ARTICLE IN PRESS JID:YICAR AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.1 (1-15) Icarus ••• (••••) •••–••• www.elsevier.com/locate/icarus Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign C.M. Lisse a,∗ , K. Dennerl b , D.J. Christian c , S.J. Wolk d , D. Bodewits e , T.H. Zurbuchen f , K.C. Hansen f , R. Hoekstra e , M. Combi f , C.D. Fry g , M. Dryer h,i , T. Mäkinen i , W. Sun j a Planetary Exploration Group, Space Department, Johns Hopkins University Applied Physics Laboratory, 11100 Johns Hopkins Rd., Laurel, MD 20723, USA b Max-Planck-Institut für Extraterrestrische Physik, Giessenbachstrasse, 85748 Garching, Germany c Department of Astronomy and Physics, Queens University Belfast, Belfast, BT7 1NN, UK d Chandra X-ray Center, Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA e KVI Atomic Physics, Rijkuniversiteit Groningen, Zernikelaan 25, 9747 AA Groningen, The Netherlands f Department of Atmospheric, Oceanic and Space Sciences, Space Research Building, The University of Michigan, Ann Arbor, MI 48109-2143, USA g Exploration Physics International, Inc., 6275 University Drive NW, Suite 37-105, Huntsville, AL 35806-1776, USA h NOAA Space Environment Center, 325 Broadway, Boulder, CO 80305, USA i Finnish Meteorological Institute, Space Research, P.O. Box 503, FIN-00101 Helsinki, Finland j Geophysical Institute, University of Alaska, Fairbanks, AK 99775, USA Received 17 April 2006; revised 5 March 2007 Abstract We present results from the Chandra X-ray Observatory’s extensive campaign studying Comet 9P/Tempel 1 (T1) in support of NASA’s Deep Impact (DI) mission. T1 was observed for ∼295 ks between 30th June and 24th July 2005, and continuously for ∼64 ks on July 4th during the impact event. X-ray emission qualitatively similar to that observed for the collisionally thin Comet 2P/Encke system [Lisse, C.M., Christian, D.J., Dennerl, K., Wolk, S.J., Bodewits, D., Hoekstra, R., Combi, M.R., Mäkinen, T., Dryer, M., Fry, C.D., Weaver, H., 2005b. Astrophys. J. 635 (2005) 1329–1347] was found, with emission morphology centered on the nucleus and emission lines due to C, N, O, and Ne solar wind minor ions. The comet was relatively faint on July 4th, and the total increase in X-ray flux due to the Deep Impact event was small, ∼20% of the immediate pre-impact value, consistent with estimates that the total coma neutral gas release due to the impact was 5 × 106 kg (∼10 h of normal emission). No obvious prompt X-ray flash due to the impact was seen. Extension of the emission in the direction of outflow of the ejecta was observed, suggesting the presence of continued outgassing of this material. Variable spectral features due to changing solar wind flux densities and charge states were clearly seen. Two peaks, much stronger than the man-made increase due to Deep Impact, were found in the observed X-rays on June 30th and July 8th, 2005, and are coincident with increases in the solar wind flux arriving at the comet. Modeling of the Chandra data using observed gas production rates and ACE solar wind ion fluxes with a CXE mechanism for the emission is consistent, overall, with the temporal and spectral behavior expected for a slow, hot wind typical of low latitude emission from the solar corona interacting with the comet’s neutral coma, with intermittent impulsive events due to solar flares and coronal mass ejections.  2007 Elsevier Inc. All rights reserved. Keywords: Comet Tempel-1; Comets, coma; Comets, plasma; Solar wind; Spectroscopy 1. Introduction Since the surprise discovery of X-rays from comets in the mid-1990s using ROSAT and EUVE (Lisse et al., 1996; * Corresponding author. Fax: +1 240 228 8939. E-mail address: carey.lisse@jhuapl.edu (C.M. Lisse). Mumma et al., 1997; Dennerl et al., 1997; Krasnopolsky et al., 2000) over 20 comets have been detected in the X-ray regime, including many with the most advanced X-ray observatories, Chandra, CHIPS, and XMM-Newton (Lisse et al., 2001; Krasnopolsky et al., 2002; Dennerl et al., 2003; Wegmann et al., 2004; Lisse et al., 2005a, 2005b; Sasseen et al., 2006; Bodewits et al., 2007). These results have further refined our understand- 0019-1035/$ – see front matter  2007 Elsevier Inc. All rights reserved. doi:10.1016/j.icarus.2007.03.004 Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR 2 AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.2 (1-15) C.M. Lisse et al. / Icarus ••• (••••) •••–••• ing of cometary X-ray emission via charge exchange between instreaming solar wind minor ions and outflowing gaseous neutral cometary material. However, these all have been passive observations. NASA’s Deep Impact (DI) mission changed all that, by impacting Comet 9P/Tempel 1 (hereafter T1) with a copper impactor on 04 July at 05:44 UT at a relative velocity of 10.2 km s−1 . The collision of the 364 kg impactor released 106 –107 kg of fresh neutral cometary material into the extended atmosphere above the comet (A’Hearn et al., 2005; Meech et al., 2005; Schultz et al., 2007), increasing the neutral gas mass of the coma temporarily by up to 25%, and providing a controlled, highly monitored experiment to observe possible changes to the X-ray emission. Knowing exactly when this new neutral material arrived allows us to test the charge exchange emission mechanism (CXE; Cravens, 1997; Lisse et al., 2001; Kharchenko and Dalgarno, 2001; Kharchenko et al., 2003; Beiersdorfer et al., 2003; Bodewits et al., 2006). Because of the comet’s low outgassing rate (Schleicher et al., 2006), comparable to that of 2P/Encke (Lisse et al., 2005b), it was collisionally thin to charge exchange. The new material should therefore have directly increased the emission measure for the comet, and possibly changed the signature of the CXE emission to a collisionally thick system. (Changes in the solar wind flux at the comet also cause changes in the observed X-ray flux, and need to be carefully accounted for.) One of the other exciting aspects of the impact was to see if enough energy was released to create prompt X-rays by hyper-velocity impact processes, as suggested by ROSAT observations of the impact of the Kfragment of Comet D/Shoemaker–Levy 9 on Jupiter (Waite et al., 1995). In this paper, we report the results of our extensive Chandra X-ray Observatory (Chandra) observing campaign of T1 during the Deep Impact encounter. The Chandra observations, in addition to showing possible changes in the X-ray emission as the result of the impact, provide one of the longest duration studies of changes to cometary emission in the X-ray. We describe the details of the Chandra and ACE instruments used, observing strategy, and our data analysis techniques in Section 2. In Section 3 we discuss the solar wind environment during the Chandra observations and the comet’s outgassing behavior. In Section 4 we present the Chandra observational results, including T1’s morphology, temporal changes and spectra, and place these in the context of previous cometary X-ray observations. We compare changes in the X-ray emission to the behavior of the solar wind and coma, and discuss the most dominant processes responsible for the observed variations. We also report Advanced Composition Explorer measurements of the solar wind during the Chandra measurements, critical to understanding the observed X-ray emission. We then discuss the changes caused by the Deep Impact experiment. In Section 5 we summarize our findings, the most plausible physical model for Tempel 1, and what was learned from the X-ray observations of the DI experiment. 2. Instruments and observations 2.1. Chandra X-ray Observatory The Chandra observations and data analysis were done in the same manner as for our previous comet observations (Lisse et al., 2001, 2005b) and the reader is referred to these papers for details of the observational scheme. Briefly, the ACIS-S CCD was used, providing high resolution X-ray imaging with platescale of 0.5′′ /pixel, instantaneous field of view 8.3′ × 8.3′ , and moderate resolution spectra (E ∼ 110 eV FWHM, σGaussian ∼ 50 eV) in the 300–2500 eV energy range. The ACIS-S optical axis was pointed on the sky 2′ in front of the comet along the direction of its apparent motion, and the exposures were taken as the comet moved through the ∼4′ field of view. T1 was centered in the back-illuminated CCD chip S3, the most sensitive CCD for energies below 1 keV. No active guiding on the comet was attempted, and Chandra was able to follow the comet’s non-sidereal motion using multiple pointings. Chandra observations of comet T1 were planned to maximize our detection of any changes caused by the impact. For this reason preliminary observations were obtained on June 30th to establish the baseline emission of the comet, and long duration observations were obtained immediately before, during, and after the impact on July 4th and 5th. These observations were followed by several monitoring “snapshot” observations on July 8, 10, 13, and 24 (Table 1). Each pointing obtained between 28 and 64 ks on the comet for a total integration time of ∼295 ks. Combined images from each set of observations were remapped into a coordinate system moving with the comet using the “sso_freeze” algorithm, part of the Chandra Interactive Analysis of Observations (CIAO) software. Source photons from the comet were extracted from a circular region 3.7′ in radius and centered on the comet, while background photons were extracted from a region defined between 4′ from the nucleus and the edges of the S3 chip. Images, light curves and spectra were extracted with the CIAO tools and analyzed with a combination of IDL and FTOOL packages. The X-ray spectral analysis software package XSPEC (Arnaud, 1996) was used for the spectral fitting. 