x-ray and extreme ultraviolet emission from cometsplorer (euve), and more recently, chandra and...

Lisse et al.: X-Ray and Extreme UV Emission from Comets 631

631

X-Ray and Extreme Ultraviolet Emission from Comets

C. M. LisseUniversity of Maryland

T. E. CravensUniversity of Kansas

K. DennerlMax-Planck-Institut für Extraterrestrische Physik

The discovery of high energy X-ray emission in 1996 from C/1996 B2 (Hyakutake) hascreated a surprising new class of X-ray emitting objects. The original discovery (Lisse et al.,1996) and subsequent detection of X-rays from 17 other comets (Table 1) have shown that thevery soft (E < 1 keV) emission is due to an interaction between the solar wind and the comet’satmosphere, and that X-ray emission is a fundamental property of comets. Theoretical and ob-servational work has demonstrated that charge exchange collisions of highly charged solar windions with cometary neutral species is the best explanation for the emission. Now a rapidly chang-ing and expanding field, the study of cometary X-ray emission appears to be able to lead us toa better understanding of a number of physical phenomena: the nature of the cometary coma,other sources of X-ray emission in the solar system, the structure of the solar wind in the helio-sphere, and the source of the local soft X-ray background.

1. INTRODUCTION

Astrophysical X-ray emission is generally found to origi-nate from hot collisional plasmas, such as the million-degreegas found in the solar corona (e.g., Foukal, 1990), the 100-million-degree gas observed in supernova remnants (e.g.,Cioffi, 1990), or the accretion disks around neutron stars andblack holes. As electromagnetic radiation with wavelength λbetween about 0.01 nm and 100 nm (1 nm = 10–9 m), ex-treme ultraviolet (EUV) and X-ray radiation are importantfor solar system and astrophysical applications because thephotons are sufficiently energetic and penetrating to ionizeneutral atoms and molecules, and can thus drive chemicalreactions. In fact, current estimates of the X-ray burden peratom in the young solar system are some 103–104 photonsper atom, even as far out as the proto-Kuiper belt, as theyoung Sun was much more X-ray active than now (Feigel-son, 1982; Dorren et al., 1995).

The Sun is not the only source of X-rays in the solar sys-tem (Cravens, 2000a, 2002a). Prior to 1996, X-rays werefound in scattering of solar X-rays from the terrestrial atmos-phere and in the terrestrial aurora, as scattered solar X-raysoff the illuminated surface of the Moon, and from the jovianaurora. Nonetheless, the 1996 discovery, using the RöntgenSatellite (ROSAT), by Lisse and co-workers (Lisse et al.,1996) (Fig. 1) of strong X-ray emission from Comet C/1996 B2 (Hyakutake) was very surprising because cometaryatmospheres are known to be cold and tenuous, with charac-teristic temperatures between 10 and 1000 K. Compared toother X-ray sources, comets are moderately weak — the

total X-ray power, or luminosity, of C/Hyakutake was mea-sured to be approximately 109 W. The emission was alsoextremely “soft”, or of low characteristic photon energy —only X-rays of energies less than ~1 keV (or wavelengthslonger than ~1.2 nm) were detected from C/Hyakutake. Thetotal amount of energy emitted in X-rays from a comet isapproximately 10–4 the energy delivered to a comet fromthe Sun due to photon insolation and solar wind impact(Lisse et al., 2001).

2. OBSERVED CHARACTERISTICS OFCOMETARY X-RAY EMISSION

Shortly after the initial C/Hyakutake detection, soft X-rayemission from four other comets was found in the ROSATarchival database, confirming the discovery (Dennerl et al.,1997). X-ray emission has now been detected from 18 com-ets to date (Table 1) using a variety of X-ray sensitive space-craft — BeppoSAX, ROSAT, the Extreme Ultraviolet Ex-plorer (EUVE), and more recently, Chandra and XMM-Newton. All comets within 2 AU of the Sun and brighterthan V = 12 have been detected when observed. We nowrecognize that X-ray emission is a characteristic of all ac-tive comets.

The observed characteristics of the emission can orga-nized into the following four categories: (1) spatial mor-phology, (2) total X-ray luminosity, (3) temporal variation,and (4) energy spectrum. Any physical mechanism that pur-ports to explain cometary X-ray emission must account forall these characteristics.

632 Comets II

TABLE 1. Observation times, instruments, and energies* for comets detected†

through December 2002 in the X-ray or extreme ultraviolet.

Comet Time Instrument Energy (keV) Detection Reference

45P/Honda-Mrkos-Pajdušáková Jul 1990 ROSAT PSPC/WFC 0.09–2.0 Yes [1]

C/1990 K1 (Levy) Sep 1990 ROSAT PSPC /WFC 0.09–2.0 Yes [1]Jan 1991 0.09–2.0 Yes [1]

C/1991 A2 Arai Nov 1990 ROSAT PSPC/WFC 0.09–2.0 Yes [1]

C/1990 N1 (Tsuchiya-Kiuchi ) Nov 1990 ROSAT PSPC/WFC 0.09–2.0 Yes [1]Jan 1991 0.09–2.0 Yes [1]

2P/Encke Nov 1993 EUVE DS 0.02–0.10 No [8]Jul 1997 EUVE Scanners 0.02–0.18 Yes [9]Jul 1997 ROSAT HRI/WFC 0.09–2.0 Yes [9]

19P/Borrelly Nov 1994 EUVE DS 0.02–0.10 Yes [8]

6P/d’Arrest Sep 1995 EUVE DS 0.02–0.10 Yes [8]

C/1996 B2 (Hyakutake) Mar 1996 ROSAT HRI /WFC 0.09–2.0 Yes [2]Mar 1996 EUVE DS 0.02–0.10 Yes [3]Mar 1996 ALEXIS 0.06–0.10 NoApr 1996 XTE PCA 2.0–10.0 NoJun 1996 ASCA 0.20–6.0 NoJun 1996 ROSAT HRI /WFC 0.09–2.0 Yes [4]Jul 1996 ROSAT HRI /WFC 0.09–2.0 Yes

Aug 1996 ROSAT HRI /WFC 0.09–2.0 YesSep 1996 ROSAT HRI /WFC 0.09–2.0 Yes

C/1995 O1 (Hale-Bopp) Apr 1996 ROSAT HRI/WFC 0.09–2.0 NoSep 1996 EUVE Scanners 0.02–0.10 Yes [5]Sep 1996 BeppoSAX 0.1–200 Yes [6]Sep 1996 ROSAT HRI/WFC 0.09–2.0 Yes? [7]Sep 1996 ASCA 0.20–6.0 No [5]Mar 1997 XTE PCA 2.0–10.0 NoOct 1997 ROSAT HRI /WFC 0.09–2.0 NoNov 1997 ROSAT HRI /WFC 0.09–2.0 NoNov 1997 EUVE DS 0.02–0.10 Yes [8]Feb 1998 ROSAT PSPC/WFC 0.09–2.0 No

C/1996 Q1 (Tabur) Sep 1996 ASCA 0.20–6.0 NoSep 1996 ROSAT HRI /WFC 0.09–2.0 Yes [1]Oct 1996 ROSAT HRI /WFC 0.09–2.0 Yes [1]

55P/Temple-Tuttle Jan 1998 EUVE Scanners 0.02–0.18 YesJan 1998 ASCA SIS 0.20–6.0 NoJan 1998 ROSAT HRI /WFC 0.09–2.0 YesFeb 1998 ROSAT HRI /WFC 0.09–2.0 Yes

103P/Hartley 2 Feb 1998 ROSAT PSPC/WFC 0.09–2.0 Yes

C/1998 U5 (LINEAR) Dec 1998 ROSAT PSPC/WFC 0.09–2.0 Yes

C/2000 S4 (LINEAR) Jul 2000 Chandra ACIS-S 0.2–10.0 Yes [10]Aug 2000 Chandra ACIS-S 0.2–10.0 Yes [10]

C/1999 T1 McNaught-Hartley Jan 2001 Chandra ACIS-S 0.2–10.0 Yes [11]Jan 2001 XMM-Newton 0.2–12.0 YesFeb 2001 FUSE 0.113–0.117 No [12]


