed field boundary and thus maps to the vicinity of the magnetopause, consistent with previous results. The equatorward boundary of the oval then maps typically near the outer boundary of the equatorial ring current appropriate to the compressed conditions prevailing. Similar results are also found for related images from the January 2004 HST data set. These new results thus show that the mapped dayside UV oval typically spans the outer magnetosphere between the outer part of the ring current and the magnetopause. It does not encompass the region of primary corotation flow breakdown within the inner Enceladus torus.


Introduction
This paper is concerned with the structure of the magnetic field in the outer regions of Saturn's magnetosphere, specifically the mapping of field lines from the high-latitude ionosphere to the equatorial regions and magnetopause, and what one can learn from this about the mapping and origins of Saturn's polar auroras.Previous work using the paraboloid model of Saturn's magnetosphere (Alexeev et al., 2006;Belenkaya et al., 2006a), also employed here, has shown that this field structure is complex, and depends on the strength and direction of the interplanetary magnetic field (IMF) (Belenkaya et al., 2006b(Belenkaya et al., , 2007(Belenkaya et al., , 2008)).The field also depends on the solar wind dynamic pressure which changes the overall size of the magnetosphere via pressure balance at the magnetopause boundary (Kanani et al., 2010).This also modulates the strength of the ring current, such that the field in the middle magnetosphere is quasi-dipolar in form when the magnetosphere is compressed but extends into a magnetodisc when it is expanded (Alexeev et al., 2006;Bunce et al., 2007Bunce et al., , 2008b;;Arridge et al., 2008).
As discussed recently e.g. by Stallard et al. (2007a, b) and Kurth et al. (2009), the physical processes which result in the generation of Saturn's auroras are not yet well established.With increasing equatorial distance from the planet, field-aligned currents and auroras could firstly result from corotation breakdown in the Enceladus torus through plasma production and transport processes analogous to those at Jupiter discussed by Hill (1979) and Pontius and Hill (1982).At larger distances in the outer magnetosphere Sittler et al. (2006) have discussed the role of centrifugally-driven instabilities, while Cowley et al. (2004aCowley et al. ( , b, 2008) ) have proposed that corotation breakdown near the open-closed field line boundary plays an essential role.We now briefly discuss each of these in turn.
Published by Copernicus Publications on behalf of the European Geosciences Union.1234 E. S. Belenkaya et al.: Magnetospheric mapping of the dayside UV auroral oval at Saturn Firstly, field-aligned currents and auroras could be associated with the corotation breakdown in the inner and middle magnetosphere beginning at radial distances of ∼3-4 R S (Wilson et al., 2009), which is associated with water-group ion production and outward transport in the Enceladus torus (Saur et al., 2004;Pontius and Hill, 2009).According to this picture, the plasma flow should return to near-corotation immediately equatorward of the main auroral oval.However, Stallard et al. (2007b) have presented a detailed analysis of profiles of the infrared (IR) H + 3 emission intensity and Doppler velocity across Saturn's polar auroral region, and show that this signature is not observed in their data.These authors thus conclude that this process is not responsible for the generation of the main auroral oval at Saturn.However, they identify IR emissions at lower latitudes than the main oval which may be associated with the corotation enforcement currents in the inner and middle magnetosphere, which have been further studied by Stallard et al. (2010).
Secondly, Sittler et al. (2006) have presented a model of Saturn's global auroral response to the solar wind as observed in simultaneous Hubble Space Telescope (HST) auroral images and Cassini upstream measurements of the solar wind and IMF obtained in January 2004, emphasizing that Saturn's magnetosphere is a fast rotator.They argue that the torques on Saturn's outer magnetosphere are relatively low, such that its outer magnetosphere will tend to conserve angular momentum.When compressed on the dayside the outer magnetosphere will then spin up, and when it expands it will spin down.These authors consider the outer boundary of the plasma sheet at L ∼ 15 (where L is the equatorial radial distance in planetary radii) to be the primary source of precipitating auroral particles.They suggest that radial transport is dominated by centrifugally-driven flux tube interchange motions, such that when the magnetosphere is compressed and spins up, outward transport will increase, and the precipitating particles will move radially outward.This mechanism will thus cause the auroral oval to move to higher latitudes as observed.