2.2. Advanced Composition Explorer The Advanced Composition Explorer (ACE) carries solar wind plasma and composition instruments that, since its launch in late 1997 have provided the best coordinated solar wind dynamics and composition measurements, including elemental, ionic and isotopic abundances and fluxes of the relevant particle species found in interplanetary space (Stone et al., 1998). ACE is located in a halo orbit near the Lagrangian libration point L1. Due to its real-time capability for space environment data, ACE data are being used for real-time space weather forecasting world-wide (Zwickl et al., 1998). The instrument of main interest for this study is the Solar Wind Ion Composition Spectrometer (SWICS), an ion mass Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.3 (1-15) Chandra-DI observations of Comet 9P/Tempel 1 3 Table 1 Chandra 9P/Tempel 1 observation times and observing parametersa Observation time (UT 2005) DOY 30 June 10:46 04 July 04:12 05 July 17:12 08 July 09:13 10 July 23:39 13 July 15:31 24 July 05:54 181.7 185.6 187.0 189.6 191.2 194.8 205.3 rh (AU)  (AU) Phase (degree) Heliocentric Lon. (degree) Lat. (degree) 1.51 1.51 1.51 1.51 1.51 1.51 1.51 0.875 0.894 0.902 0.916 0.931 0.945 0.976 40.6 40.9 41.1 41.2 41.4 41.5 41.6 245 247 248 250 252 253 256 0.79 0.34 0.16 −0.14 −0.47 −0.77 −1.36 QOH (1027 mol s−1 ) Texp (ks) Total counts in energy range after background removal (eV) 300–500 500–1000 300–1000 4.73b 5.01b 8.18b 6.31b 5.01c 5.6c 5.6c 52 64 48 28 35 34 34 1130 300 470 570 85 100 175 1610 1255 1170 1500 975 340 815 2740 1560 1640 2070 1060 440 990 a DOY (Day of Year) is the center of the observation time. r is the comet–Sun distance,  is the comet–Earth distance. The phase angle is the angle Sun–comet– h observer; the heliocentric ecliptic latitude and longitude are the comet’s position as seen from the Sun, and Qgas is the gas production rate from the nucleus surface. For reference, the heliocentric longitude and latitude of the Earth were in the range 281◦ –292◦ , and 0◦ , respectively. b Schleicher et al. (2006). c Lisse et al. (2006). spectrometer that resolves the elemental and ionic composition of all significant solar wind species. SWICS is a combination of three key systems: an electrostatic analyzer, a time-of-flight telescope, and a solid state detector array (Gloeckler et al., 1998). SWICS resolves solar wind ion species by measuring, at the same time, solar wind speed, mass per charge ratio with ∼5% resolution, and ion mass with a resolution of the order 15–20%. The larger the mass of the solar wind ion detected, the coarser the mass resolution. The analysis of SWICS data therefore uses a process separating neighboring and overlapping peaks (von Steiger et al., 2000). This process does a very good job analyzing the most abundant ion species and charge states, but has limits when analyzing very small fluxes of lowabundance charge states near multiple large peaks, e.g., the extremely rare but CXE important ion N7+ . 3. Solar wind and comet characteristics at the time of the Chandra observations The flux and the heavy ion composition of the solar wind at the comet, and the comet’s outgassing rate both strongly affect the observed X-ray emission (Lisse et al., 1999, 2001). A detailed analysis of the solar wind dynamics, composition/charge state, and the comet’s activity are therefore a crucial part of our analysis, and we go into some detail describing them here. 3.1. Large scale modeling of the solar weather The nearly 30 day time span of the Chandra observations contained several significant changes in the solar wind heavy ion and proton environment (Figs. 1 and 2, respectively). To understand the large scale solar wind interplanetary structures influencing the comet’s X-ray emission behavior, we utilize here the well tested Hakamada–Akasofu–Fry solar wind model (HAFv2; Fry et al., 2001, 2003, 2004; Fig. 2, Table 2), which takes into account the source solar coronal structure behavior, and follows the solar wind propagation through interplanetary space while allowing for solar wind plasma density, temperature and velocity dispersion relationships. We refer the reader Fig. 1. ACE/SWICS solar wind parameters. The top panel shows densities and relative compositions of a few solar wind heavy ions, as indicated by the legend at the right. The middle panel indicates the velocity of some of these ions, indicating a very tight correspondence of all speeds. The labels refer to the text. DOY 178 corresponds to June 28, DOY 213 to August 1. The lowest panel shows thermal speeds, which also appear to track each other very well. Unusual states of the solar wind can be easily located by searching for deviations from normal in the O7+ /O6+ (brown/olive curve) in the top panel. to Fry et al. (2001, 2003, 2004), Dryer (1994, 1998), and Dryer et al. (2001, 2004) for more details of the model. On June 30, at the beginning of the period, the solar wind was dominated by a slow stream at ∼350 km s−1 . This condition quickly changed on the latter part of June 30th as the heavy ions’ wind speeds rose to 650 km s−1 (Fig. 1) while maintaining roughly the same density. Within 3 days after the Co-rotating Interaction Region’s (CIR’s) passage, as suggested by the compressed Interplanetary Magnetic Field (IMF) lines in Fig. 2, the solar wind ramped down in a smooth fashion to the preJune 30th state. On July 4, Tempel 1 was about 35◦ east of the Earth–Sun line essentially in the ecliptic. The western flank of the modeled shock in Fig. 2 (from a July 1st C5.3 solar flare at N14E83 as noted as FF#608 in Table 2) might have intercepted the comet on July 4th with a somewhat higher flux of proton Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR 4 AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.4 (1-15) C.M. Lisse et al. / Icarus ••• (••••) •••–••• (a) (b) (c) (d) (e) (f) (g) Fig. 2. Solar wind interplanetary magnetic field structures, estimated using the HAFv2 model at the times of the Chandra observations. Red: magnetic polarity away from Sun; blue: toward the Sun. The coordinate system is semi-inertial (HEQ, heliocentric solar equatorial) with 0◦ longitude at the intersection of the solar equatorial plane and the ecliptic plane. (The ecliptic plane is inclined by 7.25◦ to the solar equatorial plane. The HAF model resolution is 5◦ in heliolatitude and heliolongitude, so the ecliptic plane and equatorial plane IMF plots are quite similar, particularly in areas near the intersection of the equatorial and ecliptic plane. Since T1 is near the 0◦ –180◦ line on the IMF plots, its low heliocentric ecliptic latitude puts it close to the solar equatorial plane during this period.) There are no obvious ICME structures in the comet-ward directions during the times of the Chandra–Tempel 1 observations except for the west flank of the interplanetary shock on July 4 (DOY 185). Interplanetary shocks from parent solar flares (indicated in Table 2 as FF#611, #613, #614, #615, and #617) passed ACE, respectively on 12, 16, 16, 17, and 19 July as given by real-time-reported plasma and magnetometer Level 1 data from ACE (Zwickl et al., 1998). Their eastern flanks however are considered to have been very weak at Tempel 1. On the other hand, CIRs are indicated to intercept the comet on June 30 (DOY 181), July 8 (DOY 189), and July 24 (DOY 205). On the basis of the model’s output with ACE proton data, the simulated FF#611 shock (see the panels for July 11/DOY 192 and 13/DOY 194) arrived a day earlier than the actual one; thus, the implied eastern flank’s shock speed and strength at the comet are too high. and heavier ions (C, N, O, Ne). We expect, however, any solar wind signature on July 4th from the above-mentioned shock to have been relatively weak (Fig. 2). There was a significant shock (FF#611, Table 2; see caption for Fig. 2) from another solar flare on July 9th at N19W28 accompanied by a fast (∼1600 km s−1 at the Sun) halo Coronal Mass Ejection (CME). The shock appears in both our model and in the ACE solar wind proton measurements (Fig. 1, Section 3.2). The further, weakened eastern flank (see Fig.’s 2 caption) of this flare shock passed over Tempel 1 on July 13th (DOY 194), causing milder changes in the observed X-ray behavior. What makes this event very different than the June 30th event is that the accompanying Interplanetary Coronal Mass Ejection (ICME) produced prompt, hot ions moving rapidly and radially from the Sun, widening and prolonging the solar wind perturbation before the much slower shock reached the comet. [A similar situation with a radially traveling ICME was found for Comet C/1999 S4 (LINEAR) in 2000 (Lisse et al., 2001).] This led to a prolonged increase in the solar wind flux at the comet and ACE, starting on July 9th and producing a prolonged X-ray response. Finally, it should be noted that the solar flares in Table 2, from FF #612 onward, were likely ineffective in “reaching” T1 because of their far-westward and solar backside locations. Instead, CIRs were the major global solar wind structures passing over T1 in our case, thereby justifying the mapping procedure discussed earlier. Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.5 (1-15) Chandra-DI observations of Comet 9P/Tempel 1 5 Table 2 Solar flares with coronal shock waves during the Chandra/Tempel 1 epoch with potential interplanetary consequences monitored in real timea FF# Date DOY Time Lat. Lon. Vs TAU X-ray classif. Opt. classif. Comments 606 607 608 609 610 611 612 613 614 615 616 617 618 619 03 June 16 June 01 July 07 July 08 July 09 July 10 July 12 July 13 July 14 July 16 July 17 July 21 July 24 July 154 167 182 188 189 190 191 193 194 195 197 198 202 205 23:55 20:10 5:01 16:29 6:34 21:58 15:19 16:27 15:27 11:20 7:13 9:50 3:24 13:36 S 17 N 09 N14 N 09 N 19 N 12 N 00 N 09 N 11 N 11 S 10 N 11 N 11 N 11 E 10 W 87 E 83 E 03 E 31 W 28 W 90 W 64 W 90 W 90 W 74 W130 E167 E120 499 989 1200 1000 330 1651 869 1000 1360 1236 667 2000 2000 3000 100 20 15 130 100 100 40 300 400 230 200 100 100 100 C 6.2 XXXX C 5.3 M 4.9 B 5.5 M 2.8 C 9.9 M 1.5 M 5.0 X 1.2 C 2.2 XXXX XXXX XXXX SN SF XX SN SF 1N SF SF SF XX XX XX XX XX SOHO/LASCO NOT AVAILABLE Same as above NARROW CME HALO CME SOHO/LASCO NOT AVAILABLE HALO CME SOHO/LASCO NOT AVAILABLE PARTIAL HALO CME HALO CME Same as above SOHO/LASCO NOT AVAILABLE HALO CME Same as above Same as above a FF# refers to the events that have been continuously “fearlessly forecasted” in real time since 1997 (see, for instance, Fry et al., 2003, 2004; Dryer et al., 2004). ‘Time’ refers to the start time of US Air Force ground-based observations of metric Type II radio bursts signifying a coronal shock. Lat. and Long. refer to the flare location on the Sun in solar polar coordinates. ‘Vs ’ is the shock speed in the low corona using the Type II’s frequency drift rate. “Comments” refer to the time if—or when—coronagraph observations as reported by NASA/SOHO operations scientists shortly after CME imagery is obtained. XXXX and XX indicate that flare X-ray and optical (H-alpha) classifications were not possible. Backside flare locations are based on time elapsed since West limb passage. TAU (in HMM) refers to the shocks’ piston-driving times assuming the flares’ X-ray emission durations as a proxy. The HAFv2 model uses only Columns 2 and 4–8 as inputs. 3.2. ACE measurements during the Chandra observations ACE, the source of all our major solar wind data, was located at roughly 34◦ –42◦ west of the comet (as observed from the Sun) during the observations. Analysis of the ACE solar wind data has to be done in two successive steps: In the first step, the calibrated heavy ion fluxes are computed from the spacecraft telemetry and the resulting temporal trends are produced (Fig. 1a). The second step in the determination of the solar wind ion flux at the comet is the mapping of the ACE data from the near-Earth ACE orbit to the comet’s heliocentric location. During the Chandra observations, the comet was essentially in the ecliptic plane and therefore had a negligible latitudinal separation from ACE (see Table 1). This allowed us to use a 1-D MHD mapping as described in Prange et al. (2004). This process takes into account an expansion of the solar wind, and its dynamic evolution during this expansion. Due to the relatively modest, ∼0.5 AU expansion, all dynamic signatures seen in Figs. 1 and 2 remain visible. We then assume persistence in the solar wind structure during 35◦ of solar rotation. (This is equivalent to 2.5 days of a rotation of a Carrington spiral, so our assumption is valid for the large-scale solar wind expansion, but may not be very good for CMEs.) The middle panel of Fig. 1 shows the solar wind speed patterns after the detailed radial calculation, shifted to account for the longitudinal offset. The simulated solar wind conditions are discussed in Fig. 2. Overall, we see a tremendous degree of variability of heavy ion densities in the ACE data, due mainly to three different effects. (1) There are dynamic changes in the solar wind. E.g., there are a number of fast velocity changes caused by transitions from a slow/coronal streamer associated wind, to a fast/coronal hole associated wind. Such transitions occur near July 2, 25, and 31 (DOY 183, 206, and 212) and are labeled A, B, and C in the figure, respectively. All speed increases also lead to large increases in temperature, and hence the thermal speed of heavy ions. Some of these transitions are shock-like, but other compressions have not succeeded to form shocks at the time of the observation. There is another abrupt increase in speed on July 10 (DOY 191, denoted D) and a less dramatic increase on July 18 (day 199, denoted E). Both of them are different in nature and associated with either a co-rotating interaction region (CIR) or an Interplanetary Coronal Mass Ejection (ICME, a high velocity, transient solar flare-generated CME preceded by its low corona-formed shock wave). There is a long term persistence of fast solar wind streams (formed by high speed flows from coronal holes that compress preceding slower flow, thereby forming shocks), lasting many days, as demonstrated by Smith and Dryer (1991). (2) The heavy ion compositions can change due to variations in elemental abundance. The variability of the elemental abundances is of the order of a factor of 2 (von Steiger et al., 2000, and references therein), and are therefore a minor part of the variability shown here, which is over 2 orders of magnitude. We see such effects on July 10 (DOY 191). (3) The solar wind (SW) heavy ion composition changes due to the electron temperature of the solar wind sources (see Zurbuchen and Richardson, 2007). These variations have shown to be the best signatures of CMEs and coronal holes, and are clearly visible in the top of Fig. 1. There are correlated decreases of O7+ and N6+ , associated with coronal hole winds, on July 2, 25, and 31 (DOY 183, 206, and 212). CME dominated winds, as seen on June 30 (DOY 181) do not show such decreases, but instead lead to significant increases of high charge states, such as O8+ (and presumably also N7+ ). In summary, we expect from the HAFv2 models, and see from the ACE data, large changes in the solar wind fluxes and charge states at the comet on June 30, July 9–10, 13, 18, 25, and 31 of 2005. We note that continuous modeling is essential Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR 6 AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.6 (1-15) C.M. Lisse et al. / Icarus ••• (••••) •••–••• to gain a global temporal and spatial perspective of the highly variable heliosphere with both co-rotating and transient structures, the latter considerably altering the former. 3.3. Measures of Tempel 1’s outgassing behavior Unlike our previous Chandra comet observations, the comet was being heavily monitored at other wavelengths sensitive to the neutral gas emission rate from the nucleus. The solar wind and cometary dust and gas emission activity were all studied in depth and length, by over 80 observatories around and above the Earth. In fact, one of the key science results to come from the Deep Impact experiment is the verification of the CXE mechanism as driving cometary X-ray emission using multiple lines of evidence. The Deep Impact spacecraft observed the comet’s behavior in detail from 30 June 2005–5 July 2005. A number of brief, but large explosive outbursts were seen by the DI and also by the HST spacecraft over the period of June–July 2005 (A’Hearn et al., 2005; Farnham et al., 2007). Rise times for these outbursts were typically on the order of minutes, and fall times typically on the order of hours. A direct comparison of the DI detected outbursts and the Chandra detected outbursts does show a correspondence for the “weak” outburst of 30 June 2005 with the strong Chandra outburst of the same day. There is no DI data corresponding to the 8 July 2005 strong Chandra outburst, and observations by other observatories do not show enhanced neutral gas emission from the comet on this day. Lowell Observatory observed the production rate of OH, the direct daughter product of H2 O, the majority neutral gas species emitted by the comet, over the time span June 29 to July 11. The Spitzer Space Telescope monitored the total 16 µm brightness due to thermal continuum emission from dust and water ice emitted by the comet over the time span July 2– July 14, 2005. Correcting both of the water daughter product datasets for the mean lifetime of water vs UV photolysis at 1.51 AU, ∼5 × 104 s, we find good agreement with the relative Spitzer dust production temporal trends where they overlap in time. Here we use the combined curve as a scaled proxy for the production rate of neutral gas molecules at the nucleus, Qgas . Corresponding measurements of the Ly-α emission from neutral H-atoms over the time range 5–30 July by the SWAN UV instrument onboard the SOHO spacecraft (Mäkinen et al., 2001, 2007) was obtained as a check of our method. The neutral H-atom coma of a comet is produced by the same UV photolysis of water that produces OH. While sparsely sampled, the SOHO/SWAN lightcurve is taken by a stable spacecraft platform, is long in duration, and has been calibrated and verified for numerous comets. The water production rate found by SWAN was relatively flat from 5 July through 30 July, in agreement with our derived Qgas curve. (A small apparent upturn, of amplitude 50–150% and uncertain duration, was seen on July 14–15, with the production rate returning to the average quiescent value of ∼1 × 1028 mol s−1 by July 17.) 