2.1. Spatial Morphology

X-ray and EUV images of C/1996 B2 (Hyakutake) madeby the ROSAT and EUVE satellites look very similar (Lisseet al., 1996; Mumma et al., 1997) (Fig. 1). Except for im-ages of C/1990 N1 (Dennerl et al., 1997) and C/Hale-Bopp1995 O1 (Krasnopolsky et al., 1997), all EUV and X-rayimages of comets have exhibited similar spatial morpholo-gies. The emission is largely confined to the cometary comabetween the nucleus and the Sun; no emission is found inthe extended dust or plasma tails. The peak X-ray bright-ness gradually decreases with increasing cometocentric dis-tance r with a dependence of about r–1 (Krasnopolsky, 1997).The brightness merges with the soft X-ray backgroundemission (McCammon and Sanders, 1990; McCammon et

al., 2002) at distances that exceed 104 km for weakly ac-tive comets, and can exceed 106 km for the most luminous(Dennerl et al., 1997) (Fig. 2). The spatial extent for themost extended comets is independent of the rate of gas emis-sion from the comet. The region of peak emission is cres-cent-shaped with a brightness peak displaced toward theSun from the nucleus (Lisse et al., 1996, 1999). The dis-tance of this peak from the nucleus appears to increase withincreasing values of Q, and for Hyakutake was located atrpeak ≈ 2 × 104 km.

2.2. Luminosity

The observed X-ray luminosity, Lx, of C/1996 B2 (Hya-kutake) was 4 × 1015 ergs s–1 (Lisse et al., 1996) for an aper-

TABLE 1. (continued).

Comet Time Instrument Energy (keV) Detection Reference

C/2001 A2 (LINEAR) Jun 2001 Chandra HRC/LETG 0.2–2.0 NoJun 2001 XMM-Newton 0.2–12.0 Yes?Jul 2001 FUSE 0.113–0.117 No [12]

C/2000 WM1 (LINEAR) Dec 2001 Chandra ACIS/LETG 0.2–2.0 YesDec 2001 FUSE 0.113–0.117 Yes [12]Jan 2002 XMM-Newton 0.2–12.0 Yes [13]

C/2002 C1 (Ikeya-Zhang) Apr 2002 Chandra ACIS-S 0.2–10.0 Yes [13]May 2002 XMM-Newton 0.2–12.0 Yes? [13]

*The full energy range of the observing instrument is given.†This table summarizes published and unpublished (1) dedicated observations, whether successful or not; and (2) successful serendipitousobservations. The table is sorted according to time of first observation of the comet.

References: [1] Dennerl et al. (1997); [2] Lisse et al. (1996); [3] Mumma et al. (1997); [4] Lisse et al. (1997a); [5] Krasnopolsky etal. (1997); [6] Owens et al. (1998); [7] Lisse et al. (1997b); [8] Krasnopolsky et al. (2000); [9] Lisse et al. (1999); [10] Lisse et al.(2001); [11] Krasnopolsky et al. (2002); [12] Weaver et al. (2002); [13] K. Dennerl et al. (personal communication, 2003).

Fig. 1. Images of C/Hyakutake 1996 B2 on 26–28 March 1996 UT: (a) ROSAT HRI 0.1–2.0 keV X-ray; (b) ROSAT WFC 0.09–0.2 keV extreme ultraviolet; and (c) visible light, showing a coma and tail, with the X-ray emission contours superimposed. The Sun istoward the right, “+” marks the position of the nucleus, and the orbital motion of the comet is toward the lower right in each image.From Lisse et al. (1996).

634 Comets II

ture radius at the comet of 1.2 × 105 km. [Note that the pho-tometric luminosity depends on the energy bandpass andon the observational aperture at the comet. The quoted valueassumes a ROSAT photon emission rate of PX ≈ 1025 s–1

(0.1–0.6 keV), in comparison to Krasnopolsky et al.’s (2000)EUVE estimate of PEUV ≈ 7.5 × 1025 s–1 (0.07–0.18 keV and120,000 km aperture.] A positive correlation between opti-cal and X-ray luminosities was demonstrated using obser-vations of several comets having similar gas (QH2O) to dust[Afρ, following A’Hearn et al. (1984)] emission rate ratios(Fig. 3) (Lisse et al., 1997b, 1999, 2001; Dennerl et al.,1997; Mumma et al., 1997; Krasnopolsky et al., 2000). Lxcorrelates more strongly with the gas production rate Qgasthan it does with Lopt ~ Qdust ~ Afρ (Figs. 2 and 3). Particu-larly dusty comets, like Hale-Bopp, appear to have less X-rayemission than would be expected from their overall opticalluminosity Lopt. The peak X-ray surface brightness decreaseswith increasing heliocentric distance r, independent of Q(Dennerl et al., 1997), although the total luminosity appearsroughly independent of r. The maximum soft X-ray lumi-nosity observed for a comet to date is ~2 × 1016 erg s–1 forC/Levy at 0.2–0.5 keV (Dennerl et al., 1997) (Fig. 3).

2.3. Temporal Variation

Photometric lightcurves of the X-ray and EUV emissiontypically show a long-term baseline level with superimposed

Fig. 2. Spatial extent of the X-ray emission vs. the comet’s out-gassing rate. Plot of the gas production rate Qgas vs. radial dis-tance from the comet nucleus required to encircle 95% of the totalobserved cometary X-ray flux (triangles). Upper curve: Brokenpower law with radial extent ~Qgas

1.00 up to Qgas ~ 1029 mol s–1 and~106 km for higher values fits the imaging data well. Lower curve:Estimated radius of the bow shock for each observation, allowingfor variable cometary outgassing activity and heliocentric distance(boxes). X-ray emission has been found outside the bow shock forall comets except C/1996 B2 (Hyakutake) in March 1996.

Fig. 3. X-ray vs. optical luminosity plot for the eight detectedROSAT comets and the Chandra comets C/1999 S4 (LINEAR)and C/1999 T1 (McNaught-Hartley) observed at 1–3 AU. Groupsof equal emitted dust mass to emitted gas mass ratio (D/G), asmeasured in the optical by the ratio Afρ/QH2O, are also shown.For Encke and other “gassy”, optically faint comets, the resultingslope Lx/Lopt is roughly constant. Above Lopt ~ few × 1019 erg s–1,however, Lx appears to reach an asymptote of ~5 × 1016 erg s–1.A possible explanation is that the coma is collisionally thick tothe solar wind within the neutral coma radius of ~106 km at 1 AU(Lisse et al., 2001). It is also possible that the relatively largeamounts of dust in these comets, as noted from their increasingD/G ratio, may be somehow inhibiting the CXE process. Follow-ing Dennerl et al. (1997); copyright journal Science (1997).

impulsive spikes of a few hours’ duration, and maximumamplitude 3 to 4 times that of the baseline emission level(Lisse et al., 1996, 1999, 2001). Figure 4 demonstrates thestrong correlation found between the time histories of thesolar wind proton flux (a proxy for the solar wind minorion flux), the solar wind magnetic field intensity, and a com-et’s X-ray emission for the case of Comet 2P/Encke 1997(Lisse et al., 1999). Neugebauer et al. (2000) compared theROSAT and EUVE luminosity of C/1996 B2 (Hyakutake)with time histories of the solar wind proton flux, oxygenion flux, and solar X-ray flux, as measured by spacecraftresiding in the solar wind. They found the strongest corre-lation between the cometary emission and the solar windoxygen ion flux, a good correlation between the comet’semission and the solar wind proton flux, but no correlationbetween the cometary emission and the solar X-ray flux.

