Search

International audience; Throughout 2017, the Hubble Space Telescope (HST) observed the northern far-ultraviolet aurorae of Saturn at northern solstice, during the Cassini Grand Finale. These conditions provided a complete viewing of the northern auroral region from Earth and a maximal solar illumination, expected to maximize the ionosphere-magnetosphere coupling. In this study, we analyze 24 HST images concurrently with Cassini measurements of Saturn's Kilometric Radiation and solar wind parameters predicted by two MHD models. The aurorae reveal highly variable components, down to timescales of minutes, radiating 7 to 124 +/-11 GW. They include a nightside-shifted main oval, unexpectedly frequent and bright cusp emissions and a dayside low latitude oval. On average, these emissions display a strong Local Time dependence with two maxima at dawn and pre-midnight, the latter being newly observed and attributed to nightside injections possibly associated with solstice conditions. These results provide a reference frame to analyze Cassini in situ measurements, whether simultaneous or not.

The magnetospheric cusps are important sites of the coupling of a magnetosphere with the solar wind. The combination of both ground- and space-based observations at Earth have enabled considerable progress to be made in understanding the terrestrial cusp and its role in the coupling of the magnetosphere to the solar wind via the polar magnetosphere. Voyager 2 fully explored Neptune's cusp in 1989 but highly inclined orbits of the Cassini spacecraft at Saturn present the most recent opportunity to repeatedly studying the polar magnetosphere of a rapidly rotating planet. In this paper we discuss observations made by Cassini during two passes through Saturn's southern polar magnetosphere. Our main findings are that i) Cassini directly encounters the southern polar cusp with evidence for the entry of magnetosheath plasma into the cusp via magnetopause reconnection, ii) magnetopause reconnection and entry of plasma into the cusp can occur over a range of solar wind conditions, and iii) double cusp morphologies are consistent with the position of the cusp oscillating in phase with Saturn's global magnetospheric periodicities., Journal accepted version before copy-editing: 55 pages, 12 figures. Accepted for publication in J. Geophys. Res. Space Physics

[1] Saturn's auroral activities have been suggested to be controlled by the seasonal variations of the polar ionospheric conductivities and atmospheric conditions associated with the solar extreme ultraviolet (EUV) flux. However, they have not yet been explained self-consistently by only the seasonal solar EUV effects. This study investigates the long-term variations of Saturnian Kilometric Radiation (SKR) as a proxy of the auroral activities, which were observed by Cassini's Radio and Plasma Wave Science experiment mostly during the southern summer (DOY (day of year) 001 2004 to DOY 193 2010). We deduced the height distribution of the SKR source region in the Northern (winter) and Southern (summer) Hemispheres from the remote sensing of SKR spectra. The peak spectral density of the southern (summer) SKR was found to be up to 100 times greater than that of the northern (winter) SKR, and the altitude of the peak flux was similar (∼ 0.8 Rs) in the Northern and Southern Hemispheres. The spectral densities in both hemispheres became comparable with each other around equinox in August 2009. These results suggest a stronger SKR source region during the summer than the winter related to the seasonal EUV effect, which is opposite to the trend observed in the Earth's kilometric radiation. A long-term correlation analysis was performed for the SKR, solar EUV flux, and solar wind parameters extrapolated from Earth's orbit by an magnetohydrodynamical simulation focusing on variations on timescales longer than several weeks. We confirmed clear positive correlations between the solar wind dynamic pressure and peak flux density in both the Southern and Northern Hemispheres during the declining phase of the solar cycle. We conclude that the solar wind variations on the timescale of the solar cycle control the SKR source region. In addition, it was also confirmed that the south-to-north ratios of SKR power flux and source altitudes are positively correlated with the solar EUV flux. This result strongly supports a seasonal EUV effect on the SKR source region. The variations of SKR activity over both seasonal and solar cycles are discussed in comparison to the terrestrial case.

