Line-of-sight gas sloshing in the cool core of Abell 907
aa r X i v : . [ a s t r o - ph . H E ] D ec Draft version December 20, 2018
Typeset using L A TEX twocolumn style in AASTeX62
Line-of-sight gas sloshing in the cool core of Abell 907
Shutaro Ueda, Yuto Ichinohe, Tetsu Kitayama, and Keiichi Umetsu Academia Sinica Institute of Astronomy and Astrophysics (ASIAA), No. 1, Section 4, Roosevelt Road, Taipei 10617, Taiwan Department of Physics, Rikkyo University, 3-34-1 Nishi-Ikebukuro, Toshima-ku, Tokyo 171-8501, Japan Department of Physics, Toho University, 2-2-1 Miyama, Funabashi, Chiba 274-8510, Japan (Received; Revised; )
Submitted to ApJABSTRACTWe present line-of-sight gas sloshing first found in a cool core in a galaxy cluster. The galaxycluster Abell 907 is identified as a relaxed cluster owing to its global X-ray surface brightness takenby the
Chandra X-ray Observatory . The X-ray residual image after removing the global emissionof the intracluster medium (ICM), however, shows an arc-like positive excess and a negative excesssurrounding the central positive excess in the cluster core, which in turn indicates a disturbanceof the ICM. We analyze the X-ray spectra extracted from both regions and find that (1) the ICMtemperature and the metal abundance in the positive excess are lower and higher than those in thenegative excess, respectively, and (2) the ICM is nearly in pressure equilibrium. We also find a slightredshift difference between the positive and the negative excesses, which corresponds to the velocityshear of 1680 +1300 − km s − (1 σ ). The X-ray residual image and the ICM properties are consistent withthose expected by line-of-sight gas sloshing. Assuming that the gas is moving toward inverse-parallelto each other along the line-of-sight, the shear velocity is expected to be ∼
800 km s − . The velocityfield of this level is able to provide non-thermal pressure support by ∼
34% relative to the thermalone. The total kinetic energy inferred from the shear velocity corresponds to ∼
30 % of the bolometricluminosity of the sloshing ICM. Abell 907 is therefore complementary to galaxy clusters in which gassloshing takes place in the plane of the sky, and is important for understanding gas dynamics drivenby sloshing and its influence on the heating to prevent runaway cooling.
Keywords: galaxies: clusters: individual: (Abell 907) — X-rays: galaxies: clusters — galaxies: clus-ters: general INTRODUCTIONGalaxy clusters are dynamically young systems in theuniverse. A large amount of baryons fills in the gravi-tational potential well of galaxy clusters that is formedby dark matter. Most of the baryons in galaxy clus-ters are in the form of X-ray emitting hot gas so-calledthe intracluster medium (ICM). A cool core composedof dense, cool, and metal rich ICM is often found in thecenter of galaxy clusters. It is considered that cool coresare formed by radiatively cooling gas, because of thefact that their cooling time estimated by electron den-
Corresponding author: Shutaro [email protected] sity in the core is much shorter than the age of galaxyclusters (e.g., Peterson & Fabian 2006). The abundanceand properties of cool cores in galaxy clusters provide uswith a wealth of information about not only the thermalevolution of cosmic baryons, but also the chemical evo-lution of the ICM around the brightest cluster galaxy(BCG).The presence of cool cores in galaxy clusters, on theother hand, poses us a challenge regarding the thermalevolution of intracluster baryons. The cooling time ofX-ray emitting hot gas is inversely proportional to itselectron density squared. Since the gas in the cool coreis dense, its cooling time is much shorter than the age ofgalaxy clusters (e.g., Peterson & Fabian 2006). This in-dicates that the gas in the cool core is not able to survivestably for a long time without heating. This is recog-
