ENA periodicities at Saturn

Transcription

1 Click Here for Full Article GEOPHYSICAL RESEARCH LETTERS, VOL. 35, L07102, doi: /2008gl033230, 2008 ENA periodicities at Saturn J. F. Carbary, 1 D. G. Mitchell, 1 P. Brandt, 1 C. Paranicas, 1 and S. M. Krimigis 1 Received 9 January 2007; revised 22 February 2008; accepted 3 March 2008; published 2 April [1] The Magnetospheric Imaging Instrument (MIMI) on the Cassini spacecraft has observed energetic neutral atoms (ENA) at Saturn for three years from 2004 to The kev hydrogen and kev oxygen neutrals were examined for periodic behavior by projecting their images onto a plane perpendicular to Saturn s equatorial plane and summing the emissions along the vertical. The resulting strip images were then subjected to a Lomb periodogram analyses for the period between mid-2004 to mid The hydrogens display erratic periods in the 8 13 hour range, while the oxygens exhibit a very strong, repeatable period of 10.8 ± 0.2 hours, which is close to the nominal period of Saturn kilometric radiation as well as the period of energetic particles in Saturn s outer magnetosphere. Citation: Carbary, J. F., D. G. Mitchell, P. Brandt, C. Paranicas, and S. M. Krimigis (2008), ENA periodicities at Saturn, Geophys. Res. Lett., 35, L07102, doi: /2008gl Introduction [2] Periodicities in the Saturn magnetosphere are ubiquitous. Voyager-era observations discovered periodicities in kilometric radiation [Desch and Kaiser, 1981], the spectral behavior of charged particles [Carbary and Krimigis, 1982], and even in the spoke phenomena of Saturn s rings [Porco and Danielson, 1982]. The apparent regularity of the Saturn kilometric radiation (SKR) led to the definition of a longitude system based on its 10 hour 46 minute period [Davies et al., 1996]. Later analysis of Voyager magnetometer data revealed periodic structures in the magnetic field of Saturn [Espinosa and Dougherty, 2000]. The Cassini mission has confirmed these magnetic field periodicities [Giampieri et al., 2006], and extended Saturn periodicities to energetic neutral particles [Krimigis et al., 2005] and fluxes of charged particles [Carbary et al., 2007a]. [3] While the basic periodicity of magnetospheric phenomena at Saturn is not questioned, its exact period is known to slowly change. Observations of SKR from the Ulysses spacecraft indicated small but noticeable variations in its period over long time intervals [Galopeau and Lecacheux, 2000]. Cassini measurements confirmed a shift of nearly six minutes in the SKR period [Gurnett et al., 2005], which led to the construction of a variable-period longitude system [Kurth et al., 2007]. [4] Various explanations have emerged to explain the variations in SKR period [e.g., Galopeau and Lecacheux, 2000; Cecconi and Zarka, 2005; Gurnett et al., 2007], but this paper will not address the variation. However, Paranicas 1 Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland, USA. Copyright 2008 by the American Geophysical Union /08/2008GL033230$05.00 et al. [2005] have explained the periodicity in ENA emissions in terms of a rotating point source imposed on a constant, extended source. They took into account the observational geometry of the instrument and the relative motion of the source and observer, but did not explicitly determine an ENA period nor its possible variation. The present paper examines Saturn s energetic hydrogen (20 50 kev) and oxygen ( kev) emissions for about three years of observation and specifically determines their periodicity or lack thereof. 2. Instrument and Data Set [5] Measurements of energetic neutral atoms at Saturn were obtained by the Ion Neutral Camera (INCA), which is part of the Magnetospheric IMaging Instrument (MIMI) on the Cassini spacecraft [Krimigis et al., 2004]. INCA uses the time-of-flight technique to observe neutral atoms in discrete energy passbands ranging from 3 to >220 kev. INCA employs a field of view of that, in this case, was divided into an array of pixels. This study uses INCA measurements of neutral hydrogen atoms with energies of kev and neutral oxygens with energies of kev. For each pixel count rate data were converted to ENA radiances (counts/cm 2 srs) using instrument response functions and geometric factors. ENA images were obtained at a frequency of one per four minutes, but were averaged into half-hour time intervals for this analysis. [6] For most of the Cassini mission from orbit insertion in July 2004 to mid-2007, INCA viewed Saturn at shallow angles (latitudes <20 ) relative to Saturn s equatorial plane, while some of the viewing occurred when Cassini visited high latitudes and the view angles were steep (latitudes >20 ). This investigation uses ENA images without regard to the viewing orientation. INCA observations were intermittent, but generally extended to over 12 hours per day and usually lasted longer than the 10.8 hour SKR period. Gaps of several hours may interrupt this viewing, but otherwise the ENA observations are continuous for months at a time. This continuity afforded INCA the opportunity to observe ENA dynamics over many Saturn rotations, and, with the proper techniques, allows construction of a rather good statistical database for the investigation of ENA periodicities. 