For the four comets for which extended X-ray light-curves were obtained during quiet Sun conditions, the timedelay between the solar wind proton flux and the comet’sX-ray impulse (Table 2) was well predicted by assuming asimple latitude-independent solar wind flow, a quadrupolesolar magnetic field, and propagation of the sector bound-aries radially at the speed of the solar wind and azimuth-


ally with period one-half the solar rotation period of 28 d(Lisse et al., 1997b, 1999; Neugebauer et al., 2000)

400 km/s ⋅ 86400 s/d14.7˚/d

longitudecomet – longitudeEarth (rcomet – rEarth)+

∆ttotal = ∆tCarrington rotation + ∆tradial =

2.4. Spectrum

Until 2001, all published cometary X-ray spectra hadvery low spectral energy resolution (∆E/E ~ 1 at 300–600 eV),and the best spectra were those obtained by ROSAT for C/1990 K1 (Levy) (Dennerl et al., 1997) and by BeppoSAX

for Comet C/1995 O1 (Hale-Bopp) (Owens et al., 1998).These observations were capable of showing that the spec-trum was very soft (characteristic thermal bremsstrahlungtemperature kT ~ 0.23 ± 0.04 keV) with intensity increasingtoward lower energy in the 0.01–0.60 keV energy range,and established upper limits to the contribution of the fluxfrom K-shell resonance fluorescence of carbon at 0.28 keVand oxygen at 0.53 keV. However, even in these “best” spec-tra, continuum emission could not be distinguished from amultiline spectrum. Nondetections of Comets C/Hyakutake,C/Tabur, C/Hale-Bopp, and 55P/Temple-Tuttle using theXTE PCA (2–30 keV) and ASCA SIS (0.6–4 keV) imagingspectrometers were consistent with an extremely soft spec-trum (Lisse et al., 1996, 1997b).

Higher-resolution spectra of cometary X-ray emissionhave just appeared in the literature. The Chandra X-ray Ob-servatory (CXO) detected soft X-ray spectra from CometC/1999 S4 (LINEAR) (Lisse et al., 2001) over an energyrange of 0.2–0.8 keV, using an energy resolution with a full-

Fig. 4. Temporal trends for Comet 2P/Encke 1997 on 4–9 July1997 UT. = ROSAT HRI lightcurve, 4–8 July 1997. = EUVEscanner Lexan B lightcurve 6–8 July 1997 UT, taken contempo-raneously with the HRI observations, and scaled by a factor of 1.2.All error bars are ±1σ. Also plotted are the WIND total magneticfield Btotal ( ), the SOHO CELIAS/SEM 1.0–500-Å solar X-rayflux ( ), and the SOHO CELIAS solar wind proton flux ( ).There is a strong correlation between the solar wind magneticfield/density and the comet’s emission. There is no direct corre-lation between outbursts of solar X-rays and the comet’s outbursts.After Lisse et al. (1997a).

TABLE 2. Predicted and observed lightcurve phase shifts using the latitude-independent model.

Time of Impulse ∆tlong ∆tradial ∆ttotal ∆tobservedComet (00:00H UT) (d) (d) (d) (d)

Hyakutake 27 Mar 1996 –0.23 0.032 –0.20 –0.24Hale-Bopp 11 Sep 1996 –4.60 5.9 1.30 +1.4Encke 7 Jul 1997 –0.26 0.093 –0.17 –0.1Temple-Tuttle 29 Jan 1998 –2.31 0.37 –1.94 –2.5

Time shifts assume solar wind velocity as measured near-Earth; positive time shifts = impulse happens at Earth first,comet next; negative time shifts = boundary hits comet first, Earth next.

Fig. 5. Chandra ACIS-S medium resolution CCD X-ray spec-trum for Comet C/1999 S4 (LINEAR). Soft X-ray spectrum ofC/1999 S4 (LINEAR) obtained by the Chandra X-ray Observa-tory (crosses) and a six-line best-fit “model” spectrum (solid line).The positions of several possible atomic lines are noted. Adaptedfrom Lisse et al. (2001).

636 Comets II

width half-maximum (FWHM) of ∆E = 0.11 keV (Fig. 5).The spectrum is dominated by line emission, not by con-tinuum. Using the CXO, a new spectrum of Comet C/1999T1 (McNaught-Hartley) (Krasnopolsky et al., 2002) showssimilar line-emission features. Line emission is also foundin XMM-Newton spectra of Comet C/1999 T1 (McNaught

Hartley) and, more recently, in CXO spectra of C/2001 WM1(LINEAR) and C/2002 Ikeya-Zhang (K. Dennerl et al. andC. M. Lisse et al., personal communication, 2003). AnXMM-Newton spectrum of C/2001 WM1 (LINEAR) showscharacteristic CXE X-ray signatures in unprecedented detail(K. Dennerl et al., personal communication, 2003). A re-

Fig. 6. EUVE observations of line emission from C/1996 B2 (Hyakutake), following Krasnopolsky and Mumma (2001). (a) MW(middle wavelength) 0.034–0.073 keV spectrum on March 23, 1996. (b) LW (long-wavelength) 0.018–0.04 keV. The extreme ultra-violet spectra are clearly dominated by line emission. The best agreement with CXE model predictions are for the O4+, C4+, and Ne7+

lines. (c) FUSE observations of three comets, with a marginal detection of the CXE OVI line in C/2001 WM1 (LINEAR) at 1032 Å(following Weaver et al., 2002). The nondetections in Comets C/2001 A2 (LINEAR) and C/1999 T1 (McNaught-Hartley) and the mar-ginal detection in C/2001 WM1 (LINEAR) are consistent with CXE predictions for the luminosity of these lines.


analysis of archival EUVE Deep Survey spectrometer spec-tra (Krasnopolsky and Mumma, 2001) suggests EUV line-emission features from Comet C/1996 B2 (Hyakutake)(Figs. 6a,b). Recent FUSE observations (Weaver et al.,2002) also indicate the presence of possible O VI 1032-Åemission lines in far-UV spectra of C/2001 WM1 (LIN-EAR) (Fig. 6c).

3. PROPOSED X-RAY MECHANISMS

A large number of explanations for cometary X-rayswere suggested following the discovery paper in 1996.These included thermal bremsstrahlung (German for “brak-ing radiation”) emission due to solar wind electron colli-sions with neutral gas and dust in the coma (Bingham et al.,1997; Dawson et al., 1997; Northrop, 1997; Northrop et al.,1997; Uchida et al., 1998; Shapiro et al., 1999), microdustcollisions (Ibadov, 1990; Ip and Chow, 1997), K-shell ion-ization of neutrals by electron impact (Krasnopolsky, 1997),scattering or fluorescence of solar X-rays by cometary gas orby small dust grains (Lisse et al., 1996; Wickramasinghe andHoyle, 1996; Owens et al., 1998), and by charge exchangebetween highly ionized solar wind ions and neutral speciesin the cometary coma (CXE) (Cravens, 1997a; Häberli etal., 1997; Wegmann et al., 1998; Kharchenko and Dalgarno,2000; Kharchenko et al., 2003; Schwadron and Cravens,2000). In the thermal bremsstrahlung mechanism, fast elec-trons are deflected in collisions with charged targets, suchas the nuclei of atoms, and emit continuum radiation. Elec-tron energies in excess of 100 eV (T > 106 K) are needed forthe production of X-ray photons. In the K-shell mechanism,a fast electron collision removes an orbital electron froman inner shell of the target atom. Early evaluation of thesevarious mechanisms (Dennerl et al., 1997; Krasnopolsky,1997; Lisse et al., 1999) favored only three of them: theCXE mechanism, thermal bremsstrahlung, and scattering ofsolar radiation from very small (i.e., attogram; 1 attogram =10–19 g) dust grains.

A significant problem with mechanisms involving solarwind electrons (i.e., bremsstrahlung or K-shell ionization)is that the predicted emission luminosities are too small byfactors of 100–1000 compared to observations. The flux ofhigh-energy solar wind electrons near comets is too low(Krasnopolsky, 1997; Cravens, 2000b, 2002a). Furthermore,X-ray emission has been observed out to great distancesfrom the nucleus, beyond the bow shock (Fig. 2), and thethermal energy of unshocked solar wind electrons at thesedistances is about 10 eV. No emission has ever been foundto be associated with the plasma tail of a comet, which hassimilar plasma densities and temperatures. Finally, the new,high-resolution spectra demonstrating multiple atomic linesare inconsistent with a continuum-type mechanism or amechanism producing only a couple of K-shell lines as theprimary source of cometary X-rays. Lisse et al. (2001) triedseveral thermal bremsstrahlung continuum model fits to theC/1999 S4 spectrum, and Krasnopolsky and Mumma (2001)tried the same for the C/1996 B2 (Hyakutake) spectrum, butneither was successful.