Thirdly, based on Voyager plasma observations on closed field lines in the near-equatorial magnetosphere and IR Doppler observations on open field lines in the polar cap, Cowley et al. (2004a) show that a ring of upward-directed currents should flow in the vicinity of the boundary between open and closed field lines that is of sufficient intensity to require significant acceleration of magnetospheric electrons, resulting in ultra violet (UV) auroral emissions of a few tens of kilo-Rayleighs (kR).Stallard et al. (2007b) note that the characteristic signature of this model is a significant velocity shear in the region of the main oval, corresponding to the open-closed field line boundary, in which the plasma on the equatorward side should be closer to rigid corotation than that on the poleward side.Stallard et al. (2007b) suggested that the fact that this signature is not observed in their data may be explained by their spatial resolution, which did not enable them to resolve a possible narrow band of enhanced rotation equatorward of the main auroral oval.Initial evidence in favour of this model has been presented by Bunce et al. (2008a), who analysed near-simultaneous observations of Saturn's southern auroras using HST UV images and concurrent Cassini measurements of auroral field-aligned currents in the polar magnetosphere.The Cassini data provide evidence of strong upward currents flowing in the open-closed boundary region at noon, co-located with the main oval observed in the HST images.Belenkaya et al. (2006bBelenkaya et al. ( , 2007Belenkaya et al. ( , 2008Belenkaya et al. ( , 2010) ) have studied the relationship between Saturn's aurora and the IMFdependent field structure of the outer magnetosphere using the paraboloid field model, specifically examining the correspondence between the auroral oval and the boundary between open and closed field lines.Such studies require the simultaneous availability of both UV auroral images obtained by the HST and IMF data upstream from Saturn obtained by Cassini, of which only two such joint campaign data sets exist at present.Belenkaya et al. (2006bBelenkaya et al. ( , 2007Belenkaya et al. ( , 2008) ) studied the joint data obtained during Cassini approach to the planet in January 2004 when the spacecraft lay at radial distances of ∼1300 R S , such that the spacecraft-planet IMF propagation time was ∼20 h with uncertainties of several hours.(R S is Saturn's 1 bar equatorial radius equal to 60 268 km.) Subsequently, Belenkaya et al. (2010) studied HST auroral images obtained on an approximately daily basis during a twoweek period in February 2008 (see Clarke et al., 2009, for further campaign details).For much of this interval Cassini was located inside Saturn's magnetosphere, but it emerged into the solar wind near apoapsis on the dayside of the planet during DOY (day of year) 43 to 46, such that the upstream IMF could then be monitored with much reduced propagation delay and uncertainty.Belenkaya et al. (2010) used the paraboloid magnetosphere model, including the fields of the ring current, tail current, and magnetopause current, together with a partially-penetrated IMF, to calculate the boundary of the open field line region in the southern ionosphere, it being noted that only the southern aurora could be observed during both campaign intervals due to the pre-equinox conditions prevailing (vernal equinox occurred at Saturn in August 2009).Overall, the results from both the 2004 and 2008 data intervals show a close relationship, with the openclosed boundary lying close to the poleward boundary of the auroras.
In this paper we further analyse and discuss the modelling results obtained using the joint HST-Cassini data from February 2008.Specifically, we elucidate the structure of the magnetospheric magnetic field pertaining to the auroral imaging intervals using the paraboloid model, thus allowing us to understand the connection between the southern polar ionosphere and domains within the equatorial magnetosphere and/or magnetopause.We then map the model field lines from the poleward and equatorward boundaries of the observed dayside UV oval into the magnetosphere in order to understand the physical domains concerned, thus helping to Ann.Geophys., 29, 1233Geophys., 29, -1246Geophys., 29, , 2011 www  From top to bottom the data panels show the three components of the magnetic field in KSM coordinates, and the field magnitude.In the colour-coded region identifier at the top of the plot green corresponds to the magnetosphere, red to the magnetosheath, and blue to the solar wind.Principal bow shock crossings (red-blue transitions) are identified by vertical black dashed lines.The data at the bottom of the plot give the radial distance (R S ), latitude (deg), and local time (h) of the spacecraft.The 1 h blue vertical stripes labelled "A" to "D" correspond to the suitably lagged times of four HST imaging intervals.(From Belenkaya et al., 2010).
clarify the relationship between the observed emissions and the suggested auroral mechanisms discussed above.We also briefly compare our results with those derived from selected related images from the January 2004 data set.