Ueda et al. nized as the cooling problem of the ICM. To maintainthe balance between cooling and heating, the feedback ofactive galactic nuclei (AGN) in the BCG is considered tobe a plausible mechanism (see e.g., McNamara & Nulsen2007; Fabian 2012, for reviews). In fact, X-ray cavi-ties are found in the X-ray surface brightness of a largesample of galaxy clusters (e.g., Hlavacek-Larrondo et al.2015). Some of them seem to be associated with jet-likeradio emissions (e.g., McNamara et al. 2005), which in-dicates that mechanical energy provided by AGN is adominant heating source. On the other hand, gas slosh-ing in the core of galaxy clusters is expected to be an-other possible heating source. Gas sloshing is inducedby minor mergers with a non-zero impact parameter (seeMarkevitch & Vikhlinin 2007, for a review). Numericalsimulations show that gas sloshing is able to prevent run-away cooling for at least a few Gyrs (e.g., Fujita et al.2004; ZuHone et al. 2010).Evidence of gas sloshing is often found in relaxedclusters as a spiral pattern in the residual image ofX-ray surface brightness after removing its globalprofile (e.g., Churazov et al. 2003; Clarke et al. 2004;Blanton et al. 2011; Owers et al. 2011; O’Sullivan et al.2012; Canning et al. 2013; Rossetti et al. 2013; Ghizzardi et al.2014; Sanders et al. 2014; Ichinohe et al. 2015; Ueda et al.2017; Liu et al. 2018). The observed spiral pattern islikely caused by a merger occurring nearly in the planeof the sky, through transport of angular momentumfrom an infalling galaxy cluster. Since such mergerscan occur in all directions, the resulting sloshing planesshould be distributed uniformly, and sloshing motionsshould be found not only in the plane of the sky but alsoalong the line-of-sight (LOS). Although a large fractionof LOS gas sloshing is therefore expected, only a singlecase in the Virgo cluster has been reported thus far(Roediger et al. 2011). This suggests that a large num-ber of LOS gas sloshing systems have been overlookedor misidentified. Recently, Su et al. (2017) studied theFornax cluster focusing on the Kelvin-Helmholtz insta-bility (KHI) and showed that gas sloshing is in facta reasonable mechanism to suspend runaway cooling.In addition to gas mixing by KHI, the kinetic energyof ICM motions driven by gas sloshing is expected tobe an additional heat source in the ICM through itsdissipation by turbulence. This contribution, however,strongly depends on the velocity of the sloshing ICM.Only systems of LOS gas motions allow us to directlymeasure the shear velocity of sloshing motions. Detailedstudies of galaxy clusters that are experiencing LOS gassloshing are, therefore, essential to understand the wholepicture of gas sloshing, especially in the context of gas dynamics. To this end, the first step is to identify sucha system.Abell 907 (hereafter A907) is located at a redshift of z = 0 .
167 (B¨ohringer et al. 2007). The ICM tempera-ture decreases from ∼ ∼ ∼ ∼ . ∼ × M ⊙ from the Subaru weak-lensingobservations (Okabe & Smith 2016). A907 is thereforeclassified into one of typical massive, relaxed, cool coreclusters. However, the lack of cavities may suggest thatalternative sources of heating are at play. Therefore,A907 is an ideal target to investigate observationallythe role of gas sloshing in preventing runaway cooling inthe core.Throughout the paper, we adopt Ω m = 0 .
27, Ω Λ =0 .
73, and the Hubble constant of H = 70 km s − Mpc − (Komatsu et al. 2011). In this cosmology, an angularsize of 1 ′′ corresponds to a physical scale of 2.87 kpc atthe cluster redshift z = 0 . σ . OBSERVATIONS AND DATA REDUCTIONSWe used the archival X-ray data of A907 takenwith the Advanced CCD Imaging Spectrometer (ACIS;Garmire et al. 2003) on board the
Chandra X-ray Ob-servatory . All three datasets analyzed were taken bythe ACIS-I (ObsID: 535, 3185, and 3205). We used theversions of 4.9 and 4.7.8 for
Chandra
Interactive Anal-ysis of Observations (CIAO; Fruscione et al. 2006) andthe calibration database (CALDB), respectively. Afterapplying lc clean to the data to exclude the durationof flare, we obtained a net exposure of 106.1 ksec. Weadopted the blank-sky data included in the CALDB asour background data. We extracted the X-ray spec-trum of the ICM from interested region of each datasetwith specexctract in CIAO and combined them aftermaking individual spectrum, response, and ancillaryresponse files for the spectral fitting. We used
XSPEC version 12.10.0 (Arnaud 1996) and the atomic databasefor plasma emission modeling version 3.0.9 in the X-rayspectral analysis, assuming that the ICM is in collisionalionization equilibrium. We also used the abundance ta-ble of Lodders & Palme (2009). ANALYSES AND RESULTS3.1.
X-ray imaging analysis ine-of-sight gas sloshing in the cool core of Abell 907 . − . . ± .