4. Chandra results 4.1. Morphology Chandra ACIS-S3 images of the 9P/Tempel 1 X-ray emission are shown in Fig. 3. T1’s emission morphology is centered on the nucleus and is nearly point-like, unlike the majority of the 20+ comets observed in the X-ray to date (Lisse et al., 2005a). The point-like morphology is very similar to the morphology seen for 2P/Encke 2003 (Lisse et al., 2005b), which was determined to be due to a collisionally thin coma, a result of the low rate of neutral gas emission from the nucleus. The reported rate of water gas production for T1, 1–3 × 1028 mol s−1 , was very close to the 2 × 1028 mol s−1 for 2P/Encke in 2003. In the collisionally thin regime, the maximum of the comet’s emission occurs very near the nucleus, where the maximum density of coma neutral gas is found. With T1 at 0.88–0.98 AU from the Earth, 1′′ corresponded to 630–710 km at the nucleus, so that the ACIS-S3 pixel (0.5′′ × 0.5′′ ) scale was 315–355 km pixel−1 at the comet. However, the actual size of a statistically significant (i.e., 5σ ) spatial resolution element was determined by the limited number of photons detected. In regions of high net signal, e.g., near the nucleus, this resolution was on the order of 2′′ (∼1300 km) per resolution element. The average photon limited effective pixel size realized over the entire detected X-ray emission region was 7′′ (∼4700 km) per resolution element. We find a maximum extent of detectable X-ray emission of ∼200,000 km for the comet. This implies an X-ray extent larger than the bowshock distance: following the analysis of Wegmann et al. (2004) and Wegmann and Dennerl (2005), we expect rbowshock = 1.7 × 104 km on June 30; 4.5 × 103 km on July 4 before impact and the other quiescent periods; and 1.0 × 104 km on July 8. As the comet was weakly active (Qgas ∼ 1028 mol s−1 , Mäkinen et al., 2007), we can expect penetration of the SW to near the nucleus. Using the results Giotto Grig– Skellerjup flyby (Johnstone et al., 1993; Lebedev, 2000), we estimate a radius for the SW ion contact surface, where molecular collisions become dominant in the coma, of <100 km, much less than one Chandra ACIS-S3 pixel in extent. The strongest evidence for deviations from spherically symmetric gas emission behavior from the nucleus occurred in our observations on the days of 30 June 2005 and 4–5 July 2005 (DOY 181 and 185–186). On June 30th, a brightening was seen in the anti-solar direction, increasing the radial extent of the emission by a factor of 50 ± 25%. We interpret this as evidence for an outburst of neutral material from the comet with projected direction in the anti-solar direction. On 4–5 July 2005, an extension was seen in the images in the same SW direction as the flow observed in near-nucleus images of the ejecta produced by the impact (A’Hearn et al., 2005; Meech et al., 2005; Sugita et al., 2005). The statistical significance of the extension is ∼5σ above background in each pointing. Larger scale imaging by the ROSETTA spacecraft over the course of the next few days confirms the progression of the water gas and OH ejected by the impact (Küppers et al., 2005). The estimated propagation velocity of the extension is 0.5–1 Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.7 (1-15) Chandra-DI observations of Comet 9P/Tempel 1 7 (a) (b) Fig. 3. Chandra ACIS-S3 images of the 9P/Tempel 1 X-ray emission. Images were created from all photons with 0.5 < energy < 0.7 keV, smoothed with a 5 × 5 pixel Gaussian filter. Note that the imagery from the July 24th observation has been omitted for ease of presentation; the observed structure is nearly identical to that found on July 13th. North is up, East to the left, and the projected direction to the Sun is to the right in each image. The images cover an extent that is roughly 3.5 × 105 km by 3.5 × 105 km. Inset: July 4th KPNO 4 m R-band image, rotated and scaled to match the Chandra images (Farnham et al., 2007). Arrows denote the path of the ejecta, which is moving down and to the right (SW). The position of the nucleus in the KPNO and Chandra images is denoted by a black cross. (a) The brightness in each image has been autoscaled to the maximum value of the image, enhancing the extension in the fainter, non-flare observations. (b) Same as (a), but with each image hard-scaled to the maximum brightness on the July 11th image. km s−1 for the ∼33 h duration of our 4–5 July observations, but we are limited by total signal and possibly viewing geometry. The derived expansion velocity is higher than the ∼0.2 km s−1 found for the leading edge of the solid ejecta (Lisse et al., 2006; Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR 8 AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.8 (1-15) C.M. Lisse et al. / Icarus ••• (••••) •••–••• Schleicher et al., 2006), and we associate it instead with the higher expansion velocity of neutral gaseous material produced by the impact, as expected for CXE driven X-ray emission. We use the estimated total mass of water ejected by DI (∼5 × 106 kg), and our estimate of a 10σ X-ray detection by Chandra over 4–6 July in a total on-target time of 130 ks to estimate the limiting sensitivity of Chandra to cometary activity as 5 × 106 kg/130 ks/10σ = 4 kg s−1 (1σ , equivalent to ∼1026 mol s−1 of water). Comparison of the Chandra X-ray morphology to the ROSETTA OH images shows that X-rays are not the most sensitive way to detect OH neutral gas; optical narrow band imaging has a superior emitted flux/background statistic. 4.2. Light curve A Chandra X-ray light curve was created for Tempel 1 using a 300–1000 eV passband, with corrections applied for the 24% increase in projected area from June 30 to July 24. The 300–1000 eV range allows the best tradeoff between exclusion of instrumental background and inclusion of maximal comet signal (Fig. 4a). Impulsive events were found in our data and these are similar to previous observations of Comets Hyakutake, Encke, and Linear S4 (Neugebauer et al., 2000; Lisse et al., 1999, 2001). Overall, for an average distance to T1 of 1.507 AU, the 300–1000 eV X-ray average luminosity, Lx was 4.6 × 1014 erg s−1 , the fourth lowest detected to date. As Comet Tempel 1 was collisionally thin to charge exchange to within a fraction of the central pixel, the X-ray lightcurve is directly proportional to the comet’s neutral gas emission activity, the flux of solar wind minor ions in the coma, and the cross-section for CXE between the two. The cometary X-ray emission observed is driven by the relatively abundant fully stripped, hydrogenlike, and helium-like C, N, O, and Ne ions. All Lyman- or K-series photons emitted after electron capture by these ions are within the energy range detectable by Chandra (0.3–1.0 keV), and therefore the total cross-sections for photon emission are equal to the total charge exchange cross-sections (Bodewits et al., 2007). As high signal-to-noise measures of C, N, O, and Ne ion flux measurements are not always available for use in our analysis, we use both the direct C, N, O and Ne measurements where they are available, and the more frequent and higher signal-to-noise proton flux measurements to bridge the gaps in coverage, assuming they trend in the same overall fashion as the minor ion fluxes. The long term Chandra lightcurve is shown in Fig. 4a, along with the time history of the comet’s gas production rate estimated from Lowell narrowband OH filter measurements (Schleicher et al., 2006) and Spitzer continuum data (Lisse et al., 2006) and the time history of ACE/SWICS solar wind proton flux data. Qgas has been estimated using the QOH outgassing rates from Schleicher et al. (2006), shifted by −1 day, to allow for the ∼105 s required to photolytically form OH from the parent water molecule emitted by the nucleus, combined directly with 16 µm Spitzer lightcurve data. (The Spitzer data are dominated by thermal emission from coma dust and water ice, and thus measure changes in coma emission activ- (a) (b) Fig. 4. (a) Total 0.3–1.0 keV Chandra count rate versus time for the period June 30–July 14, 2005, with 500 s bins smoothed by an 11 point boxcar filter (small solid squares). The sharp impulse on DOY 181.7 (June 30) is in strong contrast to the wide observed increase in X-ray count rate in DOY 189.6 (July 8). Points within a boxcar width of the beginning or end of each Chandra visit to the comet are not shown. The dashed line marks the time of impact of the DI spacecraft. Also plotted, on an arbitrary scale, are the interpolated and scaled Qgas temporal trends (triangles), np × vp solar wind trends (stars), SWIFT photometry (Willingale et al., 2006; open large squares) and np × vp × Qgas trends (solid line). Qgas has been estimated using the QOH outgassing rates from Schleicher et al. (2006), shifted by −1 day, combined with the 16 µm Spitzer lightcurve data of Lisse et al. (2006). (b) Chandra lightcurve phased to the diurnal rotation rate of 1.6 days. Error bars are 1σ . Zero phase corresponds to Day 181 00:00 UTC of Year 2005. ity promptly, assuming the dust and gas emission rates from the nucleus are coupled.) This assumption seems valid, as the Lowell and Spitzer trends agree very well over the 8 day overlap period 2–10 July 2006. The predicted lightcurve formed from the product of Qgas and the wind factor is also shown. The observed X-ray fluxes mirror the solar wind behavior well, except on June 30th (DOY 181), and at impact on July 4th (DOY 185). Tempel 1 also represents the first comet to be observed and Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.9 (1-15) Chandra-DI observations of Comet 9P/Tempel 1 9 detected simultaneously by multiple orbiting X-ray telescopes (Chandra, SWIFT, and XMM). In Fig. 4a we have included the SWIFT photometric data for the comet (Willingale et al., 2006), which agrees well with the trends observed by Chandra and our model flux predictions. On June 30th we find that the measured solar wind flux increase can account for at most 50% of the observed