On August 18, 1999, the Cassini spacecraft flew by Earth at an altitude of 1186 km on its way to Saturn. Although the flyby was performed exclusively to provide the spacecraft with sufficient velocity to get to Saturn, the radio and plasma wave science (RPWS) instrument, along with several others, was operated to gain valuable calibration data and to validate the operation of a number of capabilities. In addition, an opportunity to study the terrestrial radio and plasma wave environment with a highly capable instrument on a swift fly-through of the magnetosphere was afforded by the encounter. This paper provides an overview of the RPWS observations at Earth, including the identification of a number of magnetospheric plasma wave modes, an accurate measurement of the plasma density over a significant portion of the trajectory using the natural wave spectrum in addition to a relaxation sounder and Langmuir probe, the detection of natural and human-produced radio emissions, and the validation of the capability to measure the wave normal angle and Poynting flux of whistler-mode chorus emissions. The results include the observation of a double-banded structure at closest approach including a band of Cerenkov emission bounded by electron plasma and upper hybrid frequencies and an electron cyclotron harmonic band just above the second harmonic of the electron cyclotron frequency. In the near-Earth plasma sheet, evidence for electron phase space holes is observed, similar to those first reported by Geotail in the magnetotail. The wave normal analysis confirms the Polar result that chorus is generated very close to the magnetic equator and propagates to higher latitudes. The integrated power flux of auroral kilometric radiation is also used to identify a series of substorms observed during the outbound passage through the magnetotail.

By using two rotating noncollinear antennas or three spatially fixed noncoplanar antennas on a spacecraft, full information on the polarization and the direction of arrival of an electromagnetic wave can be obtained by measuring the voltages created by the electric field of the incident wave. The physical parameters (polarization and direction of arrival) of the incoming wave are related to the received voltages on the antenna system by the so-called direction-finding equations. Since the used antennas are generally of small directivity (electrically short monopoles or dipoles), the resulting system of equations is numerically close to singular, and generally no unique solution can be obtained for the physical parameters of the wave throughout the inversion process. However, there exists a very powerful tool for dealing with sets of equations that are singular or close to singular, known as singular-value decomposition (SVD), which precisely focuses the problem. For illustration, this paper analyzes the direction-finding equations for the Radio and Plasma Wave Science (RPWS) experiment on the Cassini spacecraft by using SVD techniques. It also compares the expected performances of RPWS with those of the Ulysses Unified Radio and Plasma Wave (URAP) experiment achieved at Jupiter for the kilometer and hectometer emissions. The RPWS experiment on Cassini, which will be launched in 1997, is supposed to observe wave phenomena between a few hundred Hertz and 16 MHz in the Saturnian magnetosphere.

[1] The determination of the internal magnetic field of Jupiter has been the object of many studies and publications. These models have been computed from the Pioneer, Voyager, and Ulysses measurements. Some models also use the position of the Io footprints as a constraint: the magnetic field lines mapping to the footprints must have their origins along Io's orbit. The use of this latter constraint to determine the internal magnetic field models greatly improved the modeling of the auroral emissions, in particular the radio ones, which strongly depends on the magnetic field geometry. This constraint is, however, not sufficient for allowing a completely accurate modeling. The fact that the footprint field line should map to a longitude close to Io's was not used, so that the azimuthal component of the magnetic field could not be precisely constrained. Moreover, a recent study showed the presence of a magnetic anomaly in the northern hemisphere, which has never been included in any spherical harmonic decomposition of the internal magnetic field. We compute a decomposition of the Jovian internal magnetic field into spherical harmonics, which allows for a more accurate mapping of the magnetic field lines crossing Io, Europa, and Ganymede orbits to the satellite footprints observed in UV. This model, named VIPAL, is mostly constrained by the Io footprint positions, including the longitudinal constraint, and normalized by the Voyager and Pioneer magnetic field measurements. We show that the surface magnetic fields predicted by our model are more consistent with the observed frequencies of the Jovian radio emissions than those predicted by previous models.

Radio bursts from Saturn lightning have been observed by the Cassini Radio and Plasma Wave Science instrument at frequencies of a few megahertz during several month-long storms since 2004. As the radio waves traverse Saturn's ionosphere on their way to the spacecraft, one can determine the peak electron density from the measurement of the low-frequency cutoff below which the radio bursts are not detected. In this way we obtained 231 profiles of peak electron densities that cover all Saturnian local times at a kronocentric latitude of 35°S, where the storms were spotted by the Cassini camera. Peak electron densities show a large variation at dawn and dusk and are around 5 × 10^4 cm^(−3), in fair agreement with radio occultation measurements at midlatitudes. At noon and midnight, the densities are typically somewhat above 10^5 cm^(−3) and around 10^4 cm^(−3), respectively. The diurnal variation is about 1 to 2 orders of magnitude for averaged profiles over one storm at 35°S. This is somewhat less compared to previous Voyager measurements which showed more than 2 orders of magnitude variation. The diurnal variation as well as the peak electron densities of Saturn's ionosphere tend to decrease with the decreasing solar EUV flux from 2004 until the end of 2009.