Cassini and HST observations during DOY 43-46 2008
In this section we first discuss the Cassini data and HST images which form the basis of the analysis presented here, taken from the results of Belenkaya et al. (2010).In Fig. 1 we show magnetic field data obtained by the Cassini fluxgate magnetometer (Dougherty et al., 2004) for DOY 43 to 46 of 2008.From top to bottom we show the three components of the magnetic field in kronocentric solar-magnetospheric (KSM) coordinates, together with the field magnitude.In the KSM system X points towards the Sun (approximately anti-parallel to the solar wind flow), the X-Z plane contains the planet's magnetic (and spin) axis, and Y completes the right-hand orthogonal triad pointing towards dusk.Spacecraft position information is given at the bottom of the fig-  Belenkaya et al. (2010) estimated that for this interval of data, the configuration of the ionospheric field and flow responds to solar wind conditions with an overall delay of about 6 h.This, together with the ∼1 h light propagation time to Earth, was thus adopted to determine the Cassini solar wind interval corresponding to particular HST images.Four such images were obtained during the interval, labelled "A" to "D", each consisting of the sum of nine consecutive 100 s images of Saturn's southern oval obtained using  the F115LP long-pass filter, combined together to increase signal-to-noise.The blue vertical stripes in Fig. 1 thus show the solar wind times corresponding to these images, labelled A to D at the top of the figure, each stripe showing the 1 h interval over which the IMF data were averaged in order to obtain a representative value.It can be seen in particular that the north-south field component B Z was negative (southward) during images A and B, but was positive (northward) during images C and D. In this paper we focus on images A and C as representative of these two conditions.
The images are presented in Fig. 2, projected onto a southern latitude and longitude grid shown by the yellow dotted lines at intervals of 10 • each.The view is "through" the planet from the north with noon at the bottom of each plot and dawn to the left.Because the sub-Earth latitude of Saturn was −8 • at the time of these observations, only the dayside portion of the oval was well-observed, such that the images are truncated somewhat beyond the dawn-dusk meridian.The red crosses in Fig. 2 indicate the poleward and equatorward boundaries of the auroral emission, used below to map the auroral boundaries along field lines into the magnetosphere.These were determined by averaging the emission over 10 • longitude bins to increase the signal to noise, and applying a simple emission threshold of 5 kR.This value was set sufficiently high that the boundary locations are not greatly affected by noise in the data, and sufficiently low that significant auroral emission is not omitted.We also eliminate a few points whose co-latitude lies more than two standard deviations from the mean in each image, which occur in regions of weak emission where a clear visual boundary cannot be identified.Even so, the poleward boundary of the auroras is notably variable in these images, while the equatorward boundary is relatively unvarying.We note that no corrections have been made to the boundaries for artificial poleward stretching due to the finite height of the auroral curtain, since this effect is small compared with the significant difference in morphology between the images.It can be seen that the UV oval in image A is brightest in the dawn-to-noon sector, extending between ∼6 • and ∼18 • southern co-latitude, while being more uniformly distributed in image C, extending between ∼10 • and ∼16 • .These values are comparable with the co-latitudes of the boundaries of the southern UV dayside oval determined from a wider set of UV images by Badman et al. (2006), who found that the median poleward boundary lies at 14 • and the median equatorward boundary at 16 • , both within considerable variability.We thus note that these auroral distributions are entirely representative of a much wider body of such data (e.g., Clarke et al., 2009).