Mechanisms based on dust grains also have a numberof problems. It has been known since 1996 that Rayleighscattering of solar X-ray radiation from ordinary cometarydust grains (i.e., about 1 µm in size) cannot produce theobserved luminosities — the cross section for this processis too small (Lisse et al., 1996). A potential solution to thisproblem is to invoke a population of very small, attogram(10–19 g) grains with radii on the order of the wavelengthsof the observed X-ray radiation, 10–100 Å, which can reso-nantly scatter the incident X-ray radiation. The abundanceof such attogram dust grains is not well understood in com-ets, as they are undetectable by remote optical observations;however, there were reports from the VEGA Halley flybyof a detection of an attogram dust component using thePUMA dust monitor (Vaisberg et al., 1987; Sagdeev et al.,1990). However, the statistical studies of the properties ofseveral comets (Figs. 2 and 3) demonstrate that X-ray emis-sion varies with a comet’s gas production rate and not thedust production rate (Dennerl et al., 1997; Lisse et al., 1999,2001). Furthermore, the cometary X-ray lightcurves (Lisseet al., 1996, 1999, 2001; Neugebauer et al., 2000) corre-late with the solar wind ion flux and not with solar X-rayintensity. Finally, dust-scattering mechanisms cannot ac-count for the pronounced lines seen in the new high-reso-lution spectra — emission resulting from dust scattering ofsolar X-rays should mimic the Sun’s X-ray spectral con-tinuum, similar to what is observed in the terrestrial atmos-phere for Rayleigh scattering of sunlight (Krasnopolsky,1997).

The CXE mechanism requires that the observed X-rayemission is driven by the solar wind flux and that the bulkof the observed X-ray emission be in lines. Localization ofthe emission to the sunward half of the coma, a solar windflux-like time dependence, and a line-emission-dominatedspectral signature of the observed emission all stronglypoint to the solar wind charge exchange mechanism asbeing responsible for cometary X-rays.

4. SOLAR WIND CHARGE EXCHANGEX-RAY MECHANISM

The solar wind is a highly ionized but tenuous gas (i.e.,a plasma) (Cravens, 1997b). At its source in the solar co-rona, the million-degree gas is relatively dense and in colli-sional equilibrium, but its density drops within a few solarradii into a freeflow regime wherein collisions are infre-quent. Both the solar wind and corona have “solar” com-position — 92% hydrogen by volume, 8% helium, and 0.1%heavier elements. The heavier, “minor ion” species arehighly charged (e.g., oxygen in the form of hydrogen-likeO7+ or helium-like O6+ ions, N6+/N5+, C5+/C4+, Ne8+, Si9+,Fe12+, etc.) due to the high coronal temperatures (Bame,1972; Bocshler, 1987; Neugebauer et al., 2000).

The solar wind flow starts out slowly in the corona butbecomes supersonic at a distance of few solar radii (Parker,1963; Cravens, 1997b). The gas cools as it expands, fall-ing from T ≈ 106 K down to about 105 K at 1 AU. The aver-age properties of the solar wind at 1 AU are proton number

638 Comets II

density ≈7 cm–3, speed ≈450 km s–1, temperature ≈105 K,magnetic field strength ≈5 nT, and Mach number ≈8 (Hund-hausen et al., 1968). However, the composition and chargestate distribution far from the Sun are “frozen in” at coro-nal values due to the low collision frequency outside thecorona. The solar wind contains structure, such as slow(400 km s–1) and fast (700 km s–1) streams, which can bemapped back to the Sun. The solar wind “terminates” in ashock called the heliopause, where the ram pressure of thestreaming solar wind has fallen to that of the instellar ma-terial (ISM) gas (Suess, 1990). The region of space that con-tains plasma of solar origin, from the corona to the helio-pause at ~100 AU, is called the heliosphere. A very smallpart of the solar wind interacts with the planets and comets;the bulk of the wind interacts with neutral ISM gas in theheliosphere and neutral and ionized ISM material at theheliopause.

As the solar wind streams into a comet’s atmosphere,cometary ion species produced from solar UV photoioniza-tion of neutral coma gas species are added to the flow as“pick-up ions.” The resulting mass addition slows down thesolar wind due to momentum conservation and a bow shockforms upwind of the comet (Galeev, 1991; Szegö et al.,2000) (Fig. 7). The flow changes from supersonic to sub-sonic across the shock, and the magnetic field strength in-creases by a factor of ~5. Closer to the nucleus, where thecometary gas density is higher and collisions more frequent,the flow almost completely stagnates (Flammer, 1991). Theouter boundary of this stagnation region is often called thecometopause (cf. review by Cravens, 1991). The observedX-ray brightness peak resides within this boundary. Mag-

netic field lines pile up into a “magnetic barrier” in this stag-nation region and drape around the head of the comet formingthe plasma tail in the downwind direction (Brandt, 1982).

From experimental and theoretical work in atomic andmolecular physics it is found that solar wind minor ionsreadily undergo charge transfer (or exchange) reactions(Phaneuf et al., 1982; Dijkkamp et al., 1985; Gilbody, 1986;Janev et al., 1988; Wu et al., 1988) when they are within~1 nm of a neutral atomic species

Aq+ + B → A(q – 1)+* + B+ (1)

where A denotes the solar wind projectile ion (e.g., O, C,Si . . . ), q is the projectile charge (e.g., q = 5, 6, 7) and Bdenotes the neutral target species (e.g., H2O, OH, CO, O,H . . . for cometary comae) (Fig. 8). The cross section forthis process is large, on the order of 10–15 cm2, about 1 or-der of magnitude larger than the hardsphere collisional crosssection.

The product ion deexcites by emitting one or more pho-tons (A(q – 1)+* → A(q – 1)+* + hν, where hν represents a pho-ton). It is the characteristic radiation of the product ion thatis measured with astronomical X-ray instrumentation, andso one labels the radiation detected by the charge state ofthe final ion. The deexcitation usually takes place via a cas-cade through intermediate states rather than in one step tothe ground state. For large enough values of q, the deexci-tation transitions lead to the emission of X-ray photons. Forspecies and charge states relevant to comets, the principalquantum number of the ion A(q – 1)+ is about n = 4 or 5

Fig. 7. Spatial schematic of the solar wind-comet interaction.The relative locations of the bow shock, the magnetic barrier, andthe tail are shown (not to scale). The Sun is toward the left. Alsorepresented is a charge transfer collision between a heavy solarwind ion and a cometary neutral water molecule, followed by theemission of an X-ray photon. After Cravens (2002b); copyrightjournal Science (2002).

Fig. 8. Energy level diagram for a CXE process. Electron po-tential energy (atomic units, a.u.) vs. distance from the target atomnucleus (assumed here to be an H atom) for a charge transferreaction involving a projectile ion Be4+. The internuclear distancechosen is 10 Bohr radii or 5.29 × 10–10 m, the curve-crossingdistance for the n = 3 ion final state. The target energy level (andbinding energy Eb) and product ion (Be3+) energy levels are shownin units of hartrees (1 hartree = 27.2 eV). A possible cascadingpathway for the deexcitation by photon emission is shown. AfterCravens (2002b); copyright journal Science (2002).


(Ryufuku et al., 1980; Mann et al., 1981). The cross sectionfor charge exchange between a high charge state solar windion and an ionized coma gas species is negligible in com-parison, due to the effects of Coulomb repulsion betweenthe two reactants. Once a coma neutral atom is ionized,either by CXE processes or solar UV flux, the CXE mecha-nism is no longer an important energy transfer process.

4.1. Charge Exchange Morphology

Numerical simulations of the solar wind interaction withHyakutake including CXE have been used to generate X-rayimages. A global magnetohydrodynamic (MHD) model(Häberli et al., 1997) and a hydrodynamic model (Wegmannet al., 1998) were used to predict solar wind speeds anddensities and the X-ray emission around a comet. The simu-lated X-ray images are similar to the observed images,which is relatively unsurprising, as any dissipative solarwind–coma process with total optical depth near unitywould create the observed morphology (Fig. 1). The emis-sion is found to lie in the sunward hemisphere of the neu-tral coma, varying from collisionally thin to collisionallythick as the solar wind approaches the nucleus. Emissionis predicted to be in the soft X-ray, UV, and optical wave-lengths, with the softest photons emitted closest to the nu-cleus. However, the spatial models to date have includedonly highly simplified models of the CXE deexcitation cas-cade, as compared to the detailed spectral models of the glo-bal behavior discussed below.