[1] The Cassini mission crossed the source region of the Saturn kilometric radiation (SKR) on 17 October 2008. On this occasion, the Radio and Plasma Wave Science (RPWS) experiment detected both local and distant radio sources, while plasma parameters were measured in situ by the magnetometer and the Cassini Plasma Spectrometer. A goniopolarimetric inversion was applied to RPWS three-antenna electric measurements to determine the wave vector k and the complete state of polarization of detected waves. We identify broadband extraordinary (X) mode as well as narrowband ordinary (O) mode SKR at low frequencies. Within the source region, SKR is emitted just above the X mode cutoff frequency in a hot plasma, with a typical electron-to-wave energy conversion efficiency of ∼1% (2% peak). The knowledge of the k vector is then used to derive the locus of SKR sources in the kronian magnetosphere, which shows X and O components emanating from the same regions. We also compute the associated beaming angle at the source θ′ = (k, −B) either from (1) in situ measurements or a model of the magnetic field vector (for local to distant sources) or (2) polarization measurements (for local sources). Obtained results, similar for both modes, suggest quasi-perpendicular emission for local sources, whereas the beaming pattern of distant sources appears as a hollow cone with a frequency-dependent constant aperture angle: θ′ = 75° ± 15° below 300 kHz, decreasing at higher frequencies to reach θ′ (1000 kHz) = 50° ± 25°. Finally, we investigate quantitatively the SKR polarization state, observed to be strongly elliptical at the source, and quasi-purely circular for sources located beyond approximately two kronian radii. We show that conditions of weak mode coupling are achieved along the raypath, under which the magnetoionic theory satisfactorily describes the evolution of the observed polarization. These results are analyzed comparatively with the auroral kilometric radiation at Earth.

The two main, low frequency radio components discovered at Neptune by the Planetary Radio Astronomy experiment carried aboard Voyager 2 have a well defined periodicity at about 16.1 hour. By analyzing all the available data, i.e. about 60 days around the closest approach for the 'Burst' component and 15 days for the 'Smooth' component, we determine, for each component, the best estimate of the radio period. We conclude that the two estimates are not statistically different. While the two kinds of radio emissions have very different characteristics (in particular their frequency ranges and beaming properties), and probably correspond to different emission processes, their modulation is very likely due to the rotation of the planetary magnetic field tied to the core of the planet, as it has already been assumed for the other giant planets. The deduced estimate of the sidereal rotation period of Neptune is 16.108 +/- 0.006 (or 16h06.5m +/- 0.04 m).

During the Ulysses inbound cruise to Jupiter the Unified Radio and Plasma Wave (URAP) experiment observed a variety of the planet's radio components in the frequency range below 1 MHz. Most of these emissions were already detected by the Voyager Radio Astronomy and Plasma Wave experiments, however, with much less sensitivity and different spectral coverage. These different radio components within the URAP dynamic spectra are identified, and their appearance with the previous Voyager observations are compared.

The Ulysses URAP experiment has detected Jovian radio emissions in the VLF range at distances from Jupiter in excess of 1.5 AU. The URAP observations represent the first synoptic observations of Jupiter in the VLF band, from 3 to 30 kHz. In this band lie the low-frequency extent of the bKOM emission, the escaping continuum emission, and the Jovian type IIIs. Initial results indicate that the continuum varies in frequency with the solar wind ram pressure at Jupiter, whereas, the Jovian type IIIs appear to be controlled to some extent by the planetary rotation, often appearing when system III longitude 100 deg faces the spacecraft.

[1] The high-inclination orbits of the Cassini spacecraft from autumn 2006 until spring 2007 allowed the Cassini/RPWS (Radio and Plasma Wave Science) instrument to observe Saturn Kilometric Radiation (SKR) from latitudes up to 60° for the first time. This has revealed a surprising new property of SKR: above ∼30° in observational latitude, a significant amount of SKR is strongly elliptically polarized, in marked contrast to previous observations from low latitudes, which showed only circular polarization. There are transitional latitudes where the elliptical polarization occurs in “patches” in the time-frequency spectrograms next to regions of still completely circularly polarized SKR. From ∼45° to 60° in northern latitude, it is found that most of the SKR is elliptically polarized throughout its entire frequency range with an average degree of ∼0.7 in linear polarization. We demonstrate the ellipticity of SKR by using the concept of “apparent polarization” in case of two-antenna measurements, but also show three-antenna measurements from which the polarization can be unambiguously determined. Possible reasons for the variation of SKR polarization with the observer's latitude will be discussed.