3. Analysis and Data Samples [7] Half-hour averages of ENA images were projected onto a plane perpendicular to the cylindrical position vector r of the spacecraft (i.e., r =xx 1 +yy 1, where unit vector x 1 points sunward in Saturn s equatorial plane, z 1 lies along the Saturn spin axis, and y 1 completes the right hand system). In the projected coordinate plane, the z 2 axis is L of5

2 Figure 1. Projection geometry (top left), a sample projected image from kev hydrogens (top right) and the vertical-summation profile (bottom). The projected z 2 axis is the spin axis of Saturn, while the projected y 2 axis is perpendicular to the spacecraft location vector in the equatorial plane. Intensities are color-coded with reds being most intense and purples being least intense; white dots indicate centers of projected pixels. The profile in the bottom panel was obtained by vertically summing image pixels within Dy 2 =5R S bins in the image. the same as z 1 (Saturn s spin axis), the x 2 axis lies along r, and the y 2 axis is perpendicular to r. This projection plane slowly rotates as the spacecraft orbits Saturn. Only pixels within a 60 x 40 R S box centered on Saturn are analyzed. Furthermore, such a projected image will remain coherent if the observer is sufficiently far from the planet (>20 R S ) and at a sufficiently low latitude (<40 ). The top left panel of Figure 1 indicates the projection geometry and defines the x 1 y 1 z 1 and x 2 y 2 z 2 coordinate systems. The top right panel shows a sample projected image of hydrogen ENA emissions. Projected pixels appear as white dots in the projected image. [8] The bottom panel of Figure 1 represents a summation of pixel intensities in the vertical dimension of the image. The summations were performed in 5 R S bins. In this instance, the projected image exhibits two peaks, one on either side of the y 2 = 0 line. The double peaks appear in the bottom profile also. The peak intensities often seem to vary periodically, which led Paranicas et al. [2005] to model the phenomenon in terms of a source rotating about the z axis. Because of the Compton-Getting effect, a source corotating (or nearly corotating) with Saturn would appear more intense on the left side of the projected image and less intense on the right side, which is the case in Figure 1. [9] Each projected image generates one summation profile in the vertical dimension. The summation profiles are then stacked horizontally to yield a compact representation of the viewing as a function of time. Figure 2 shows seven days of stacked profiles. The time period corresponds to that examined by Paranicas et al. [2005] when the spacecraft was outbound on the morning side of the magnetosphere at low latitude. INCA observations enjoyed nearly continuous, optimal viewing during this period; more gaps usually appear in the data set. The reader will readily apprehend the periodic behavior of the ENA emissions in the stacked profiles. The hydrogens (top) exhibit a curious spiralized structure suggesting back and forth or sweeping motion, while the oxygens (bottom) display a different periodicity reminiscent of simple flashing. 4. Lomb Periodogram Results [10] The exact period for these emissions can be obtained by applying a Lomb periodogram analysis to the central bins in the stacked images. The central bins are those situated along the centerlines (horizontal dashed lines) in Figure 2. Use of the central bins will eliminate Compton- Getting effects of azimuthal motion about Saturn [see Paranicas et al., 2005, equation (1)] A Lomb analysis offers a robust method of finding periodic behavior that overcomes both gaps and noise in the data [e.g., Press et al., 1992]. Lomb analyses have previously been used to determine periodicities in magnetometer and charged particle data [Giampieri et al., 2006; Carbary et al., 2007a]. To be effective, however, the Lomb analysis should be applied over long time intervals. In this case, one periodogram was generated for each species for each half a year (183 days) from mid-2004 (pre-soi) to mid No detrending or smoothing was performed. To overcome possible Doppler effects as well as de-focusing from being too close to the source, images were excluded when the spacecraft was within 20 R S of Saturn. Observations from all latitudes were included to promote continuity of the analysis; the 2of5