01 and its posi-tion angle is 165 ◦ ± ◦ . The position of the center ofthe ellipse model is located at (RA, Dec) = (9:58:21.979,-11:03:50.065), which is ∼ . ′′ offset from that of theBCG. We then subtracted the mean profile from theoriginal X-ray surface brightness. The right panel ofFigure 1 shows the resultant X-ray residual image ofA907. We found a bumpy structure in the X-ray resid-ual image. One is a positive excess with respect to themean surface brightness profile and the other is a nega-tive excess.An arc-like positive excess region is located at a posi-tion close to the cluster center. The length of this regionis ∼
75 kpc (26 ′′ ) and its width is ∼
30 kpc (10 ′′ ). Theangular resolution of Chandra only allows us to detectsuch small structure. The positive excess region is sur-rounded by a negative excess region.In addition, a low-level excess is found in the west sideof A907 as indicated with white arrows in the right panelof Figure 1. We call this a weak positive excess region.This region is along the negative excess region and itsmorphology is arc-like.3.2.
X-ray spectral analysis
To study the origin of the perturbation presented inSection 3.1, we analyzed the X-ray spectra extractedfrom the positive and the negative excess regions. Weshow the regions with the positive and the negative ex-cess in the right panel of Figure 1 as solid white el-lipses and dashed white ellipse, respectively. When weextracted the X-ray spectrum of the negative excess re-gion, we excluded the positive excess region. Figure 2shows the X-ray spectrum of each region and the best-fit model with respect to each X-ray spectrum. The netcounts in these regions are shown in Table 1. In theX-ray spectral analysis, we allowed the column densityof the Galactic absorption ( N H ) to vary in order to re-duce the contamination uncertainty. Table 1 shows thebest-fit parameters derived by the X-ray spectral anal-ysis of both regions. The electron pressure ( kT × n e )and the entropy ( kT × n − / ) that are readily calcu-lated from the best-fit parameters are also presented in The position angle is measured for the major axis of an ellipsefrom north (0 ◦ ) to east (90 ◦ ) Table 1. We assumed the LOS length of 140 kpc in thisanalysis, which is comparable to the projected size ofthe negative excess region. Note that, although the fit-ted values of N H are a factor of ∼ . × cm − ), they areconsistent with one another within large uncertainties.This is likely due to the uncertainty of contaminationcalibration and a low spatial resolution map of N H . Theflux of background data (i.e., cosmic X-ray backgroundand non X-ray background) is two orders of magnitudelower than the source flux, even though the energy bandis 2 . − . σ ) than that of the neg-ative excess region, (2) the abundance in the positiveexcess region is higher (3 σ ) than that in the negativeexcess region, (3) the electron pressure is comparable toeach other ( < σ ), and (4) the entropy in the positiveexcess region is lower (10 σ ) than that in the negativeexcess region. Only statistical errors are considered incalculating the significance levels. In addition, we find aslight redshift difference of ∆ z = 0 . +0 . − . betweenthe positive and the negative excess regions, which cor-responds to ∆ v = 1680 +1310 − km s − . The uncertaintyof the detector gain of the ACIS is estimated at 0 . , which corresponds to ∼
20 eV at 6 keV. A furtherstudy of the gain uncertainty of the ACIS-I was doneby Liu et al. (2015), using the background data insidethe field of view during the observations of the Bulletcluster. They found a gain uncertainty of ∼
10 eV andno spatial variation on the CCD chip in their observa-tions. The measured redshift offset is about 30 eV, whichis larger than the systematic uncertainty. In addition,both regions are close to each other on the CCD chip,so that this implies that the relative gain uncertaintybetween the two regions are expected to be canceled, aspointed out by Liu et al. (2015).We also studied the weak positive excess region byanalyzing the X-ray spectrum extracted from this re-gion. In the right panel of Figure 1, a half ellipse(black dashed) represents the weak positive excess re-gion. The X-ray spectrum of the weak excess region wasextracted after excluding the positive and the negativeexcess regions. The best-fit parameters for this regionare summarized in Table 1. We assumed the LOS depthof 140 kpc in the calculation of the electron density tomatch with the other excess regions. We find that the http://cxc.harvard.edu/cal/summary/Calibration Status Report.html Ueda et al. -7 -6100 kpc : . - : : . Right ascension D ec li n a ti on log I X (photon/s/arcsec /cm ) -2E-07 -1E-07 0 1E-07 2E-07 : . - : : . Right ascension D ec li n a ti on Δ I X (photon/s/arcsec /cm )
100 kpc
Figure 1.