International audience; Jupiter's radio emissions are dominated in intensity by decametric radio emissions due to the Io-Jupiter interaction. A significant part of these emissions consists of short radio bursts (so-called S-bursts) drifting in time and frequency. Previous analyses suggest that these emissions are cyclotron-maser emissions in the flux tube connecting Io or Io's wake to Jupiter. We present simulations of these electrons under the assumption of acceleration by Alfvén waves in the Io flux tube. Near Jupiter, a loss cone and a ring distribution appear in the magnetically mirrored electron population, which can then amplify extraordinary (X) mode radio waves. The X-mode growth rate is computed, which allows us to build theoretical dynamic spectra of the resulting Jovian radio emissions. Additional potential structures are assumed in the Jovian auroral region. We reconstruct their impact on the morphology of the emission. They match some of the time-frequency patterns observed with Jovian S-bursts. This provides the first evidence of bipolar electrostatic structures in the Jovian auroral region.

[1] Since Cassini's arrival at Saturn in 2004, the Radio and Plasma Wave Science instrument has detected numerous narrowband (NB) radio emission events. These emissions, mostly detected around 5 and 20 kHz, usually occur periodically for several days after intensifications of Saturn kilometric radiation. We present calculations based on an electron density profile of Saturn's plasma torus and a dipole magnetic field model showing that the NB emissions originate from the northern and southern edges of Saturn's plasma torus at L shells ∼ 8 to 10 for 5-kHz NB and L ∼ 4 to 7 for 20-kHz NB. In many cases, Cassini passes through the source region of the 20-kHz NB, as indicated by intense electrostatic upper hybrid (ESUH) waves in close proximity to electromagnetic emissions on spectrograms. The positions of the spacecraft when intense ESUH waves are observed agree with the model predictions of the NB source locations. Source locations determined by goniopolarimetric (also known as direction-finding) analysis of the NB emissions also support the above results, although sometimes the directions of arrival point toward the region interior to Saturn's plasma torus. A polarization reversal technique is applied to localize the NB emissions observed during spacecraft rotation, on the basis of the fact that the source is within the antenna plane when the apparent circular polarization degree switches sign. The NB emissions are found to be L-O mode polarized, which is consistent with the prediction of linear/nonlinear mode conversion theory. It is also found that sometimes right-hand polarized NB emissions are generated at second harmonic frequencies of the 20-kHz NB; in which case, wave-wave interactions between oppositely propagating ESUH waves may play an important role in the mode conversion process.

[1] We analyze the records of Jupiter's decameter radio emissions obtained during an Io-A S-burst storm on 15 March 2005. The observations were performed at the world's largest decameter array, UTR-2, which is equipped with a digital receiver capable of catching waveforms of duration ∼3 s with temporal resolution defined by the sampling rate of ∼66 MHz. A Hilbert transform based algorithm has been applied to study narrow-band spectral patterns demonstrating quasi-linear drift over time-frequency plane. The instantaneous amplitude and phase information has been extracted from the recorded waveforms with the purpose of analyzing microsecond-scale coherent events in the S-burst emission. A statistical model of narrow band random process is proposed for describing such features in the observed waveforms as coherent segments, phase jumps, nonlinear frequency drift, etc. It is shown that the study of coherence properties in terms of instantaneous phase is equivalent to Fourier analysis of a narrowband signal. This implies that no particular mechanism (such as superimposed modulation or oscillation) is required for generating the observed coherent phase structures of S-burst emission: those, as well as the pulse-like envelope structures, emerge naturally at the output of a narrow band filter applied to a random noise. It is further suggested that probability distribution function of instantaneous amplitude gives an important insight into the underlying physical mechanism of S-burst generation. In particular, it is demonstrated that models based on the concept of “generator,” i.e., a nonlinear system with feedback, are less suitable for reproducing the observational characteristics of S-bursts at microsecond time scale resolution. On the other hand, the concept of “amplifier,” i.e., a linear system (without feedback) that enhances the fluctuations within a narrow band, fits the observational data well. This conclusion is consistent with S-burst generation mechanism via cyclotron-maser instability, which is indeed a resonant wave amplification process.