3 Figure 2. Stacked summation profiles for kev hydrogens (top) and kev oxygens for a period of several days in 2004 (the same period examined by Paranicas et al. [2005]). Vertical dotted lines indicate radial distances of the spacecraft from Saturn. For the observing geometry in this case, negative values of the y 2 represent nightside locations in the magnetotail, while positive values represent dayside locations. high latitude observations (>40 ) comprise only a small part of the data set (5%). [11] Figure 3 shows a typical Lomb periodogram from one half-year segment of the data set. The hydrogens present a noisy spectrum with many possible periods, none of which dominate the spectrum. In this case, the main hydrogen peak was at 13.2 hours, although a strong secondary peak also appears near 10.6 hours. On the other hand, a very strong peak at 10.8 hours emerges from the oxygen periodogram. This peak has a signal-to-noise ratio of 10 (peak strength divided by standard deviation), which is the strongest periodic signal from particles measured in Saturn s magnetosphere [Carbary et al., 2007a]. The oxygen period corresponds closely with the 10.8 hour base period of the SKR emissions [Kurth et al., 2007]. Periodograms from different time periods would show behavior similar to that of Figure 3. [12] Figure 4 summarizes the Lomb periods derived from six half-year periods during the Cassini mission. The six H periods have an average of 11.7 hours and a standard deviation of ±2.4 hours, while the six O periods average 10.8 hours ± 0.2 hours. Thus, the hydrogens have erratic periods as a result of noisy periodograms, and the oxygens exhibit a very stable period over the entire three-years. For comparison, a dashed line shows the variable period recently derived from fitting a fifth-order polynomial to Saturn kilometric radio emissions [Kurth et al., 2008]. Deviations in the oxygen period from the SKR period occur during the last year of this record when Cassini visited high latitudes. Notably, the high latitude observations are those which most stress the projection method used here, so periodicities in this time period are relatively more uncertain (i.e., lower SNRs) than those at earlier times. 5. Discussion [13] Three years of periodogram analyses reveal that hydrogen ENA has no reproducible period, while the oxygen ENA has a very strong, repeatable period of 10.8 hours. The oxygen period is indistinguishable from the base period for the SKR radiation [Kurth et al., 2007, 2008] or the period found for charged particles in Saturn s outer magnetosphere [Carbary et al., 2007a]. The similarity between the SKR and oxygen periods suggests the two may have the same cause. That is, the same mechanism generating the SKR periodicity may also be causing the oxygen periodicity. Paranicas et al. [2005] have suggested a rotating point source in Saturn s inner magnetosphere may cause such a periodicity. Alternatively, a flashing source that does not rotate (or rotates very little) would also produce a strong peak in a periodogram. Similar to the Paranicas model, a rotating source that emits only within a limited local time sector would emulate a flashing source. This type of flashing source, limited to the magnetotail sector, would be consistent with possible substorm-like activity postulated by Mitchell et al. [2005]. If this is the case, the pseudosubstorms would be periodic and occur at approximately the rotation period of Saturn. 3of5

4 Figure 3. Lomb periodogram spectra for (top) kev hydrogens and (bottom) kev oxygens for the last half of The time profiles input to the Lomb analyses derive from the centerline profiles of stacked summation profiles such as those in Figure 2. [14] Differences between the oxygen and hydrogen periodicities remain problematic. The energetic oxygen atoms result from charge-exchange collisions of energetic oxygen (O+) ions with neutral H and O, while the energetic hydrogen atoms result from collisions of energetic protons (H+) with the same neutral H and O. Differences in the ENA emissions should therefore arise from differences in the source populations of O+ and H+. These source populations might be energized by different mechanisms at different times, as has been observed and explained for terrestrial substorms [Fok et al., 2006]. Notably, O+ and H+ with energies in the kev range have considerably different speeds and gyroradii, and differences in periods may reflect promptness with which the emissions are generated. [15] The oxygen periodicity changes slightly during the final year covered by this analysis. This variation might result from some unrecognized bias in the projection, or might be caused by an actual deviation of the periodicity apparent only from high latitudes [Carbary et al., 2007b]. Alternately, the actual period of the ENA emissions may be changing in a manner similar to that of the SKR emissions [Kurth et al., 2008]. 