X-ray flux increase; the rest must come from increases in the comet’s neutral gas outflow rate. A corresponding increase in brightness was indeed found in the Deep Impact approach lightcurves for this date (Farnham et al., 2007). This was followed by a period of relatively quiet wind and low outgassing around the time of impact. On July 4th we examined the observed X-rays for any direct indication of changes as the result of the impact. While there is a possible prompt spike jump in the very first points of the lightcurve measurements after impact, it is only found after smoothing the data with effective bin sizes of 5000 s. We thus do not claim a detection of an impact flash in the X-ray. We estimate the maximum amplitude of any flash to be less than 0.02 cps (30% of the pre-impact quiescent count rate). Using this upper limit, we estimate that it would take at least 5 times as much signal, and hence at least 5 times as much energy deposited onto a cometary body at a distance of 0.9 AU from Chandra, to make a significant X-ray flare we could detect. We also searched the July 4 light curves for any possible enhancements immediately after impact and also searched for any decreases in emission in the interval following the impact, assuming either absorption or quenching were possible. We find we can rule out changes in the solar wind as the cause of the small, but significant (18 ± 5%) increase found in the average trend of the X-ray emission over the course of July 4. This emission rate stayed high on July 5, although the solar wind flux trends did not show this level of increase. The increase in the observed X-ray flux on July 4th is consistent with a total amount of water gas released by Deep Impact of ∼5 × 106 kg, as reported from ROSETTA measurements of coma OH emission during and after the impact (Keller et al., 2005; Küppers et al., 2005). This amount of water gas, equivalent to approximately 10 hours of the normal ambient outflow rate of about 1 × 1028 mol s−1 , is roughly 30% of the total gas contained in a coma of radius 1 × 105 km. The July 8th (DOY 189) flare appears to be driven mainly by a strong, prolonged increase in the solar wind flux (Section 3). After another period of relative quiescence, the comet again encountered increased solar wind fluxes on July 16 (DOY 197), which had disappeared by July 21. By the time of the very last Chandra observation on July 24, the solar wind flux had begun to increase once again. It is important to note here that there is no evidence for the large upsurge in material (1.9 ± 0.4 × 108 kg of water or 3.9 ± 0.5 × 108 kg of carbon dioxide) emitted from the comet into the coma due to the DI experiment as reported by Willingale et al. (2006), starting on July 8; instead, the broad increase in X-ray flux seen in the combined Chandra/SWIFT X-ray lightcurves is explained by increases in the solar wind flux at the comet (Fig. 4a). There is also no obvious increase in the SWIFT X-ray count rate during the increased outgassing on June 30th–July 1, 2005. The long term temporal behavior of the X-ray luminosity, before, during, and after the Deep Impact experiment, can be entirely understood using measures of the coma neutral gas density (Küppers et al., 2005; Jehin et al., 2006; Lisse et al., 2006; Schleicher et al., 2006) that show an elevated density of material for the first two days after impact, then normal cometary behavior afterwards. 4.2.1. Nucleus rotation effects For 2P/Encke, we found a lightcurve responding to diurnal changes in outgassing due to the rotation of the asymmetric nucleus with discrete regions of enhanced outgassing (Lisse et al., 2005a, 2005b, 2005c). Because the comet was collisionally thin to CXE, we expected and found prompt response to increases in coma neutral density, although our detection was somewhat noisy. Here we have attempted a similar experiment to confirm this result: taking the ensemble of Chandra data and searching for the best-fit long term lightcurve, to see if it is consistent with other measures of activity modulation due to nucleus rotation (e.g., the Deep Impact spacecraft approach imagery lightcurve (A’Hearn et al., 2005) or the ground based CN narrow band photometry of Jehin et al., 2006). Allowing for effects due to the windowing function, episodes of cometary outburst, episodes of increased solar wind flux and artifacts due to the fixed pointing + comet drift method, we find that phasing the data to a period of 1.70 days (as reported by Lisse et al., 2005c; A’Hearn et al., 2005; Jehin et al., 2006; Farnham et al., 2007) yields a periodic waveform. Running a χ 2 period search, we find the best fit period to be 1.6 ± 0.15 days (Fig. 4b). Jehin et al. show a T1 folded CN/NH lightcurve from 3–7 July 2005 with peak-to-peak amplitude of 50% to have the highest peaks between phases 0.0 and 0.3, similar to the maxima at phases of 0.1–0.4 in the Chandra data. Removing the high count rate flare data of June 30th and July 8th from the fit, however, reduces the significance of the periodic fit substantially. As the flare data represent only 20% of all photons, this change should not be large, but a simple calculations shows that the flares are separated by exactly 5 rotations if we assume a 1.63 day period. We thus consider the rotational detection in the X-rays to be somewhat weak, and only report here that the X-ray lightcurve is consistent with a 1.6–1.7 day period. 4.3. X-ray spectroscopy The Chandra 9P/Tempel 1 spectra demonstrated the classical emission features of cometary X-ray emission (Fig. 5), with strong emission due to highly charged C, N, O, and Ne solar wind ions charge exchanging with neutral gas species in the cometary coma (Lisse et al., 2001, 2005a; Krasnopolsky et al., 2002; Krasnopolsky, 2006). In modeling the faint T1 comet spectrum, we have required our spectral model to be consistent with current models of charge exchange emission (Bodewits et al., 2004) and models of the observed spectra of 7 other comets measured by Chandra (Bodewits et al., 2007). Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR 10 AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.10 (1-15) C.M. Lisse et al. / Icarus ••• (••••) •••–••• (a) (b) Fig. 5. (a) Spectra (background subtracted) for a sample of the Chandra ACIS observations plotted from 300 to 1000 eV. Each spectrum is offset from the previous by 0.2 count s−1 keV−1 . 1σ error bars are shown. The spectra show strong day-to-day changes in the emission between 500 and 650 eV, attributable to changes in the OVII and OVIII emission (see text). The natural X-ray outbursts of June 30th and July 8th caused much larger changes than the man-made outburst of July 4–5. The spectra of June 30th (flare) and July 8th (quiet) are very similar, while the July 8th (flare) spectrum, though noisy, has much stronger C, N, and O charge exchange emission lines, suggesting the dominant driver for the observed spectral changes is the solar wind ion composition. (b) Ion abundances for O8+ (top) and C6+ (middle) relative to O7+ . In the lower panel, the MHD results for the proton velocity are given for comparison. In all panels, the lines are ACE/SWICS data, as obtained with the ’simple extrapolation’ (corotational + distance), and the error bars are 90% C.L. (±1.6σ ). The dashed lines are averaged slow wind abundances from Schwadron and Cravens (2000). We have done this by fixing the line energies of the model to the C, N, O, and Ne emission lines detected in the other comets, effectively minimizing the number of free parameters in our line emission models. While we found that other spectral models, such as the multiple emission line + thermal Bremsstrahlung continuum (Lisse et al., 2001) could also produce acceptable fits to the data, we argue that the observed “continuum” could be understood to be the blending of many weaker lines, and is not a self-consistent or physically viable model. Specifically, we have combined the theoretical predictions of the CXE model of Bodewits et al. (2004, 2006, 2007) with a 1-D Haser model of the neutral coma gas density (Qneutral = QH2 O + QOH + QCO2 + QCO + QH + QC + QO ; Fig. 6). The Bodewits model provided the line energies in the 300–1000 eV energy range from solar wind minor ion species CV, CVI, NVI, NVII, OVII, OVIII, and Ne IX. The resulting CXE model has 34 lines, but we included only the 11 most significant in our modeling (and no continuum): those based on the strongest lines expected from the solar wind composition for our CXE model, as well as the most statistically significant lines derived from fitting spectra also be those first determined from the very strong June 30th flare spectrum. The data were then fit at these energies in XSPEC as a series of Gaussian emission lines, each at a fixed line energy and set to the ACIS-S3 instrument resolution. Model fits from all seven of the Chandra pointings on T1 spectral observations were compared to determine the significant emission lines in a consistent manner. Fig. 6. Neutral coma gas model used for the spectral modeling. Line styles are: solid black—water; dashed black—CO; dotted black—H; solid gray—OH; dotted gray—C; dash-dotted gray—O. QH2 O = 1 × 1028 mol s−1 , τH2 O = 1.3 × 105 s; QCO = 0.12 × QH2 O , τCO = 9.0 × 105 s; QCO2 = 0.05 × QH2 O , τCO2 = 3.3 × 105 s. Lifetimes are from Huebner et al. (1992) for an active Sun, scaled to 1.51 AU. Note that the neutral abundance is dominated by water and its dissociation products H2 O, OH, and O throughout the coma. We present results of the best fitting multi-emission line