[1] Similar to past studies at Jupiter, we conduct an investigation of possible associations of radio emission occurrence probability with the orbital phases of Saturnian moons. We use a new definition of the Saturn longitude system (SLS) based on the results of Kurth et al. (2007). This paper presents results of our findings to date, sampling a large portion of the Radio and Plasma Wave Science (RPWS) instrument data over the frequency range 12 kHz < f < 16 MHz. We also investigate the intensity of Saturn kilometric radiation (SKR) as a function of the local time and subsolar longitude. When Titan is near a local time of midnight, there is a significant increase in the occurrence probability of SKR and a diminution of SKR when Titan is near local noon and afternoon. This indicates Titan may play a role in the process of substorm generation at Saturn perhaps due to its large plasma wake. Rhea displays a marginal orbital phase ‘‘control’’ of a subset of the SKR. In the past there have been conflicting reports of the absence of SKR emission at particular local times of Dione. We find no long-term, statistical influence of Dione on SKR occurrence probability. In addition, we find no significant statistical influence of Enceladus or Tethys on SKR occurrence probability.

[1] During Cassini's orbit insertion about Saturn, the spacecraft passed within 1.4 Rs of the planet passing from dayside into the nightside region. During this nightside passage, the onboard Radio and Plasma Wave (RPWS) instrument surprisingly detected Saturn kilometric radiation (SKR). Prior to this encounter, it was believed that SKR originated from a high-latitude dayside source, and radio beams from such a source would not be viewable in this nearplanet night-side location. Subsequent analysis presented here reveals that this SKR did indeed originate from the near-midnight region on field lines near L ∼ 10–15. Such a radio source suggests the presence of an active region in the night-side inner magnetosphere; this source possibly being near the outer edge of the icy-moon created plasma torus surrounding the planet. The implication is that some of the SKR is driven by an internal energy source that may also account for recent UV aurora observations.

[1] We present an analytical inversion method to achieve direction-finding (DF) (i.e., retrieve the direction of arrival of an incoming electromagnetic wave, its flux, and its full polarization state) using radio measurements performed using a system of two or three electric dipole antennas on a three-axis stabilized spacecraft. The Radio and Plasma Wave Science (RPWS) radio receiver on board Cassini includes such instantaneous DF capabilities, and so does the Solar Terrestrial Relations Observatory (STEREO) Waves radio receiver. We also present an analytical solution of the inverse problem which consists of calibrating the electric dipole orientations and effective lengths using a known radio source. Error sources (imperfect knowledge of antenna parameters, digitization errors, signal to noise ratio, etc.) and their propagation through the analytical inversion have been studied. The typical expected accuracy of our DF inversion is 1 dB [V2/Hz] for flux measurements, about 1–2° for source position and a few percent for degrees of polarization. For the antenna calibration procedure the expected accuracy is of the order of 2° on antenna direction and of 1% on antenna length. We define the data selection criteria to be used during both DF analysis and antenna calibration. We also discuss the limitations of the methods and the ways to improve their accuracy.

International audience; We apply the calibration method developed by [2001] to the data from the Cassini/Radio and Plasma Wave Science (RPWS) High-Frequency Receiver in order to derive flux density measurements of six components of the Jovian low-frequency radio spectrum over the full frequency range of the instrument (3.5 kHz to 16.1 MHz). The estimated accuracy is better than 50%, i.e., much less than the intrinsic variations of the flux densities of these radiosources. It is mainly limited by the accuracy of the model used for the radio galactic background. Instrumental parameters such as the antennas' effective lengths and base capacitance are constrained in the calibration process. From 6 months of calibrated data centered on the Cassini-Jupiter flyby, we derive the average and peak Jovian radio spectrum between 3.5 and 16.1 MHz and its range of fluctuations, from which we deduce constraints on the beaming of the various radio components and estimate the power emitted by each component. Our calibration procedure also allows us to compare Cassini measurements of the Jovian radio spectrum with ground-based measurements performed, e.g., in Nançay above the ionospheric cutoff (10-15 MHz). It will be used to derive absolute flux measurements during the Saturn tour.