The data at the bottom of each image in Fig. 2 give the corresponding rounded averaged IMF vector determined from the Cassini data in Fig. 1.As we go on in the next section to describe, the (unrounded) averaged IMF vectors have been employed in the paraboloid model of Saturn's magnetosphere to calculate the position of the open-closed field line boundary, shown by the orange lines in each image.Those 1237 shown here represent a higher-resolution refinement of those previously depicted by Belenkaya et al. (2010).

Paraboloid model calculations
In the paraboloid model (Alexeev et al., 2006;Belenkaya et al., 2006aBelenkaya et al., , 2007Belenkaya et al., , 2008Belenkaya et al., , 2010;;Alexeev, 2010), the magnetopause is taken to be a paraboloid of revolution about the Saturn-Sun line, such that the model is expressed in KSM coordinates.The main contributors to the model magnetic field are (i) the intrinsic magnetic (dipole) field of the planet, together with the shielding magnetopause current which confines the dipole field inside the boundary, (ii) the ring current and the corresponding shielding magnetopause current, (iii) the tail currents and their closure currents on the magnetopause, and (iv) the IMF which partially penetrates into the magnetosphere.The parameters which define the model magnetic field are then as follows.(i) is the tilt angle between the magnetic dipole axis and the KSM Z-axis (−8.4 • during DOY 43-46 2008), (ii) R SS is the distance from Saturn's centre to the subsolar point on the magnetopause, (iii) R rc1 and R rc2 are the distances to the outer and inner edges of the ring current, respectively, (iv) B rc1 is the radial component of the ring current field at the outer edge of the ring current, (v) R 2 is the distance from the planet's centre to the inner edge of the magnetospheric tail current sheet, and (vi) the field magnitude of the tail currents at the inner edge of the tail current sheet is B t /α 0 , where α 0 = (1 + 2R 2 /R ss ) 1/2 .For simplicity the current sheets in the model are taken to be of zero thickness, compared with observed thicknesses typically of a few R S (e.g., Kellett et al., 2009), which should have little effect on the field line mapping between ionosphere and magnetosphere.Further discussion of the paraboloid model current systems and magnetic fields may be found, e.g., in Alexeev and Belenkaya (2005).The effect of the IMF inside the magnetosphere is given by adding the uniform field k S B IMF , where B IMF is the IMF vector and 0 ≤ k S ≤ 1 is the magnetosphere penetration coefficient.We note that the model is steadystate and thus does not include representations of the global magnetic field perturbations that rotate around the planet near the planetary period (e.g., Andrews et al., 2010), related to corresponding small-amplitude "wobbles" of the auroral oval (e.g., Nichols et al., 2010), nor of the transient plasmoids possibly associated with the Vasyliunas-cycle observed in Saturn's nightside plasma sheet (e.g., Jackman et al., 2008).The model also does not include representation of the quasisteady magnetosphere-ionosphere currents that couple angular momentum between these two regions (e.g., Hill, 1979;Cowley et al., 2004a), that generally lead to "lagging" field lines in the magnetosphere in the presence of sub-corotation.No detailed empirically-based models of such fields have yet been derived for Saturn due to the dominant presence of the above "planetary period" oscillations, however, simple esti-mates indicate modest LT displacements that are largely directed around the auroral oval, thus not significantly affecting the radial mapping issue.
In the field calculations shown here we keep the basic model parameters fixed as in Belenkaya et al. (2010), so that we can clearly discern the effect of the variable IMF.Since according to the results presented by Clarke et al. (2009) the interval studied corresponds overall to one in which Saturn's magnetosphere was quite strongly compressed by the solar wind, with the dynamic pressure peaking at ∼0.1 nPa on DOY 44 and 45, here we choose to use a relatively compressed model throughout.Specifically we employ the model derived by Belenkaya et al. (2006a) from fits to the magnetic field observed during the Pioneer-11 flyby, which corresponds to a dynamic pressure of ∼0.08 nPa.This model was also successfully employed by Belenkaya et al. (2007) to describe Saturn's magnetosphere under high solar wind dynamic pressure conditions.The corresponding set of model parameters is R ss = 17.5 R S , R rc1 = 12.5 R S , R rc2 = 6.5 R S , B rc1 = 3.62 nT, R 2 = 14 R S , and B t = 8.7 nT.We note that the values for the ring current boundaries are in reasonable accord with the Cassini field modelling results of Bunce et al. (2007) corresponding to similar compressed magnetosphere conditions.