model in Table 3. The number of degrees of freedom of the fit, equivalent to the number of independent spectral channels in the Chandra observations minus the fit parameters, was 36. NH , the neutral hydrogen column density between Chandra and the comet, was set to 1013 cm−2 (i.e., effectively to zero compared Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.11 (1-15) Chandra-DI observations of Comet 9P/Tempel 1 11 to typical columns of 1019 cm−2 or greater found in astrophysical X-ray sources) for the models, as expected for the density of hydrogen in interplanetary space. The strongest emission features in the T1 spectrum are due to the OVII multiplex between 550–580 eV. Other relatively strong emission features found were the NVII/CVI and CV lines between 350 and 460 eV. Faint lines from NVII (500 eV) and Ne IX (922 eV) were also detected. This is the first time NVII lines have been definitively detected, and only the 3rd comet for which the NeIX lines were found. Spectra observed during epochs of intensified solar wind, namely June 30, July 8th and July 24th also showed an additional line at ∼605 eV. This line may possibly be the NVII Ly-β transition at 595 eV, but from the CXE model this line is expected to be weak and only ∼10% of the NVII 500 eV Ly-α emission. The June 30th flare spectrum has over 150 counts per channel in the OVII line and this spectrum was of good signal-tonoise. Arguably, several of the fainter spectra (July 5, 10th) can be modeled using only 6 or 7 emission lines, but we have included all lines from our CXE model to consistently compare spectra of the different observations and for comparison to previous results in the literature (e.g., Lisse et al., 2001, 2005a, 2005b; Krasnopolsky et al., 2002; Krasnopolsky, 2006; Kharchenko et al., 2003) The large errors (or upper limits) of these fainter lines are reflected in the summary of fitting results in Table 3. We note that an additional emission line with energy 270– 285 eV could be included in the spectral models. However, the Chandra effective area in this region is low because of the instrument’s carbon edge. This fact, coupled with the rising X-ray background makes the emission features below 300 eV suspect and we limited the spectra to start at above this region. This limitation did not cause us to lose any line identifications, while making our modeling more robust by increasing the measured S/N ratio. 4.3.1. X-ray spectra and charge state ratios Our Chandra observations sampled several very different solar wind environments (Sections 3.1 and 3.2) and thus provide at least two different spectral signatures with which to test the CXE model. Because the comet remained collisionally thin throughout the observations, compositional variations in the solar wind directly result in spectral variations. Charge exchange emission from H-like ions is dominated by strong Ly-α emission, or, equivalently for He-like ions, by the forbiddenintercombination-resonance feature (FIR). From state selective electron capture cross-sections we derived velocity dependent emission cross-sections for all emission lines of interest. These cross-sections allowed us to extract relative ion abundances from the observed spectra based on the proton speeds derived from the ACE data (Bodewits et al., 2007). Comparison of the ion abundances derived from the X-ray emission to those measured by ACE and propagated back to the comet are given in Table 4 and Fig. 5b. As discussed above, the solar wind was highly variable during the T1 observations, and this resulted in large variations in the observed abundances for C, N, O. For C and O, the obtained values agree within the uncertainties with averaged slow solar wind ion abundances, such as those listed in Schwadron and Cravens (2000) (0.35 and 1.59 for O8+ and C6+ relative to O7+ , respectively). The C6+ ion ratio appears to track the ACE SW data especially well. The carbon ion, C6+ , has the best statistical accuracy of the three ions presented in Fig. 5, and demonstrates the best agreement with the Schwadron and Cravens (2000), Beiersdorfer et al. (2003), and Otranto et al. (2006) SW trends (see Table 4). The derived N7+ abundances (0.3–0.9 relative to O7+ ), however, are consistently high by almost an order of magnitude as compared to the average solar wind values. The O+8 abundances are less problematic, but still high by about a factor of 2 compared to the Schwadron and Cravens (2000) estimates and the analysis results of Chandra observations of C/McNaught– Hartley by Krasnopolsky et al. (2002). While at first apparently alarming, careful study of our methodology has not been able to find an error in our spectral modeling that could account for these discrepancies—the required errors in the CXE cross sections are too large compared to the uncertainty of the measurements (Beiersdorfer et al., 2003; Otranto et al., 2006). Other issues may be in play. For example, our work with solar wind trending data (von Steiger et al., 2000; Zurbuchen and Richardson, 2007) has determined that the abundance ratios of the relatively rare ions used in Schwadron and Cravens (2000), such as O8+ and N7+ , have not been determined to better than an order of magnitude to date (due to poor counting statistics). Also, the Schwadron and Cravens (2000) values for average solar wind were derived on a time-interval in a very different part of the solar cycle, with different average compositions. Further, the minor ion abundances in the SW are highly variable, so that the value of average SWabundance /RMS SWabundance rarely exceeds the value of 1 over the course of a day (see Fig. 1). Given the faintness of Tempel 1 in the X-ray, however, and the relatively large uncertainties of our ion charge state estimates (Fig. 5b, Table 3) we do not feel the issue can be solved with the present dataset. We raise the issue here as one requiring further study, which is now possible, using spectral analyses of other, brighter comets observed by Chandra (Krasnopolsky, 2006; Bodewits et al., 2007). Overall, the charge state of the solar wind encountered by the comet was close to the relatively hot, dense, slow wind emitted from the Sun’s lower latitudes around the minimum in the solar activity cycle. This stands in strong contrast to Chandra results of Comet 2P/Encke on 24 Nov 2003 (Lisse et al., 2005b), where we measured a very unusual X-ray spectrum with approximately equal strength CVI and OVII CXE lines in the 300–700 eV range during a period of mixed slow and fast winds. In fact, the new Tempel 1 results, for another collisionally thin system, now allows us to empirically separate behavior induced by strong changes in the incident solar wind charge state from that due to a collisionally thin coma. Tempel 1 showed a “normal” spectrum (i.e., a ∼3:1 line ratio of the 560 eV/400 eV complexes) for most of the time it was observed. Reassuringly, however, the spectrum of July 5th, 2005 appears to be rather like that seen for Encke in 2003. Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR 12 AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.12 (1-15) C.M. Lisse et al. / Icarus ••• (••••) •••–••• Table 3 0.3–1.0 keV spectral fits to the June–July 2005 Chandra ACIS-S observations Eline Line fluxes (10−5 ph cm−2 s−1 ) and 90% C.L. (1.6σ ) (eV) June 30a June 30b July 04 July 05 July 08b July 11 July 13 July 24 Line ID 307 367.5 419.8 500.3 568d 653.5 712.7 774.6 849.1 922.0 358 ± 60 18.7 ± 5.6 <6.0 3.2 ± 1.8 8.8 ± 1.2 1.6 ± 0.7 <1.0 1.0 ± 0.5 0.5 ± 0.2 0.6 ± 0.3 508 ± 100 33.4 ± 7.4 7.7 ± 7.0 6.8 ± 3.0 24.2 ± 3.0 5.7 ± 1.4 1.5 ± 1.1 1.2 ± 0.7 1.2 ± 0.5 0.8 ± 0.5 112 ± 40 8.8 ± 4.6 4.0 ± 3.0 3.0 ± 1.0 4.9 ± 1.0 2.2 ± 0.5 1.0 ± 0.5 0.9 ± 0.3 0.8 ± 0.2 0.8 ± 0.2 174 ± 60 3.9 ± 6.0 4.5 ± 3.8 <2.5 5.1 ± 1.1 1.9 ± 0.7 1.0 ± 0.6 1.2 ± 0.4 1.0 ± 0.2 1.0 ± 0.3 490 ± 190 41 ± 19 <25 7.0 ± 1.0 27.8 ± 5.0 12.6 ± 4.0 5.3 ± 3.5 <3.0 3.4 ± 1.0 3.0 ± 1.7 – <8.0c 7.0 ± 4.0 2.3 ± 1.6 5.5 ± 0.7 3.5 ± 1.0 1.4 ± 0.8 1.2 ± 0.6 1.5 ± 0.2 0.7 ± 0.4 – 8.0 ± 6.0 <6.0 2.2 ± 1.7 3.3 ± 1.2 1.9 ± 0.8 1.0 ± 0.7 0.9 ± 0.5 <0.5 0.7 ± 0.3 – <10.0 5.2 ± 4.0 4.0 ± 2.0 7.4 ± 6.0 2.5 ± 1.0 1.1 ± 0.7 0.8 ± 0.6 0.5 ± 0.3 0.6 ± 0.4 CV f+r+i CVI Ly-α NVI f+r+i NVII Ly-α OVII f+r+i OVIII Ly-α OVII K-ε OVIII Ly-γ OVIII Ly-ζ NeIX f+r+i 0.9 1.2 1.2 1.1 0.7 1.1 1.2 1.1 χ 2 /dof a b c d Quiet state spectrum only. Flare/high state spectrum. ‘<’ indicates the 90% confidence upper limits used throughout the table. Average of OVII 561.1, 568.4 and 573.96 eV lines. Table 4 Solar wind ion abundancesc derived from the spectral fits in Table 3 Ion O8+ C6+ N7+ Date June 30a June 30b July 04 July 05 July 08a July 11 July 13 July 24 S&Cd Bei 03d Kra 04d Otr 06d 0.7 ± 0.28 3.0 ± 2.45 0.9 ± 0.4 0.4 ± 0.09 1.2 ± 0.78 0.4 ± 0.13 1.0 ± 0.22 1.5 ± 1.82 0.8 ± 0.3 1.0 ± 0.3 0.8 ± 2.2 0.4 ± 0.3 0.7 ± 0.2 1.1 ± 1.8 0.3 ± 0.4 1.2 ± 0.2 0.7 ± 1.5 0.5 ± 0.3 1.4 ± 0.5 2.1 ± 2.9 0.9 ± 0.6 0.5 ± 0.5 0.6 ± 2.0 0.7 ± 0.5 0.35 1.59 0.03 0.13 ± 0.03 0.90 ± 0.9 0.06 ± 0.02 0.13 ± 0.05 0.7 ± 0.2 – 0.35 1.02 0.03 a b c d Quiet state spectrum only. Flare/high state spectrum. Solar wind abundances (relative to O7+ ) derived from spectra ±90% C.L. Slow wind model parameters from Schwadron and Cravens (2000); Bei 03 = Beiersdorfer et al. (2003); Kra 04 = Krasnopolsky (2004); Otr 06 = Otranto et al. (2006). 