6. Conclusions [16] ENA emissions from Saturn have been examined using a Lomb periodogram analysis applied to three years of Cassini data from the MIMI/INCA camera. The kev hydrogens display an irregular period (11.7 ± 2.4 hours) characterized by a noisy signal in which no strong peak is seen in a periodogram. The kev oxygens exhibit a very strong periodicity with a single periodogram peak at 10.8 (±0.2) hours, which is the same as the SKR base period and the charged particle period seen in the outer magnetosphere. Figure 4. Lomb periods for kev hydrogens and kev oxygens from mid-2004 to mid The periods represent the peaks in the Lomb periodograms for half-year time intervals. The SLS-3 period derives from a fifth-order polynomial defining the SKR period [Kurth et al., 2008]. 4of5

5 [17] Acknowledgments. This work was supported in part by the NASA Office of Space Science under Task Order 003 of contract NAS between NASA Goddard Space Flight Center and the Johns Hopkins University, and in part by NASA Cassini Data Analysis Program grant NNX07AJ69G. We thank Ed Roelof, Barry Mauk, and Steve Christon for helpful discussions, and Martha Kusterer for reducing and managing the INCA data. References Carbary, J. F., and S. M. Krimigis (1982), Charged particle periodicity in the Saturnian magnetosphere, Geophys. Res. Lett., 9, Carbary, J. F., D. G. Mitchell, S. M. Krimigis, D. C. Hamilton, and N. Krupp (2007a), Charged particle periodicities in Saturn s outer magnetosphere, J. Geophys. Res., 112, A06246, doi: /2007ja Carbary, J. F., D. G. Mitchell, S. M. Krimigis, and N. Krupp (2007b), Evidence for spiral pattern in Saturn s magnetosphere using the new SKR longitudes, Geophys. Res. Lett., 34, L13105, doi: / 2007GL Cecconi, B., and P. Zarka (2005), Model of a variable radio period for Saturn, J. Geophys. Res., 110, A12203, doi: /2005ja Davies, M. E., et al. (1996), Report for the IAU/IAG/COSPAR working group on cartographic coordinates and rotational elements of the planets and satellites: 1994, Celestial Mech. Dyn. Astron., 63, Desch, M. D., and M. L. Kaiser (1981), Voyager measurements of the rotation period of Saturn s magnetic field, Geophys. Res. Lett., 8, Espinosa, S. A., and M. K. Dougherty (2000), Periodic perturbations in Saturn s magnetic field, Geophys. Res. Lett., 27, Fok, M.-C., T. E. Moore, P. C. Brandt, D. C. Delcourt, S. P. Slinker, and J. A. Fedder (2006), Impulsive enhancements of oxygen ions during substorms, J. Geophys. Res., 111, A10222, doi: /2006ja Galopeau, P. H. M., and A. Lecacheux (2000), Variations of Saturn s radio period measured at kilometer wavelengths, J. Geophys. Res., 105, 13,089 13,101. Giampieri, G., M. K. Dougherty, E. J. Smith, and C. T. Russell (2006), A regular period for Saturn s magnetic field that may track its internal rotation, Nature, 441, 62 64, doi: /nature Gurnett, D. A., et al. (2005), Radio and plasma wave observations at Saturn from Cassini s approach and first orbit, Science, 307, , doi: /science Gurnett, D. A., A. M. Persoon, W. S. Kurth, J. B. Groene, T. F. Averkamp, M. K. Dougherty, and D. J. Southwood (2007), The variable rotation period of the inner region of Saturn s plasma disk, Science, 316, , doi: /science Krimigis, S. M., et al. (2004), Magnetospheric imaging instrument (MIMI) on the Cassini mission to Saturn, Space Sci. Rev., 114, Krimigis, S. M., et al. (2005), Dynamics of Saturn s magnetosphere from MIMI during Cassini s orbital insertion, Science, 307, , doi: /science Kurth, W. S., A. Lecacheux, T. F. Averkamp, J. B. Groene, and D. A. Gurnett (2007), A Saturnian longitude system based on a variable kilometric radiation period,, Geophys. Res. Lett., 34, L02201, doi: / 2006GL Kurth, W. S., T. F. Averkamp, D. A. Gurnett, J. B. Groene, and A. Lecacheux (2008), An update to a Saturnian longitude system based on kilometric radio emissions, J. Geophys. Res., doi: /2007ja012861, in press. Mitchell, D. G., et al. (2005), Energetic ion acceleration in Saturn s magnetotail: Substorms at Saturn?, Geophys. Res. Lett., 32, L20S01, doi: /2005gl Paranicas, C., D. G. Mitchell, E. C. Roelof, P. C. Brandt, D. J. Williams, S. M. Krimigis, and B. H. Mauk (2005), Periodic intensity variations in global ENA images of Saturn, Geophys. Res. Lett., 32, L21101, doi: /2005gl Porco, C. C., and G. E. Danielson (1982), The periodic variation of spokes in Saturn s rings, Astron. J., 87, Press, W. H., S. A. Teukolsky, W. T. Vetterling, B. P. Flannery (1992), Numerical Recipes in C: The Art of Scientific Computing, pp , Cambridge Univ. Press, Cambridge, U.K. P. Brandt, J. F. Carbary, S. M. Krimigis, D. G. Mitchell, and C. Paranicas, Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA. (james.carbary@jhuapl.edu) 5of5