As an example of the potential of studying the behaviorof the solar wind inside the coma using CXE reactions, weconsider the dissimilar morphologies of the extended Ly-man α comae and the X-ray emitting regions of comets (cf.Keller, 1973; Festou et al., 1979; Combi et al., 2000). TheCXE mechanism should not only transfer electrons fromcometary neutrals to solar wind minor ions, but to the so-lar wind majority ions H+ and He2+ as well. These ions areroughly 1000 times more abundant than the X-ray activehighly ionized minor ions. The prompt photon-emissionenergy produced from CXE by He2+ ions produces at leastthree times the energy released by all the minor ions com-bined (D. Shemansky, personal communication, 2003).Further, the neutral atoms produced are capable of scatter-ing emission from the Sun. At luminosities of ~1016 erg s–1,and production rates of ~1027 s–1, the HI created by CXEshould be detectable in Lyman α comet images. The factthat it is not is puzzling. A possible solution is that the neu-tral hydrogen atoms produced by CXE retain relatively largevelocities with respect to the Sun, i.e., the solar wind is notappreciably slowed at the cometary bow shock. This largeremnant velocity redshifts the CXE-produced neutral hydro-gen with respect to the peak of the solar Lyman α emission,so that fluorescence from these atoms is greatly reduced inefficiency vs. H atoms produced from dissolution of comet-ary water group species. Recently Raymond et al. (2002)have reported the effects of highly redshifted neutral hydro-gen created by CXE emission in SOHO UVCS measure-ments of 2P/Encke during its 1997 apparition.

4.2. Charge Exchange Luminosityand Temporal Variation

To first order, the CXE local X-ray power density Px canbe estimated assuming only one CXE collision per solarwind ion per coma passage. This approximation yields theexpression

Px = αnswuswnn (2)

where nsw usw, and nn are the solar wind proton density, solarwind speed, and neutral target density respectively (Cra-vens, 1997a, 2002a). All the “atomic and molecular details”as well as the solar wind heavy ion fraction fh are combinedinto the parameter α, given by α ≈ fh⟨sct⟩Eave, where ⟨sct⟩ is anaverage CXE cross section for all species and charge states,and Eave an average photon energy. A simple sphericallysymmetric approximation to the neutral density in the comais given by nn = Q/[4πunr2], for r less than the ionizationscale length R = unt, where t ≈ 106 s is the ionization life-time [for 1 AU (Schleicher and A’Hearn, 1988)] and un ≈1 km s–1 the neutral gas outflow speed. Integration of Pxover the volume of the neutral coma yields an X-ray lumi-nosity typically within a factor of 2–3 of the observed lu-minosity (Cravens, 1997a, 2000a; Lisse et al., 2001). Theobserved luminosity is a function of both the solar windflux density and the com-etary neutral gas production rateup to the limit of 100% charge exchange efficiency of allsolar wind minor ions within an ionization scale length of106 km. The maximum expected X-ray luminosity at 1 AUand 0.2–0.5 keV is ~1016 erg s–1 (Fig. 3). Temporal varia-tions of the solar wind flux directly translate into time varia-tions of the X-ray emission (Fig. 4).

4.3. Charge Exchange Spectra

Model CXE spectra are in good agreement with the low-resolution X-ray spectra of cometary X-ray emission, andthe line centers of the high-resolution spectra have beensuccessfully predicted using CXE theory (Figs. 5 and 6)(Lisse et al., 1999, 2001; Krasnopolsky and Mumma, 2001;Weaver et al., 2002). The application of the CXE model tocomets has entailed a number of approaches to date. Somework has included only a few solar wind species but used acareful cascading scheme (Häberli et al., 1997). Other ap-proaches have used a simple cascading scheme and simplecollision cross sections, but included a large number ofsolar wind ions and charge states (Wegmann et al., 1998;Schwadron and Cravens, 2000). Kharchenko and colleagues(Kharchenko and Dalgarno, 2000; Kharchenko et al., 2003)have treated the atomic cascading process more carefullythan other modelers, although the spatial structure of thesolar wind–cometary neutral interaction in their model washighly simplified. They predicted the existence of a largenumber of atomic lines, including O5+ (1s25d → 1s22p) at106.5 eV, C4+ (1s2s → 1s2) at 298.9 eV, C5+ (2p → 1s) at367.3 eV, C5+ (4p → 1s) at 459.2 eV, and O6+ (1s2p → 1s2)at 568.4 eV. At least some of these lines appear in the best

640 Comets II

cometary X-ray spectra to date, the CXO ACIS-S spectra ofC/1999 S4 (LINEAR) and C/1999 T1 (McNaught-Hartley).The peak measured near 0.56 keV is certainly the combi-nation of three closely spaced helium-like O6+ (1s2p and1s2s → 1s2) transitions (which comes from CXE of solarwind O7+) and the line located at 0.32 keV is due to helium-like C4+ (1s2p and 1s2s → 1s2). Similar identifications havebeen made for the EUVE spectrum of C/1996 B2 (Hyaku-take) (Krasnopolsky and Mumma, 2001). The spectral prob-lem is still far from totally solved, however. Careful com-parisons and calculations needed to interpret the subtletiesof the high-resolution spectral observations, including therole of collisions after charge transfer, solar wind ion–dustinteractions, and the exact species present at each point inthe coma, are only now starting to be done (Krasnopolsky etal., 2002; Kellett et al., 2003).

5. THE FUTURE OF COMETARYX-RAY EMISSION STUDIES

The discovery that comets are X-ray sources can nowbe explained by charge exchange reactions of highly chargedsolar wind ions with neutral atoms and molecules residingin the cometary coma. The energy required to power thisemission originates in the hot solar corona and is stored aspotential energy in highly stripped solar wind ions (Cra-vens, 2000a, 2002b; Dennerl, 1999). Charge exchange reac-tions with neutral species, like cometary coma neutral atomsand molecules, release this potential energy in the form ofX-ray and UV radiation. This simple model can explain thegross features of the observed crescent-shaped emissionwith its sunward displaced peak, the maximum spatial ex-tent of the emission of ~106 km (Figs. 1 and 2), the maxi-mum observed luminosity of ~1016 erg s–1 (Fig. 3), and spec-tra dominated by line emission (Fig. 4).

5.1. Plasma-Neutral Interactions in theCometary Coma

However, a more careful treatment of the CXE mecha-nism is needed to fully understand the phenomenon of com-etary X-ray emission. One complication is that the multipleCXE collisions take place in regions close to the nucleuswhere the target density is high (the so-called “collisionallythick” case). The charge state for the initially hydrogen-likeor helium-like solar wind minor ion is reduced by one dur-ing each CXE collision, ultimately leading to its conversionto a neutral atom. When the ion’s charge state becomes toolow, X-ray photons are no longer emitted. When the numberof neutrals in a volume of space becomes too low, due tocoma expansion and ionization of coma gas molecules bysolar UV radiation, X-ray photons are no longer detectable.Another complication is that spectral differences are ex-pected for slow and fast solar wind streams, with the slowsolar wind, with its higher coronal freeze-in temperature,producing a harder spectrum than does the fast solar wind(Schwadron and Cravens, 2000).