4.3.2. Pre- and post-impact spectral behavior of Tempel 1 Because of the flare during the June 30th observation, which had been intended by our experimental design to establish the quiescent baseline value of the comets emission, our best estimate of the pre-impact spectral behavior of the comet comes from the day of impact, July 4th. We thus extracted the 5 ks of spectral data from the start of the July 4th observation just before the impact and compared it to a spectrum of 20 ks duration taken after impact and near the end of the July 4th observation (Fig. 7). While we see the ∼20% increase reported in the lightcurve photometry due to the impact in summing the 300– 1000 eV X-ray spectra from T1, with a 90% confidence range of 0.8 - 1.6 times the mean pre-impact count rate (equal to 2.7 × 10−13 ergs cm−2 s−1 or LX = 4.6 × 1014 ergs s−1 ), no significant spectral changes are observed in the 500–700 eV region, and although the post-impact spectrum is higher between 750–900 eV, the poor signal-to-noise makes it impossible to draw any definite conclusions. These may be consistent with the observed solar wind variations (Fig. 5b), due to the passing of a shock flank by the comet on late July 4th (Section 3.2). 5. Discussion and conclusions Here we summarize our findings from observing Tempel 1 during the extensive Chandra Deep Impact observing cam- paign. First we list the morphological, temporal, and spectral results far in time from the impact. This information gives us the comet’s behavior in the X-ray and sets the stage for the July 4th Deep Impact experiment. We then discuss the changes observed due to the impact, and compare the influence of the man-made cratering experiment with the two “natural” outbursts found. 5.1. General results • In terms of measured count rate, Tempel 1 is the second faintest comet ever detected in the X-ray. In terms of 300– 1000 eV X-ray luminosity, it ranks 4th faintest. In terms of total on-target observing time, it is the most studied comet to date. • Tempel 1 in June–July 2005 is the first comet to be observed simultaneously and repeatedly by two different Xray instruments—the Chandra ACIS-S and the SWIFT HRT. Comparison of the measured photometric fluxes and temporal trends for the two datasets shows good agreement. • The standard CXE model works for this very well studied comet. The observed X-ray behavior (morphology, lightcurve, and spectrum) for Tempel 1 can be explained with the collisionally thin model used for 2P/Encke 2003 (Lisse et al., 2005b; Bodewits et al., 2007). The relatively normal line ratios of the Tempel 1 spectrum, when com- Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.13 (1-15) Chandra-DI observations of Comet 9P/Tempel 1 13 were much stronger in intensity and total flux than that produced by the Deep Impact experiment. • The observed spectrum in and out of flare varies, while the observed morphology does not (except for the June 30th flare, where a solar wind impulse coincided with a comet neutral gas outburst). The differences in the spectrum can be explained by variation in the incident charge state ratio of the solar wind. 5.2. Deep Impact experiment results Fig. 7. Comparison of the pre- and post-impact spectra from 04 July. Shown are the spectra (with ±1σ error bars) with pre-impact in open squares and post-impact in filled circles, over plotted with the best fitting multi-emission line CXE models (dashed for pre-impact and solid line for post-impact). The post-impact spectrum has been scaled to the pre-impact levels; an overall change in the 300–1000 eV total flux of 18 ± 5 (1σ )% was found. Marginal differences in the pre- and post-spectra are seen, if at all, in the 800–900 eV region. pared to the unusual spectrum observed for 2P/Encke on 24 Nov 2003, allow us to determine that the source of this unusual spectrum was the abnormal charge state of the solar wind ions incident at Encke during the time of observation, not the low neutral number density in the comet’s coma. • The narrow June 30th flare coincides with detections of dramatically increasing outgassing rates by the Deep Impact (A’Hearn et al., 2005; Farnham et al., 2007) and Lowell Observatories (Schleicher et al., 2006), followed by an increased solar wind flux. Roughly 50% of the increased emission during this flare was due to increases in the solar wind flux incident on the comet; the rest is due to increasing cometary outgassing. While the outgassing effect was difficult to detect, as jumps in observed X-ray emission rates from fluctuations in the incident solar wind flux were large (i.e. much larger than those seen due to the material emitted due to the DI experiment), long term monitoring of the comet over 75 ks shows a definite trend attributable to an outburst in neutral gas from the comet. The extent of the detectable X-ray emitting region increased by 50 ± 25%. These results represent the first demonstration of a direct link between rapid changes in a comet’s Qgas (t) and the observed X-ray flux. • The days-long, wide X-ray flare from the comet starting on July 8 is solar wind driven, and can be accounted for entirely by changes in the solar wind time record. The July 8th X-ray flare did not increase the apparent extent of the X-ray emitting region over the range seen during quiescent periods. Both naturally-occurring X-ray flares from the comet • No definitive X-ray flash was observed at impact. It is not clear if this is due to an intrinsic effect from the impact, or the sensitivity of the Chandra/ACIS-S camera. • The overall emission increased by 18 ± 5% over the course of July 4th and stayed elevated through July 5. The increase in total neutral coma gas during the same time was ∼30%. There is no evidence for a huge change in the comet’s Xray emission behavior, moving the comet from collisionally thin (to CXE) to collisionally thick behavior, as predicted by some pre-impact estimates (Lisse et al., 2005a, 2005b, 2005c; Schultz et al., 2005; Fig. 8). Much of the reason for the small magnitude of the change is the lack of production of a new active area, and the relatively small total amount of material, 106 –107 kg, liberated by the impact. • No large spectral shifts in the emission were seen due to the impact. The level of the observed changes is within the range of statistical uncertainty. • Extended emission along the path taken by the excavated material was observed at the 5σ level. The detection is not strong enough to determine if the emission rate from the extension is changing with time, as would be expected from an extended source of neutral gas, such as outgassing from ejected water ice. • There is no evidence for a late, large upsurge in material emitted from the comet into the coma due to the Deep Impact experiment, as reported by Willingale et al. (2006) based on the analysis of SWIFT data; instead, the broad increase in X-ray flux seen in the combined Chandra/SWIFT X-ray lightcurves can be explained by increases in the solar wind flux at the comet. The long term temporal behavior of the X-ray luminosity, before, during, and after the Deep Impact experiment, can be matched using measures of the coma neutral gas density (SOHO/SWAN, ROSETTA, ODIN, ESO CN, CALAR ALTO, SST, CARA) that show an elevated density of material for the first two days after impact, then normal cometary behavior afterwards. There is also no change in the emission morphology seen on July 8th, corresponding to that found on July 4th, during the outgassing event reported by SWIFT. Acknowledgments We would like to thank V. Kharchenko, T. Sasseen, M. Torney, D. Willingale, and the SWIFT project for many helpful discussions, as well as the anonymous reviewers whose comments greatly improved the manuscript. The SOHO CELIAS Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR 14 AID:8214 /FLA YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.14 (1-15) C.M. Lisse et al. / Icarus ••• (••••) •••–••• heavy ion and MTOF proton monitor data were graciously provided by the University of Maryland SOHO project, URL http:// umtof.umd.edu/. We are grateful for the cometary ephemerides of D.K. Yeomans, et al., found at URL http://ssd.jpl.nasa.gov/ horizons.html used to reduce our data. Help with the MHD modeling was provided by Y. Jia at the University of Michigan. C. Lisse and D. Christian were supported in part by SMAO observing Grants NAG GO45167X, and D. Bodewits and R. Hoekstra would like to acknowledge support from the FOMEURATOM association agreement. References A’Hearn, M.F., and 32 colleagues, 2005. Deep Impact: Excavating Comet Tempel 1. Science 310, 258–264. Arnaud, K.A., 1996. XSPEC: The first ten years. In: Jacoby, G., Barnes, J. (Eds.), Astronomical Data Analysis Software and Systems V. In: ASP Conf. Series, vol. 101, pp. 17–20. Beiersdorfer, P., and 10 colleagues, 2003. Laboratory simulation of charge exchange-produced X-ray emission from comets. Science 300, 1558–1560. Bodewits, D., Juhasz, Z., Tielens, A.G.G.M., Hoekstra, R., 2004. Catching some Sun: Probing the solar wind with cometary X-ray and far-ultraviolet emission. Astrophys. J. 606, L81–L84. Bodewits, D., Hoekstra, R., Seredyuk, B., McCullough, R.W., Jones, G.H., Tielens, A.G.G.M., 2006. Charge exchange emission from solar wind helium ions. Astrophys. J. 642, 593–605. Bodewits, D., Christian, D.J., Torney, M., Dryer, M., Lisse, C.M., Dennerl, K., Wolk, S.J., Tielens, A.G.G.M., Hoekstra, R., 2007. Spectral analysis of the Chandra Comet Survey. Astron. Astrophys., in press. astro-ph/0704.1648. Cravens, T.E., 1997. Comet Hyakutake X-ray source: Charge transfer of solar wind heavy ions. Geophys. Res. Lett. 24, 105–108. Dennerl, K., Englhauser, J., Trümper, J., 1997. X-ray emissions from comets detected in the Rőntgen X-ray satellite All-Sky Survey. Science 277, 1625– 1630. Dennerl, K., Aschenbach, B., Burwitz, V., Englhauser, J., Lisse, C.M., Rodriguez-Pascual, P.M., 2003. A major step in understanding the X-ray generation. In: Trümper, J.E., Tananbaum, D.H. (Eds.), Comets: Recent Progress Obtained with XMM-Newton, X-Ray, and Gamma-Ray Telescopes and Instruments for Astronomy. In: Proceedings of SPIE, vol. 4851, pp. 277–288. Dryer, M., 1994. Interplanetary studies: Propagation of disturbances between the Sun and the magnetosphere. Space Sci. Rev. 67 (3/4), 363–419. Dryer, M., 1998. Multi-dimensional MHD simulation of solar-generated disturbances: Space weather forecasting of geomagnetic storms. Am. Inst. Aeronaut. Astronaut. 