The value of the IMF penetration coefficient k S is not yet well known for Saturn, such that Belenkaya et al. (2010) presented results for both low (k S = 0.2) and high (k S = 0.8) values.While the structure of the outer magnetosphere and the length of the magnetospheric tail is found to be strongly affected by this choice (see their Fig.4), the size and position of the open field region projected to the ionosphere is found to be changed only to a relatively modest degree.We also note that determinations of the IMF penetration parameter at Earth, obtained both directly using magnetospheric magnetic field data and from estimates of the "efficiency of reconnection" factor based on radar convection measurements, favour the lower value (e.g., Cowley and Hughes, 1983;Tsurutani et al., 1984;Milan et al., 2004).Here we therefore employ k S = 0.2.We start by discussing the simpler situation that occurs for northward IMF corresponding to case C.

Image C during northward IMF conditions
Image C in Fig. 2 was obtained by the HST at ∼07:00 UT on DOY 45, with a corresponding lagged KSM IMF vector of (−0.11, 0.28, 0.25) nT such that the Z-component was northward-directed in this case (Belenkaya et al., 2010).To gain a better understanding of the magnetospheric connection to the southern polar region during this interval, we first examine the model magnetic field lines whose ionospheric footprints lie at fixed southern latitudes, at steps of 1 h of LT.In Fig. 3  with the principal meridians (noon, dusk, etc.) being shown by the green lines.The field line intersections are confined to the Southern Hemisphere, and also favour the dawn side of the tail due to the positive IMF Y component conditions prevailing.
We now turn specifically to the mapping of the boundaries of the southern auroras, as defined by the red crosses in the auroral image in Fig. 2. The corresponding field lines are shown in Fig. 5 in a similar format to Fig. 3, noting, however, that the ionospheric footprints are confined mainly to the dayside sector due to the limited viewing of the auroral oval by the HST.In panel (a) we show the field lines mapping from the equatorward boundary of the dayside southern oval (blue lines), consisting wholly of closed field lines in conformity with Fig. 2.These map to equatorial distances of ∼11.5 R S in the dawn to post-noon sector, corresponding to latitudes ∼ −73 • in the southern ionosphere as can be seen in Fig. 2, moving inwards to ∼9 R S in the dusk sector as the boundary moves equatorward to ∼ −70 • .Panel (b) similarly shows field lines mapping from the poleward boundary of the dayside oval, some of which are closed (purple lines) while others are open (black lines), as also seen in Fig. 2. The closed lines map typically to radial distances of ∼16 R S near to the dayside magnetopause, while the open field lines pass into the magnetic tail in the Southern Hemisphere.
For better clarity, in Fig. 6 we show the mappings of the closed dayside oval boundary field lines in the KSM equatorial plane, together with principal features of the magnetic model.The latter features are indicated by the black solid lines, where the outer curve shows the paraboloid magnetopause, the two inner circles centred on the planet (black dot) show the inner and outer boundaries of the equatorial ring current, while the black curve across the tail on the The mapped field lines corresponding to the equatorward and poleward boundaries of the auroras are then marked by blue and purple squares in Fig. 6, respectively, all of which correspond to closed field lines since the open field lines of the poleward boundary do not intersect the equatorial plane.It can be seen that, overall, the dayside auroral oval maps from close to the dayside magnetopause at ∼17.5 R S for these compressed conditions (i. this sector.We note that these results are in excellent accord with the findings of Talboys et al. (2009) concerning the location of the dayside upward-directed magnetosphereionosphere coupling field-aligned currents at Saturn, specifically in the southern pre-noon sector in that case.They found that the region of upward current, which may be associated with downward electron acceleration and UV auroras, maps to the closed field region between the boundary of open field lines and the outer edge of the ring current.We also note that the plasma in this outer region is dominated by hot (several keV and above) tenuous (∼0.1 cm −3 ) ions and electrons, rather than the cool (∼100 eV ions and few eV electrons) dense (up to ∼100 cm −3 ) water plasma originating from Enceladus that is characteristic of the inner region (Schippers et al., 2008;Sergis et al., 2009;Kellett et al., 2010Kellett et al., , 2011)).