Further theoretical progress will require the integrationof several ingredients into a single model: (1) accurate solarwind composition for a range of solar wind types; (2) asuitable MHD model of the solar wind interaction with thecoma, in order to accurately predict the densities of ionsand neutrals in equation (2); and (3) a more detailed under-standing of the atomic processes, in order to improve ourunderstanding of the parameter α in equation (2). Successfor the first point requires new and improved measurementsof the solar wind throughout the heliosphere, and/or largenumber statistical studies of cometary X-ray emissionthroughout the heliosphere; the second point requires im-proved MHD codes on modern supercomputers; and toachieve the third point, additional laboratory measurementsof state-specific CXE cross sections will be required. Thefirst and second points are being actively pursued by as-tronomers and modelers in the field. New laboratory workis now being undertaken (Beiersdorfer et al., 2000, 2001;Greenwood et al., 2000; Hasan et al., 2001) to measureCXE cross sections for cometary target species such as H2Oat collision energies relevant to the solar wind (a few keV/amu). Recent measurements have indicated, for example,that multiple as well as single-electron CXE makes a con-tribution to the X-ray emission (Hasan et al., 2001; Green-wood et al., 2001; Gao and Kwong, 2002). Given detailedMHD models of the solar wind passing through the comaand accurate cross sections for the CXE process, we willbe able to map out the density of solar wind minor ions inthe coma.

5.2. Remote Sensing of the Solar Wind

Driven by the solar wind, cometary X-rays provide anobservable link between the solar corona, where the solarwind originates, and the solar wind where the comet resides.Once we have understood the CXE mechanism’s behaviorin cometary comae in sufficient detail, we will be able touse comets as probes to measure the solar wind through-out the heliosphere. This will be especially useful in moni-toring the solar wind in places hard to reach with space-craft — over the solar poles, at large distances above andbelow the ecliptic plane, and at heliocentric distances greaterthan a few AU (Lisse et al., 1996, 2001; Krasnopolsky etal., 2000). For example, approximately one-third of the ob-served soft X-ray emission is found in the 530–700-eVoxygen O7+ and O6+ lines; observing photons of this energyallows studies of the oxygen ion charge ratio of the solarwind, which is predicted to vary significantly between theslow and fast solar winds (Neugebauer et al., 2000; Schwad-ron and Cravens, 2000; Kharchenko and Dalgarno, 2001).

5.3. Emission from Other Planetary Systems

The CXE mechanism operates wherever the solar wind(or any highly ionized plasma) interacts with a substantialquantity of neutral gas. Motivated in part by the discoveryof the new class of cometary X-ray emitters, further meas-


urements of potential solar system sources (Holmström et al.,2001; Cravens and Maurellis, 2001) of X-ray emission havebeen undertaken recently. X-rays have now been observedfrom Venus as solar X-rays fluorescently scattered by thethick (compared to comets) venusian atmosphere (Dennerlet al., 2002), and from Mars as ~90% fluorescently scat-tered solar X-rays from the thick martian atmosphere and~10% CXE-derived X-rays (Dennerl, 2002). The solar windis known to approach very closely to these planets due totheir weak intrinsic magnetic fields, which cannot act as ef-fective obstacles to the external flow, allowing for CXE pro-cesses to occur. The small spatial extent and high densityof the gravitationally bound planetary atmospheres providesa higher X-ray luminosity due to scattering of solar X-raysthan from CXE. X-rays have now been detected at Io andEuropa and in the Io flux torus using Chandra, althoughthe exact mechanism of emission is still unknown (Elsneret al., 2002). Another neutral gas system from which X-rayemission has been predicted is the terrestrial hydrogen geo-corona (Cox, 1998; Cravens, 2000a; Dennerl, 1999; Frey-berg, 1998; Robertson and Cravens, 2003), although no de-tection has yet been made.

X-ray emission from the heliosphere is also expectedfrom the interaction of the solar wind with the interstellarneutral gas (mainly HI and HeI) that streams into the solarsystem (Cox, 1998; Cravens, 2000a,b). Cravens (2000b)demonstrated that roughly one-half of the observed 0.25-keV X-ray diffuse background can be due to this process.Solar and Heliospheric Observatory (SOHO) observationsof neutral hydrogen Lyman α emission show a clear asym-metry in the ISM flow direction, with a clear deficit ofneutral hydrogen in the downstream direction of the incom-ing neutral ISM gas, most likely created by CXE ionizationof the ISM as it transits the heliosphere. Cravens et al.(2001) have shown a strong correlation between the solarwind flux density and the ROSAT “long-term enhance-ments,” systematic variations in the soft X-ray backgroundof the ROSAT X-ray detectors. Photometric imaging obser-vations of the lunar nightside by Chandra made in Septem-ber 2001 do not show any lunar nightside emission abovea CXE background. The soft X-ray emission detected fromthe darkside of the Moon, using ROSAT, would appear tobe due not to electrons spiraling from the sunward to thedark hemisphere, as proposed by Schmitt et al. (1991), butinstead be due to CXE in the column of solar wind betweenEarth and the Moon. The analogous process applied to otherstars has been suggested as a means of detecting stellarwinds (Lisse, 2002a,b; Wargelin and Drake, 2001).

5.4. Soft X-Ray Background

Heliospheric and geocoronal X-ray emission have beensuggested (Cravens, 2000b, 2002b; Dennerl, 1999; Lisse etal., 2001) to make significant contributions to the observedsoft X-ray background [previously attributed entirely to hotinterstellar gas (Snowden et al., 1990, 1998; McCammon etal., 2002)]. This supposition is supported by the positive cor-

relations that have been found between measured solar windfluxes and measured X-ray background intensities (Cravenset al., 2001; Robertson et al., 2001). Recent large angularscale measurements of the diffuse soft X-ray background ina 100-s rocket flight by McCammon et al. (2002) and in theChandra background toward MBM12 (R. Edgar et al., per-sonal communication, 2001) show a clear peak at 560 eV, asexpected for CXE-driven emission (Fig. 4). A significant frac-tion of the line emission observed in the soft X-ray back-ground, e.g., the 560-eV emission, is probably due to CXEemission in our solar system and around other stars.

REFERENCES

A’Hearn M. F., Schleicher D. G., Millis R. L., Feldman P. D., andThompson D. T. (1984) Comet Bowell 1980b. Astron. J., 89,579–591.

Bame S. J. (1972) Spacecraft observations of the solar wind com-position. In Solar Wind (C. P. Sonnett et al., eds.), p. 529.NASA SP-308, Washington, DC.

Beiersdorfer P., Olson R. E., Brown G. V., Chen H., Harris C. L.,Neill P. A., Schweikhard L., Utter S. B., and Widmann K.(2000) X-ray emission following low-energy charge exchangecollisions of highly charged ion. Phys. Rev. Lett., 85, 5090–5093.

Beiersdorfer P., Chen H., Lisse C. M., Olson R., and Utter S. B.(2001) Laboratory simulation of cometary x-ray emission.Astrophys. J. Lett., 549, L147.

Bingham R., Dawson J. M., Shapiro V. D., Mendis D. A., andKellett B. J. (1997) Generation of X-rays from C/Hyakutake1996B2. Science, 275, 49–51.

Bochsler P. (1987) Solar wind ion composition. Phys. Scripta,T18, 55–60.

Brandt J. C. (1982) Observations and dynamics of plasma tails.In Comets (L. L. Wilkening, ed.), pp. 519–537. Univ. of Ari-zona, Tucson.

Cioffi D. F. (1990) Supernova remnants as probes of the inter-stellar medium. In Physical Processes in Hot Cosmic Plasma(W. Brinkmann et al., eds.), p. 17. Kluwer, Dordrecht.

Combi M. R., Reinard A. A., Bertaux J.-L., Quemerais E., andMakinen T. (2000) SOHO/SWAN observations of the structureand evolution of the hydrogen Lyman-α coma of Comet Hale-Bopp (1995 O1). Icarus, 144, 191–202.

Cox D. P. (1998) Modeling the local bubble. In Lecture Notes inPhysics, Vol. 506, pp. 121–131. Springer-Verlag, Berlin.

Cravens T. E. (1991) Plasma processes in the inner coma. InComets in the Post-Halley Era (R. L. Newburn Jr. et al., eds.),pp. 1211–1255. Kluwer, Dordrecht.

Cravens T. E. (1997a) Comet Hyakutake X-ray source: Chargetransfer of solar wind heavy ions. Geophys Res. Lett., 24, 105–109.

Cravens T. E. (1997b) Physics of Solar System Plasmas, p. 477.Cambridge Univ., Cambridge.

Cravens T. E. (2000a) X-ray emission from comets and planets.Adv. Space Res., 26, 1443–1451.