An exceptional Jovian aurora was detected in the FUV on December 21, 1990, by means of Vilspa and Goddard Space Flight Center (GFSC) International Ultraviolet Explorer (IUE) observations. This event included intensification by a factor of three between December 20 and 21, leading to the brightest aurora identified in the IUE data analyzed, and, in the north, to a shift of the emission peak towards larger longitudes. The Jovian radio emission simultaneously recorded at decameter wavelengths in Nancay also exhibits significant changes, from a weak and short-duration emission on December 20 to a very intense one, lasting several hours, on December 21. Confirmation of this intense radio event is also found in the observations at the University of Florida on December 21. The emissions are identified as right-handed Io-independent 'A' (or 'non Io-A') components from the northern hemisphere. The radio source region deduced from the Nancay observations lies, for both days, close to the UV peak emission, exhibiting in particular a similar shift of the source region toward larger longitudes from one day to the next. A significant broadening of the radio source was also observed and it is shown that on both days, the extent of the radio source closely followed the longitude range for which the UV brightness exceeds a given threshold. The correlated variations, both in intensity and longitude, strongly suggest that a common cause triggered the variation of the UV and radio emissions during this exceptional event. On one hand, the variation of the UV aurora could possibly be interpreted according to the Prange and Elkhamsi (1991) model of diffuse multicomponent auroral precipitation (electron and ion): it would arise from an increase in the precipitation rate of ions together with an inward shift of their precipitation locus from L approximately equal 10 to L approximately equal 6. On the other hand, the analysis of Ulysses observations in the upstream solar wind suggests that a significant disturbance in the solar wind, involving the generation of an interplanetary shock and the presence of a CME have interacted with the Jovian magnetosphere at about the time of the auroral event. Both arguments suggest that we may have observed for the first time a magnetic storm-type interaction in an outer planet magnetosphere, affecting simultaneously several auroral processes. Conversely, the observed relationship between the level of UV auroral activity and the detection of decameter emission (DAM), if it were a typical feature, might argue in favour of a more direct and permanent association between the auroral processes leading to UV and radio aurorae, possibly related to 'discrete-arc'-like activity and electron precipitation.

New observations of Jovian narrow-band kilometric (nKOM) radio emissions were made by the Unified Radio and Plasma Wave (URAP) experiment on the Ulysses spacecraft during the Ulysses-Jupiter encounter in early February 1992. These observations have demonstrated the unique capability of the URAP instrument for determining both the direction and polarization of nKOM radio sources. An important result is the discovery that nKOM radio emission originates from a number of distinct sources located at different Jovian longitudes and at the inner and outermost regions of the Io plasma torus. These sources have been tracked for several Jovian rotations, yielding their corotational lags, their spatial and temporal evolution, and their radiation characteristics at both low latitudes far from Jupiter and at high latitudes near the planet. Both right-hand and left-hand circularly polarized nKOM sources were observed. The polarizations observed for sources in the outermost regions of the torus seem to favor extraordinary mode emission.

The Neptune flyby in 1989 added a new planet to the known number of magnetized planets generating nonthermal radio emissions. We review the Neptunian radio emission morphology as observed by the planetary radio astronomy experiment on board Voyager 2 during a few weeks before and after closest approach. We present the characteristics of the two observed recurrent main components of the Neptunian kilometric radiation, i.e., the 'smooth' and the 'bursty' emissions, and we describe the many specific features of the radio spectrum during closest approach.

Although lightning-triggered whistlers were detected by the Voyager 2 plasma wave experiment at Neptune, only four possible lightning sferics were detected by the planetary radio astronomy instrument. This low level of detection places strict limits on the characteristics of Neptunian lightning, ruling out lightning with power levels as high as those observed at Saturn or Uranus, but leaving the intriguing possibility that lightning slightly stronger than normal terrestrial lightning flashes but with much lower occurrence rate could exist. Possible scenarios to explain the observations are examined, including reduced NH3 concentration in the planet's atmosphere and an unusually slow risetime discharge process.

During the Voyager Saturn mission the Planetary Radio Astronomy (PRA) experiment observed sporadic broadband emissions (20 KHz to at least 40 MHz) during a few days around the Voyager 1 and 2 closest approaches to the planet. These emissions are interpreted to arise from short-lived electrostatic discharges in the vicinity of Saturn. We have made a statistical study of their characteristics from the Voyager 1 and 2 high-band data of the PRA experiment (1.2 MHz to 40.2 MHz). In this paper, we compare the results obtained for the two encounters and describe in detail the appearance, intensity, and spectrum of the Saturn electrostatic discharges (SED). We show that the SED spectrum is likely smooth in the frequency range 1–40 MHz and extends down to 20 KHz. We have determined the period of the SED occurrence, which seems to be slightly different for Voyager 1 and 2 encounters (10 hours 9 min for the Voyager 1 encounter and 10 hours 00 min for the Voyager 2 encounter). An attempt to interpret the statistical results is presented.