36 (3), 365–370. Dryer, M., Fry, C.D., Sun, W., Deehr, C.S., Smith, Z., Akasofu, S.-I., Andrews, M.D., 2001. Prediction in real-time of the 2000 July 14 heliospheric shock wave and its companions during the “Bastille” epoch. Solar Phys. 204, 267– 286. Dryer, M., Fry, C.D., Sun, W., Deehr, C.S., Akasofu, S.-I., 2004. Real-time predictions of interplanetary shock arrivals at L1 during the “Halloween 2003” Epoch. Space Weather 2, doi:10.1029/2004SW000087. S09001. Farnham, T.L., and 11 colleagues, 2007. Dust coma morphology in the Deep Impact images of Comet 9P/Tempel 1. Icarus 187, 26–46. Fry, C.D., Sun, W., Deehr, C.S., Dryer, M., Smith, Z., Akasofu, S.-I., Tokumaru, M., Kojima, M., 2001. Improvements to the HAF solar wind model for space weather predictions. J. Geophys. Res. 106, 20985–21001. Fry, C.D., Dryer, M., Deehr, C.S., Sun, W., Akasofu, S.-I., Smith, Z., 2003. Forecasting solar wind structures and shock arrival times using an ensemble of models. J. Geophys. Res. 108 (A2), doi:10.1029/2002JA009474. 1070. Fry, C.D., Dryer, M., Sun, W., Detman, T.R., Smith, Z., Deehr, C.S., Wu, C.-C., Akasofu, S.-I., Berdichevsky, D., 2004. Solar observation-based model for multi-day predictions of interplanetary shock and CME arrivals at Earth. IEEE Trans. Plasma Phys. 32 (4), 1489–1497. Part I of III. Gloeckler, G., and 11 colleagues, 1998. Investigation of the composition of solar and interstellar matter using solar wind and pickup ion measurements with SWICS and SWIMS on the ACE spacecraft. Space Sci. Rev. 86, 497– 539. Huebner, W.F., Keady, J.J., Lyon, S.P., 1992. Solar photo rates for planetary atmospheres and atmospheric pollutants. Astrophys. Space Sci. 195, 1–289. 291–294. Jehin, E., and 9 colleagues, 2006. Deep Impact: High resolution optical spectroscopy with the ESO VLT and the Keck 1 Telescope. Astrophys. J. 641, L145–L148. Johnstone, A.D., Coates, A.J., Huddleston, D.E., Jockers, K., Wilken, B., Borg, H., Gurgiolo, C., Winningham, J.D., Amata, E., 1993. Observations of the solar wind and cometary ions during the encounter between Giotto and Comet P/Grigg–Skjellerup. Astron. Astrophys. 273, L1–L4. Keller, H.U., and 11 colleagues, 2005. Deep Impact observations by OSIRIS onboard the Rosetta spacecraft. Science 310, 281–283. Kharchenko, V., Dalgarno, A., 2001. Variability of cometary X-ray emission induced by solar wind ions. Astrophys. J. 554, L99–L102. Kharchenko, V., Rigazio, M., Dalgarno, A., Krasnopolsky, V.A., 2003. Charge abundances of the solar wind ions inferred from cometary X-ray spectra. Astrophys. J. 585, L73–L75. Krasnopolsky, V.A., 2004. Comparison of X-rays from Comets LINEAR (C/1999 S4) and McNaught–Hartley (C/1999 T1). Icarus 167, 417–423. Krasnopolsky, V.A., 2006. X-rays and solar wind composition in four comets observed with Chandra X-ray Observatory. J. Geophys. Res. 111, A12102– A12109. Krasnopolsky, V.A., Mumma, M.J., Abbott, M.J., 2000. EUVE search for Xrays 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., Stern, S.A., 2002. X-ray emission from Comet McNaught– Hartley (C/1999 T1). Icarus 160, 437–447. Küppers, M., and 40 colleagues, 2005. A large dust/ice ratio in the nucleus of Comet 9P/Tempel 1. Nature 437, 987–990. Lebedev, M.G., 2000. Comet Grigg–Skjellerup atmosphere interaction with the oncoming solar wind. Astrophys. Space Sci. 274, 221–230. Lisse, C.M., and 11 colleagues, 1996. Discovery of X-ray and extreme ultraviolet emission from Comet C/Hyakutake 1996 B2. Science 274, 205–209. Lisse, C.M., Christian, D., Dennerl, K., Desch, M., Englhauser, J., Marshall, F.E., Petre, R., Snowden, S., Truemper, J., 1999. X-ray and extreme ultraviolet emission from Comet 2P/Encke 1997. Icarus 141, 316–330. Lisse, C.M., Christian, D.J., Dennerl, K., Meech, K., Petre, R., Weaver, H.A., Wolk, S.J., 2001. Charge exchange-induced X-ray emission from Comet C/1999 S4 (LINEAR). Science 292, 1343–1348. Lisse, C.M., Cravens, T.E., Dennerl, K., 2005a. X-ray and extreme ultraviolet emission from Comets. In: Festou, M.C., Weaver, H.A., Keller, H.U. (Eds.), Comets II. Univ. of Arizona Press. Lisse, C.M., Christian, D.J., Dennerl, K., Wolk, S.J., Bodewits, D., Hoekstra, R., Combi, M.R., Mäkinen, T., Dryer, M., Fry, C.D., Weaver, H., 2005b. Chandra observations of Comet 2P/Encke 2003: First detection of a collisionally thin, fast solar wind charge exchange system. Astrophys. J. 635, 1329–1347. Lisse, C.M., A’Hearn, M.F., Groussin, O., Fernandez, Y.R., Belton, J.M., van Cleve, J.E., Charmandaris, V., Meech, K.J., McGleam, C., 2005c. Rotationally resolved 8–35 um Spitzer Space Telescope observations of the nucleus of Comet 9P/Tempel 1. Astrophys. J. 625, L139–L142. Lisse, C.M., and 16 colleagues, 2006. Spitzer spectral observations of the Deep Impact ejecta. Science 313, 635–640. Mäkinen, J.T.T., Silén, J., Schmidt, W., Kyrölä, E., Summanen, T., Bertaux, J.-L., Quémerais, E., Lallement, R., 2001. Water production of Comets 2P/Encke and 81P/Wild 2 derived from SWAN observations during the 1997 apparition. Icarus 152, 268–274. Mäkinen, J.T.T., Combi, M.R., Bertaux, J.L., Quemerais, E., Schmidt, W., 2007. SWAN observations of 9P/Tempel around the Deep Impact event. Icarus 187, 109–112. Meech, K.J., and 208 colleagues, 2005. Deep Impact: Observations from a Worldwide Earth-Based Campaign. Science 310, 265–269. Mumma, M.J., Krasnopolsky, V.A., Abbott, M.J., 1997. Soft X-rays from four comets observed with EUVE. Astrophys. J. 491, L125–L128. Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 ARTICLE IN PRESS JID:YICAR AID:8214 /FLA Neugebauer, M., Cravens , T.E., Lisse, C.M., Ipavich, F.M., Christian, D., von Steiger, R., Shah, P.D., Armstrong, T.P., 2000. The relation of temporal variations of soft X-ray emission from Comet Hyakutake to variations of ion fluxes in the solar wind. J. Geophys. Res. 105, 20949–20955. Otranto, S., Olson, R.E., Beiersdorfer, P., 2006. X-ray emission cross-sections following charge exchange by multiply charged ions of astrophysical interest. Phys. Rev. A 73, 0222723–0222731. Prange, R., Pallier, L., Hansen, K.C., Howard, R., Vourlidas, A., Courtin, R., Parkinson, C., 2004. An interplanetary shock traced by planetary auroral storms from the Sun to Saturn. Nature 432, 78–81. Sasseen, T., Hurwitz, M., Lisse, C.M., Kharchenko, V., Christian, D., Wolk, S.J., Sirk, M.M., Dalgarno, A., 2006. A search for extreme-ultraviolet emission from comets with the Cosmic Hot Interstellar Plasma Spectrometer (CHIPS). Astrophys. J. 650, 461–469. Schleicher, D.G., Barnes, K.L., Baugh, N.F., 2006. Photometry and imaging results for Comet 9P/Tempel 1 and Deep Impact: Gas production rates, postimpact light curves, and ejecta plume morphology. Astron. J. 131, 1130–1137. Schultz, P.H., Ernst, C., Anderson, J.L.B., 2005. Expectations for crater size and photometric evolution from the Deep Impact collision. Space Sci. Rev. 117, 207–239. Schultz, P.H., Eberhardy, C.A., Ernst, C.M., Sunshine, J.M., A’Hearn, M.F., Lisse, C.M., 2007. Impact cratering at Comet 9P/Tempel 1 by Deep Impact. Icarus (this issue), in press. Schwadron, N., Cravens, T.E., 2000. Implications of solar wind composition for cometary X-rays. Astrophys. J. 544, 558–566. Smith, Z., Dryer, M., 1991. Numerical simulations of high-speed solar wind streams within 1 AU and their signatures at 1 AU. Solar Phys. 131, 363– 383. Stone, E.C., and 16 colleagues, 1998. The solar isotope spectrometer for the advanced composition explorer. Space Sci. Rev. 86, 357–408. Sugita, S., and 22 colleagues, 2005. Subaru telescope observations of Deep Impact. Science 310, 274–278. von Steiger, R., Schwadron, N.A., Fisk, L.A., Geiss, J., Gloeckler, G., Hefti, S., Wilken, B., Wimmer-Schweingruber, R.F., Zurbuchen, T.H., 2000. Composition of quasi-stationary solar wind flows from Ulysses/solar wind ion composition spectrometer. J. Geophys. Res. 105, 27217– 27238. Waite, J.H., and 11 colleagues, 1995. ROSAT observations of X-ray emissions from Jupiter during the Impact of Comet Shoemaker–Levy 9. Science 268, 1598–1601. Wegmann, R., Dennerl, K., 2005. X-ray tomography of a cometary bow shock. Astron. Astrophys. 430, L33–L36. Wegmann, R., Dennerl, K., Lisse, C.M., 2004. The morphology of cometary X-ray emission. Astron. Astrophys. 428, 647–661. Willingale, R., and 16 colleagues, 2006. Swift X-ray telescope observations of the Deep Impact collision. Astrophys. J. 649, 541–552. Zurbuchen, T.H., Richardson, I.G., 2007. In situ solar wind and magnetic field signatures of interplanetary coronal mass ejections. Space Sci. Rev., in press. Zwickl, R.D., and 11 colleagues, 1998. The NOAA Real-Time Solar-Wind (RTSW) system using ACE data. Space Sci. Rev. 86, 633–648. Please cite this article in press as: C.M. Lisse et al., Chandra observations of Comet 9P/Tempel 1 during the Deep Impact campaign, Icarus (2007), doi:10.1016/j.icarus.2007.03.004 View publication stats YICAR:8214 [m5+; v 1.72; Prn:30/05/2007; 13:37] P.15 (1-15) Chandra-DI observations of Comet 9P/Tempel 1 15