Cravens T. E. (2000b) Heliospheric X-ray emission associated withcharge transfer of the solar wind with interstellar neutrals.Astrophys. J. Lett., 532, L153–L156.

Cravens T. E. (2002a) X-ray emission in the solar system. InAtomic Processes in Plasmas: 13th APS Topical Conferenceon Atomic Processes in Plasmas (D. R. Schultz et al., eds.),

642 Comets II

pp. 173–181. AIP Conference Proceedings 635, American In-stitute of Physics, Melville, New York.

Cravens T. E. (2002b) X-ray emission from comets. Science, 296,1042–1046.

Cravens T. E. and Maurellis A. N. (2001) X-ray emission fromscattering and fluorescence of solar X-rays at Venus and Mars.Geophys. Res. Lett., 28, 3043.

Cravens T. E., Robertson I. P., and Snowden S. L. (2001) Tem-poral variations of geocoronal and heliospheric x-ray emissionassociated with the solar wind interaction with neutrals. J. Geo-phys. Res., 106, 24883–24892.

Dawson J. M., Bingham R., and Shapiro V. D. (1997) X-rays fromComet Hyakutake. Plasma Phys. Control. Fusion, 39, A185–A193.

Dennerl K. (1999) X-ray emission from comets. In Atomic PhysicsVol. 16: Sixteenth International Conference on Atomic Physics(W. E. Baylis and G. W. F. Drake, eds.), pp. 361–376. AIPConference Proceedings 477, Springer-Verlag, Berlin.

Dennerl K. (2002) Discovery of X-rays from Mars with Chandra.Astron. Astrophys., 394, 1119–1128.

Dennerl K., Englhauser J., and Trümper J. (1997) X-ray emissionsfrom comets detected in the Röntgen X-ray satellite all-skysurvey. Science, 277, 1625–1629.

Dennerl K., Burwitz V., Englhauser J., Lisse C. M., and Wolk S.(2002) Discovery of X-rays from Venus with Chandra. Astron.Astrophys., 386, 319–330.

Dijkkamp D., Gordeev Y. S., Brazuk A., Drentje A. G., andde Heer F. J. (1985) Selective single-electron capture into (n, l)subshells in slow collisions of C6+, N6+, O6+ and Ne6+ with He,H2 and Ar. J. Phys. B. Atom. Mol. Phys., 18, 737–756.

Dorren J. D., Guedel M., and Guinan E. F. (1995) X-ray emissionfrom the Sun in its youth and old age. Astrophys. J., 448, 431–436.

Elsner R. F. and 15 colleagues (2002) Discovery of soft X-rayemission from Io, Europa, and the Io plasma torus. Astrophys.J., 572, 1077–1082.

Feigelson E. D. (1982) X-ray emission from young stars and im-plications for the early solar system. Icarus, 51, 155–163.

Festou M., Jenkins E. B., Barker E. S., Upson W. L., Drake J. F.,Keller H. U., and Bertaux J. L. (1979) Lyman-alpha observa-tions of comet Kobayashi-Berger-Milon (1975 IX) with Coper-nicus. Astrophys. J., 232, 318–328.

Flammer K. R. (1991) The global interaction of comets with thesolar wind. In Comets in the Post-Halley Era (R. L. NewburnJr. et al., eds.), pp. 1125–1144. Kluwer, Dordecht.

Foukal P. (1990) Solar Astrophysics. Wiley, New York. 475 pp.Freyberg M. J. (1998) On the zero-level of the soft X-ray back-

ground. In The Local Bubble and Beyond (D. Breitschwerdtet al., eds.), pp. 113–116. Lecture Notes in Physics, Vol. 506,Springer-Verlag, Berlin.

Galeev A. A. (1991) Plasma processes in the outer coma. In Com-ets in the Post-Halley Era (R. L. Newburn Jr. et al., eds.),pp. 1145–1169. Kluwer, Dordrecht.

Gao H. and Kwong V. H. S. (2002) Charge transfer of O+5 andO+4 with CO at keV energies. Astrophys. J., 567, 1272–1275.

Gilbody H. B. (1986) Measurements of charge transfer and ion-ization in collisions involving hydrogen atoms. Adv. Atom. Mol.Phys., 22, 143–195.

Greenwood J., Williams I. D., Smith S. J., and Chutjian A. (2000)Measurement of charge exchange and X-ray emission crosssections for solar wind-comet interactions. Astrophys. J. Lett.,533, L175–L178.

Greenwood J. B., Williams I. D., Smith S. J., and Chutjian A.

(2001) Experimental investigation of the processes determin-ing X-ray emission intensities from charge-exchange collisions.Phys. Rev. A, 63, 627071–627079.

Häberli R. M., Gombosi T. I., deZeeuw D. L., Combi M. R., andPowell K. G. (1997) Modeling of cometary X-rays caused bysolar wind minor ions. Science, 276, 939–942.

Hasan A. A., Eissa F., Ali R., Schultz D. R., and Stancil P. C.(2001) State-selective charge transfer studies relevant to solarwind-comet interactions. Astrophys. J. Lett., 560, L201–L205.

Holmström M. S., Barabash S., and Kallio E. (2001) X-ray imag-ing of the solar wind-Mars interaction. Geophys. Res. Lett., 28,1287.

Hundhausen A. J., Gilbert H. E., and Bame S. J. (1968) Ioniza-tion state of the interplanetary plasma. J. Geophys. Res., 73,5485–5493.

Ibadov S. (1990) On the efficiency of X-ray generation in impactsof cometary and zodiacal dust particles. Icarus, 86, 283–288.

Ip W.-H. and Chow V. W. (1997) NOTE: On hypervelocity impactphenomena of microdust and nano X-ray flares in cometarycomae. Icarus, 130, 217–221.

Janev R. K., Phaneuf R. A., and Hunter H. T. (1988) Recom-mended cross sections for electron capture and ionization incollisions of Cq+ and Oq+ ions with H, He and H2. Atomic Dataand Nuclear Data Tables, 40, 249.

Keller H. U. (1973) Lyman-alpha radiation in the hydrogen atmos-pheres of comets: A model with multiple scattering. Astron.Astrophys., 23, 269–280.

Kellett B. J., Bingham R., Lisse C. M., Torney M., Summers H. P.,and Shapiro V. D. (2003) A detailed model of the X-ray emis-sion from comets. In Plasma Physics: 11th International Con-gress on Plasma Physics (ICPP2002) (I. S. Falconer et al., eds.),pp. 704–707. AIP Conference Proceedings 669, American In-stitute of Physics, Melville, New York.

Kharchenko V. and Dalgarno A. (2000) Spectra of cometary X raysinduced by solar wind ions. J. Geophys. Res., 105, 18351–18360.

Kharchenko V. and Dalgarno A. (2001) Variability of cometaryX-ray emission induced by solar wind ions. Astrophys. J. Lett.,554, L99–L102.

Kharchenko V., Rigazio M., Dalgarno A., and Krasnopolsky V. A.(2003) Charge abundances of the solar wind ions inferred fromcometary X-ray spectra. Astrophys. J. Lett., 585, L73–L75.

Krasnopolsky V. A. (1997) On the nature of soft X-ray radiationin comets. Icarus, 128, 368–385.

Krasnopolsky V. A. and Mumma M. J. (2001) Spectroscopy ofComet Hyakutake at 80–700 Å: First detection of solar windcharge transfer emissions. Astrophys. J., 549, 629–634.

Krasnopolsky V. A., Mumma M. J., Abbott M., Flynn B. C.,Meech K. J., Yeomans D. K., Feldman P. D., and CosmoviciC. B. (1997) Discovery of soft X-rays from Comet Hale-Boppusing EUVE. Science, 277, 1488–1491.

Krasnopolsky V. A., Mumma M. J., and Abbott M. J. (2000) EUVEsearch for X-rays from Comets Encke, Mueller (C/1993 A1),Borrelly, and postperihelion Hale-Bopp. Icarus, 146, 152–160.

Krasnopolsky V. A., Christian D. J., Kharchenko V., Dalgarno A.,Wolk S. J., Lisse C. M., and Stern S. A. (2002) X-ray emissionfrom Comet McNaught-Hartley (C/1999 T1). Icarus, 160, 437–447.