As a test of the generality of these features, we have compared them with results from one of the images in the 2004 data set previously studied by Belenkaya et al. (2007Belenkaya et al. ( , 2008)), obtained under related interplanetary conditions.Specifi-cally, we have considered image "m" in this data set, obtained on 30 January 2004 when the IMF mapped from Cassini was also northward-directed, but with a somewhat larger total magnitude of ∼1 nT compared with ∼0.4 nT for image C. The compression of the magnetosphere was somewhat reduced, however, corresponding to a more usual solar wind dynamic pressure of ∼0.03 nPa.Differing paraboloid model parameters have thus been employed appropriate to this case (see Belenkaya et al., 2008, for details), in particular a larger subsolar magnetopause radius of 22 R S .Belenkaya et al. (2008) again showed that the poleward boundary of the oval in this case maps close to the open-closed field boundary, while here we have newly computed the equatorial mapping of the equatorward boundary of the oval.As in Fig. 6 for image C, we find that this boundary maps close to the outer edge of the model ring current in the equatorial plane, now located at a radial distance of ∼15 R S in this somewhat more expanded case.These results are therefore entirely compatible with those found above for image C.
As noted previously by Belenkaya et al. (2007Belenkaya et al. ( , 2008Belenkaya et al. ( , 2010)), the result that the poleward boundary of the UV oval lies close to the boundary of open field lines is in agreement with the model calculations of Cowley and Bunce (2003) and Cowley et al. (2004aCowley et al. ( , b, 2008)), that associates the oval with upward field-aligned currents generated by a shear in azimuthal flow between open and closed field lines.Sittler et al. (2006), on the other hand, consider the outer boundary of the plasma sheet (at L ∼ 15 in their model and at 12.5 in ours) to be the primary source of auroral precipitating particles.Our results show that generally the equatorward auroral boundary maps close to the outer edge of the ring current, such that this mechanism could also contribute to the observed emissions.Wilson et al. (2009) show that the primary corotation breakdown at Saturn occurs in the Enceladus torus between 3 and 4 R S , mapping to about −62 • in the southern ionosphere.This breakdown thus occurs well equatorward of the UV oval, as originally suggested by Cowley and Bunce (2003).However, corotation-enforcement currents will certainly flow in this vicinity, as discussed in the jovian context by Hill (1979), Pontius and Hill (1982) and Vasyliunas (1983).Stallard et al. (2008) reported the discovery of a lower-latitude auroral oval in IR emission at Saturn which they called a "secondary oval", lying equatorward of the main UV oval investigated here.They showed that this low-latitude emission occurs close to the point where corotation breaks down in the IR Doppler data, thus mapping to the Enceladus torus region well equatorward of the main oval (Stallard et al., 2010).Thus it seems likely that auroral emission formed by current systems caused by rigid corotation breakdown in the inner and middle magnetosphere may be realized in the lower-latitude auroral oval in IR emission, but not in the main oval.