X-ray surface brightness of A907 (left) and its residual image after removing the mean profile (right). Left: TheX-ray surface brightness in the 0 . − . − arcsec − cm − . Thisimage is smoothed by the Gaussian kernel with 2 . ′′ FWHM. Right: The X-ray residual image of A907 after subtracting themean profile from the original X-ray surface brightness (the left panel of this figure). Solid, white ellipses show the region toextract the X-ray spectrum of the positive excess region. A dashed, white ellipse is the negative excess region. A black, dashedhalf ellipse represents the weak positive excess region. The white arrows indicate this region as well.
Table 1.
Best-fit parameters of X-ray spectral analyses in the positive, the negative, and the weak positive excess regions.Region Positive excess Negative excess Weak positiveNet count (0 . − . N H (10 cm − ) 10 . +1 . − . . ± . . ± . . +0 . − . . +0 . − . . +0 . − . Abundance ( Z ⊙ ) 1 . +0 . − . . +0 . − . . +0 . − . Redshift 0 . +0 . − . . +0 . − . . ± . − ( L/
140 kpc) − / ) 0 . ± . . ± . . +0 . − . Pressure (keV cm − ( L/
140 kpc) − / ) 0 . ± .
003 0 . ± .
002 0 . +0 . − . Entropy (keV cm ( L/
140 kpc) / ) 48 . +1 . − . . +2 . − . . +5 . − . ICM temperature of this region is higher than that of thenegative excess region and the abundance is lower thanthat in the positive excess region. In addition, the red-shift of this region is consistent with that of A907 (i.e., z = 0 . DISCUSSION4.1.
LOS gas sloshing
We have measured thermodynamic properties of theICM in the cluster core through the
Chandra observa-tions. We found that the ICM temperature and abun-dance of the positive excess region are lower and higherthan those of the negative excess region, respectively. ine-of-sight gas sloshing in the cool core of Abell 907 −3 no r m a li ze d c oun t s s − k e V − r a ti o Energy (keV)
Figure 2.
X-ray spectra extracted from the positive ex-cess region (red) and the negative excess region (black) withthe best-fit model. The ratios of the data to the model areplotted in the bottom panel.
The electron pressure in both regions suggests that theICM in the core is nearly in pressure equilibrium. All ofthe differences and similarities of the ICM are in goodagreement with the thermodynamic properties expectedby gas sloshing (e,g., Clarke et al. 2004; Blanton et al.2011; Ichinohe et al. 2015; Ueda et al. 2017): i.e., theICM in the positive excess region is comprised of cool,dense, and metal rich gas originally in the cool core andis uplifted toward outside the cool core. On the otherhand, the ICM in the negative excess region consists ofrelatively hot, thin, and metal poor gas originated in theouter region and is flowing into the cool core.We also find a possible velocity shear between thetwo regions. This suggests that the two componentsare moving toward inverse-parallel to each other. Thisredshift offset appears to be higher than those measuredin several galaxy clusters. For example, in the case ofAbell 1835, the upper limit of redshift offset correspondsto ∆ v <
600 km s − (Ueda et al. 2017). Liu et al. (2018)measured the redshift offset of the cold fronts relative tothe cluster average in Abell 2142 to be 810 ±
330 km s − .For the Perseus cluster, Hitomi Collaboration et al.(2018) reported the velocity gradient with a 100 km s − amplitude across the cluster core, which is likely asso-ciated with a sloshing motion in the plane of the sky.The relatively high level of the velocity offset in A907is therefore peculiar. This indicates that the direc-tion of the ICM motion in A907 is different from othergalaxy clusters that are experiencing gas sloshing in theplane of the sky. In addition, the morphology of theX-ray residual image of A907 is similar to that expectedby LOS gas sloshing in numerical simulations. Ac-cording to numerical simulations, perturbations in theX-ray surface brightness generated by LOS gas slosh- ing are not spiral-like but ripple-like (e.g., Figure 5 ofRoediger et al. 2011). This paper therefore presents forthe first time the LOS gas sloshing in the cluster core.The left panel of Figure 3 shows the i -band optical im-age of the A907 field observed with the Pan-STARRS1survey (Chambers et al. 2016). The right panel of Fig-ure 3 presents the contours of the X-ray surface bright-ness of A907 and the positive excess region overlaid onthe optical image. The X-ray centroid coincides with theposition of the BCG. The BCG is located at the interfacebetween the positive and the negative excess regions.The