Lisse C. M. (2002a) Cometary X-ray emission — the view afterthe first Chandra observation of a comet. In The High EnergyUniverse at Sharp Focus: Chandra Science (E. M. Schlegeland S. B. Vrtilek, eds.), pp. 3–17. ASP Conference Series 262.

Lisse C. M. (2002b) Cometary X-rays — the EUVE photometric


legacy. In Continuing the Challenge of EUV Astronomy:Current Analysis and Prospects for the Future (S. B. Howellet al., eds.), pp. 254–267. ASP Conference Series 264.

Lisse C. M. and 11 colleagues (1996) Discovery of X-ray and ex-treme ultraviolet emission from Comet Hyakutake C/1996 B2.Science, 274, 205–209.

Lisse C. M., Mumma M. J., Petre R., Dennerl K., Englhauser J.,Schmitt J., and Truemper J. (1997a) Comet C/1996 B2 (Hya-kutake). IAU Circular No. 6433.

Lisse C. M., Dennerl K., Englhauser J., Trümper J., Marshall F. E.,Petre R., Valina A., Kellett B. J., and Bingham R. (1997b) X-ray emission from Comet Hale-Bopp. Earth Moon Planets, 77,283–291.

Lisse C. M., Christian D., Dennerl K., Englhauser J., Trümper J.,Desch M., Marshall F. E., Petre R., Snowden S. (1999) X-rayand extreme ultraviolet emission from Comet P/Encke 1997.Icarus, 141, 316–330.

Lisse C. M., Christian D. J., Dennerl K., Meech K. J., Petre R.,Weaver H. A., and Wolk S. J. (2001) Charge exchange-inducedX-ray emission from Comet C/1999 S4 (LINEAR). Science,292, 1343–1348.

Mann R., Folkmann F., and Beyer H. F. (1981) Selective electroncapture into highly stripped Ne and N target atoms after heavy-ion impact. J. Phys. B.: Atom. Molec Phys., 14, 1161–1181.

McCammon D. and Sanders W. T. (1990) The soft X-ray back-ground and its origins. Annu. Rev. Astron. Astrophys., 28, 657–688.

McCammon D. and 19 colleagues (2002) High spectral resolutionobservation of the soft X-ray diffuse background with thermaldetectors. Astrophys. J., 576, 188–203.

Mumma M. J., Krasnopolsky V. A., and Abbott M. J. (1997) SoftX-rays from four comets observed by EUVE. Astrophys. J.Lett., 491, L125–L128.

Neugebauer M., Cravens T. E., Lisse C. M., Ipavich F. M., Chris-tian D., von Steiger R., Bochsler P., Shah P. D., Armstrong T. P.(2000) The relation of temporal variations of soft X-ray emis-sion from Comet Hyakutake to variations of ion fluxes in thesolar wind. J. Geophys. Res., 105, 20949–20956.

Northrop T. G. (1997) The spectrum of X-rays from Comet Hya-kutake. Icarus, 128, 480–482.

Northrop T. G., Lisse C. M., Mumma M. J., and Desch M. D.(1997) A possible source of the X-rays from Comet Hyakutake.Icarus, 127, 246–250.

Owens A., Parmar A. N., Oostrbroek T., Orr A., Antonelli L. A.,Fiore F., Schultz R., Tozzi G. P., Macarone M. C., and Piro L.(1998) Evidence for dust-related X-ray emission from CometC/1995 O1 (Hale-Bopp). Astrophys J., 493, 47–51.

Parker E. N. (1963) Interplanetary Dynamical Processes. Wiley,New York. 272 pp.

Phaneuf R. A., Alvarez I., Meyer F. W., and Crandall D. H. (1982)Electron capture in low-energy collisions of Cq+ and Oq+ withH and H2. Phys. Rev. A, 26, 1892–1906.

Raymond J. C., Uzzo M., Ko Y.-K., Mancuso S., Wu R., GardnerL., Kohl J. L., Marsden B., and Smith P. L. (2002) Far-ultra-violet observations of Comet 2P/Encke at perihelion. Astro-phys. J., 564, 1054–1060.

Robertson I. P. and Cravens T. E. (2003) X-ray emission from theterrestrial magnetosheath. Geophys. Res. Lett., 30, 1439–1442.

Robertson I. P., Cravens T. E., Snowden S., and Linde T. (2001)Temporal and spatial variations of heliospheric X-ray emissionsassociated with charge transfer of the solar wind with inter-stellar neutrals. Space Sci. Rev., 97, 401–405.

Ryufuku H., Sasaki K., and Watanabe T. (1980) Oscillatory be-havior of charge transfer cross sections as a function of thecharge of projectiles in low-energy collisions. Phys. Rev. A,21, 745–750.

Sagdeev R. Z., Evlanov E. N., Zubkov B. V., Prilutskii O. F., andFomenkova M. N. (1990) Detection of very fine dust particlesnear the nucleus of Comet Halley. Sov. Astron. Lett., 16, 315–318.

Schleicher D. G. and A’Hearn M. F. (1988) The fluorescence ofcometary OH. Astrophys. J., 331, 1058–1077.

Schmitt J. H. M. M., Snowden S. L., Aschenbach B., HasingerG., Pfeffermann E., Predehl P., and Trumper J. (1991) A softX-ray image of the Moon. Nature, 349, 583–587.

Schwadron N. A. and Cravens T. E. (2000) Implications of solarwind composition for cometary X-rays. Astrophys. J., 544,558–566.

Shapiro V. D., Bingham R., Dawson J. M., Dobe Z., Kellett B. J.,and Mendis D. A. (1999) Energetic electrons produced bylower hybrid waves in the cometary environment and soft X rayemission: Bremsstrahlung and K shell radiation. J. Geophys.Res., 104, 2537–2554.

Snowden S. L., Cox D. P., McCammon D., and Sanders W. T.(1990) A model for the distribution of material generating thesoft X-ray background. Astrophys. J., 354, 211–219.

Snowden S. L., Egger R., Finkbeiner D. P., Freyberg M. J., andPlucinsky P. P. (1998) Progress on establishing the spatial distri-bution of material responsible for the 1/4 keV soft X-ray diffusebackground local and halo components. Astrophys. J., 493, 715.

Suess S. T. (1990) The heliopause. Rev. Geophys., 28, 97–115.Szegö K. and 20 colleagues (2000) Physics of mass loaded plas-

mas. Space Sci. Rev., 94, 429–671.Uchida M., Morikawa M., Kubotani H., and Mouri H. (1998) X-

ray spectra of comets. Astrophys. J., 498, 863–870.Vaisberg O. L., Smirnov V., Omel’Chenko A., Gorn L., and Iovlev

M. (1987) Spatial and mass distribution of low-mass dust parti-cles (m < 10–10 g) in Comet P/Halley’s coma. Astron. Astro-phys., 187, 753–760.

Wargelin B. J. and Drake J. J. (2001) Observability of stellar windsfrom late-type dwarfs via charge exchange X-ray emission.Astrophys. J. Lett., 546, L57–L60.

Weaver H. A., Feldman P. D., Combi M. R., Krasnopolsky V.,Lisse C. M., and Shemansky D. E. (2002) A search for argonand O VI in three comets using the Far Ultraviolet Spectro-scopic Explorer. Astrophys. J. Lett., 576, L95–L98.

Wegmann R., Schmidt H. U., Lisse C. M., Dennerl K., andEnglhauser J. (1998) X-rays from comets generated by ener-getic solar wind particles. Planet. Space Sci. 46, 603–612.

Wickramsinghe N. C. and Hoyle F. (1996) Very small dust par-ticles (VSDP’s) in Comet C/1996 B2 (Hyakutake). Astrophys.Space Sci., 239, 121–123.

Wu W. K., Huber B. A., and Wiesemann K. (1988) Cross sectionsfor electron capture by neutral and charged particles in colli-sions with He. Atomic Data and Nuclear Data Tables, 40, 57.

644 Comets II

x-ray and extreme ultraviolet emission from cometsplorer (euve), and more recently, chandra and...

Documents