Image A during southward IMF conditions
We now provide a similar discussion of image A in Fig. 2, obtained by the HST at ∼22:00 UT on DOY 43 with a corresponding lagged KSM IMF vector of (0.20, −0.85, −0.24) nT, such that the Z-component was southward-directed in this case, opposite to case C (Belenkaya et al., 2010).Model magnetic field lines with fixed latitudes in the southern ionosphere are shown in Fig. 7    The equatorial intersections of these boundary field lines are depicted in Fig. 10 in a similar format to Fig. 6, where panel (a) uses the same spatial scales as Fig. 6 showing the near-planet region, while panel (b) shows a larger region extending down the magnetospheric tail.The closed field region is contained between the red and green lines, the former being projected from the southern ionosphere and the latter from the northern, which now extends to much greater distances down the tail than for case C.These lines are connected to each other at intersections of the magnetic separator line with the equatorial plane.(For a discussion of the "open" field geometries occurring for northward-and southwarddirected IMF in these cases, corresponding to southward-and northward-directed IMF at Earth due to the differing planetary dipole directions, the reader is referred to the works of Alexeev and Belenkaya (1983) and Blomberg et al. (2005), and references therein.)The blue squares again show the equatorial intersections of the equatorward boundary of the dayside oval, which as for case C is somewhat variably located between the outer edge and centre of the ring current region, ∼8.5-12.5 R S .In this case, however, it is located near the outer edge of the ring current in the dawn and postnoon sector, and near the centre in the pre-noon sector.With regard to the poleward boundary, only one of the model field lines intersects the equator, shown by the purple square close to the magnetopause in the post-dawn sector.1244 E. S. Belenkaya et al.: Magnetospheric mapping of the dayside UV auroral oval at Saturn that the auroral distributions observed in these images are entirely representative of a much wider body of such data (e.g., Clarke et al., 2009).
Calculations of the field structure were thus undertaken using the paraboloid magnetosphere model appropriate to the conditions prevailing during these HST imaging intervals, corresponding in the February 2008 data to an interval of significant magnetospheric compression by the solar wind (Clarke et al., 2009;Belenkaya et al., 2010), with a penetrating field computed from suitably lagged and averaged Cassini IMF data.The auroral boundaries were then mapped along the model field lines in each case from the ionosphere to the magnetosphere, with the following principal results.
(a) Although the model field structure in the outer magnetosphere and tail is very different in the two cases, as expected from the differing IMF directions prevailing, the mapped dayside UV auroral oval is found nevertheless to have a consistent location relative to the field, spanning the outer magnetosphere between the outer part of the ring current and the open-closed field boundary at the magnetopause.Corresponding results have also been found for selected images within the January 2004 data set obtained under related interplanetary conditions, thus supporting the generality of these findings.The results are in excellent accord with the findings of Talboys et al. (2009), who studied Cassini observations of field-aligned current signatures in high-latitude dayside magnetic field data, and found that the region of upward currents spanned the outer magnetosphere from near the outer boundary of the ring current to near the boundary of open field lines.Upward currents are potentially associated with downward accelerated electrons and bright aurora.This outer magnetospheric region is that dominated by hot tenuous plasma, rather than by cool dense plasma characteristic of the inner region (Schippers et al., 2008;Sergis et al., 2009;Kellett et al., 2010Kellett et al., , 2011)).
(b) In more detail, the equatorward boundary of the dayside oval in images A and C was found to map typically to distances of ∼12 R S in the equatorial plane, near the outer boundary of the model ring current appropriate to the compressed magnetospheric conditions prevailing.However, this boundary was also found to map somewhat further inward to near the centre of the ring current at ∼9 R S in some restricted sectors, near dusk in image C, and in the pre-noon sector in image A. Under the more typical less compressed conditions of image "m" in the January 2004 data set, the equatorward boundary was found to map consistently to a radial distance of ∼15 R S in the equatorial plane, again close to the outer boundary of the ring current.The poleward boundary of the dayside oval, on the other hand, was found typically to straddle the boundary between open and closed field lines, in conformity with the previous conclusions of Belenkaya et al. (2006bBelenkaya et al. ( , 2007Belenkaya et al. ( , 2008Belenkaya et al. ( , 2010)).Oval boundary points on the closed field side thus map close to the dayside magnetopause.
(c) These results have implications with regard to previous theoretical considerations concerning the origins and mapping of the auroras.Specifically we note that the association of the equatorward boundary of the dayside oval with the outer edge of the ring current agrees with the discussion of Sittler et al. (2006), while the association of the poleward boundary of the oval with the open-closed field boundary relates to the discussion of Cowley et al. (2004aCowley et al. ( , b, 2008)).We find that the UV oval does not map inward to the primary corotation breakdown region found by Wilson et al. (2009) at radial distances of ∼3-4 R S within the inner Enceladus torus.This maps to the southern ionosphere at a latitude of −62 • , well equatorward of the equatorward boundary of the UV oval at ∼ −70 • to −74 • , in agreement with the prior discussion of Cowley and Bunce (2003).However, the field-aligned currents associated with this inner corotation breakdown may well be associated with a lowerlatitude "secondary" oval observed in IR data by Stallard et al. (2008Stallard et al. ( , 2010)).
We finally note the very different field structures occurring in Saturn's outer magnetosphere and tail within the paraboloid models computed here for northward and southward-directed IMF, even for the rather small interplanetary field strengths prevailing.These may then give rise to differing IMF-dependent flow regimes on high-latitude open field lines, for example, between open field lines that pass directly from the ionosphere to the magnetopause in the corresponding hemisphere, and those that pass to the magnetopause in the opposite hemisphere through the equatorial plane, where they may be significantly influenced by magnetospheric plasma dynamics.For northward IMF only the former type of open field line is present, while for southward IMF both types are present in differing regions of the open polar cap.Further study of these features may be of interest.

Fig. 1 .
Fig.1.Cassini magnetic field data for DOY 43 to 46 of 2008.From top to bottom the data panels show the three components of the magnetic field in KSM coordinates, and the field magnitude.In the colour-coded region identifier at the top of the plot green corresponds to the magnetosphere, red to the magnetosheath, and blue to the solar wind.Principal bow shock crossings (red-blue transitions) are identified by vertical black dashed lines.The data at the bottom of the plot give the radial distance (R S ), latitude (deg), and local time (h) of the spacecraft.The 1 h blue vertical stripes labelled "A" to "D" correspond to the suitably lagged times of four HST imaging intervals.(FromBelenkaya et  al., 2010).
1236 E. S. Belenkaya et al.: Magnetospheric mapping of the dayside UV auroral oval at Saturn

Fig. 2 .
Fig. 2. Plots showing two HST UV images of Saturn's southern auroras, projected onto a spheroidal surface 1100 km above the atmospheric 1 bar level, with noon at the bottom and dawn to the left.The view is thus "through" the planet from the north.Dotted yellow lines show a latitude-longitude grid at 10 • intervals in each case.Only the dayside portion of the oval was well-observed due to the pre-equinoctial conditions prevailing, such that the images are truncated somewhat beyond the dawn-dusk meridian.Image identifiers are given at the top of each panel together with the start time of the ∼20 min combined exposure time.The locations of the poleward and equatorward boundaries of the emission in each image are shown by red crosses, determined at 10 • intervals of longitude from a 5 kR limiting emission intensity.The superposed solid orange lines show the open field regions for k S = 0.2 in each case, calculated using the paraboloid magnetic field model and the corresponding averaged IMF vector determined from Cassini data, indicated in rounded form at the bottom of each panel.

Figure 3 Fig. 3 . 25 Figure 4 Fig. 4 .
Figure 3Fig.3. Plots showing model magnetic field lines whose ionospheric footprints lie at constant southern latitudes, for the field model corresponding to case C with IMF penetration coefficient k s = 0.2.Panels (a) to (f) show results for latitudes from −70 • to −90 • , respectively, at steps of 4 • .The longitude step between each field line corresponds to 1 h of LT.Note the varying X-Y spatial scales in the various panels.Blue symbols mark intersections of the field lines with the planet's equatorial plane, while red symbols mark intersections with the southern magnetopause.The solid black circle shows the position of Saturn.

Fig. 7 .
Fig. 7. Plots showing model magnetic field lines whose ionospheric footprints lie at constant southern latitudes, for the field model corresponding to case A with IMF penetration coefficient k s = 0.2.The format is similar to Fig. 3. Blue squares mark intersections of the field lines with the planet's equatorial plane, while red and green squares mark intersections with the southern and northern magnetopause, respectively.

Fig. 8 .
Fig. 8. Intersection points of the open magnetic field lines emerging from Saturn's southern ionosphere, as shown in Fig. 7, with (a) the southern and (b) the northern magnetopause, for the case of image A with k s = 0.2.The format is the same as Fig. 4.