positive excess region is extended toward the direc-tion of east from the BCG. This feature is consistentwith that expected by LOS gas sloshing in numericalsimulations (e.g., Roediger et al. 2011). This similaritysupports the scenario of LOS gas sloshing in A907. Thequality of this optical image does not allow us to identifythe second BCG, which is most likely associated with aninfalling subcluster, i.e., the cause of LOS gas sloshing.Deep, high-resolution optical data are needed to searchfor the second BCG and to study the substructure in thecentral region of A907 using strong gravitational lensingand the member galaxy distribution.The thermodynamic property of the ICM in the weakpositive excess region is far from that of the positiveand the negative excess regions, i.e., the highest ICMtemperature and entropy, and also the lowest electronpressure among them. The abundance of this region isconsistent with that of the negative excess region. Theseresults imply that the weak positive excess is not gen-erated by the same reason as the central positive andnegative excess regions. Since the redshift of this re-gion is consistent with that of A907, no large bulk mo-tion is expected. According to numerical simulations(e.g., Roediger et al. 2011; ZuHone et al. 2016), a mi-nor merger also generates a low-level positive fluctuationoutside the core, which has a higher temperature andslower motion than those in the central region. Thesepictures are consistent with those observed in the weakpositive excess region. The weak positive excess regionmay therefore be also caused by a minor merger.4.2. Velocity shear in the core
The system of LOS gas sloshing in A907 allows usto measure the velocity of the ICM motion induced bygas sloshing. This information is important for under-standing gas dynamics as well as a heating source(s)to prevent runaway cooling in the cool core. Some in-stabilities such as KHI are induced by a velocity shearat the interface of a cold front (e.g., Roediger et al. https://panstarrs.stsci.edu/ Ueda et al.
100 kpc 100 kpc
Figure 3.
Left: The optical image of A907 is taken by the Pan-STARRS1 (PS1) i -band. Right: The contours of the X-raysurface brightness (white), the positive excess (magenta), and the negative excess regions (green) are overlaid on the left panel. Hitomi for the Perseus clus-ter (Hitomi Collaboration et al. 2018), which is experi-encing gas sloshing in the plane of the sky and showsa spiral pattern in the residual X-ray surface bright-ness (e.g., Churazov et al. 2003; Simionescu et al. 2012;Walker et al. 2017). The main component of the veloc-ity field of sloshing motion is still unclear, so that thevelocity field inside the core might be underestimated.The redshift difference we find indicates a velocity off-set of ∆ v = 1680 +1310 − km s − . The velocity of each com-ponent is then inferred to be ∆ v/
2, i.e., ∼
800 km s − .This inferred velocity is in good agreement with thatexpected by gas sloshing in numerical simulations (e.g.,Ascasibar & Markevitch 2006). On the other hand, theadiabatic sound speed in the core of A907 is expectedto be ∼ − at 4 keV, which is larger than theinferred velocity. In addition, the ICM in both regionsis nearly in pressure equilibrium. These are consistentwith the picture that the nature of gas sloshing is sub-sonic and in pressure equilibrium (Ueda et al. 2018).In addition to the thermal pressure, this shear velocitycan provide additional, non-thermal pressure support forthe ICM. The ratio of the non-thermal to thermal energydensity is estimated as ǫ non /ǫ therm = ( γ/ M , where ǫ non is a non-thermal energy density, ǫ therm is a thermalenergy density, γ is the adiabatic index of γ = 5 /
3, and M is the Mach number, respectively. The ratio is then ǫ non /ǫ therm = 0 . +0 . − . . Note that this value is derivedfrom the shear velocity only between the positive and the negative excess regions. The ratio in the entire coreof A907 might be smaller than the observed ratio.In thecase of the Perseus cluster, Hitomi Collaboration et al.(2018) directly measured the non-thermal to thermal ra-tio in the core to be ∼ . − .
07 using the velocity of thebulk motion along the LOS and assuming isotropic tur-bulence recovered by the LOS velocity dispersion. Theirestimate did not include the contribution of bulk motionof sloshing in the plane of the sky. Our result indicatesthat the contribution of bulk motion along the direc-tion of gas sloshing is essential for the contribution ofnon-thermal pressure.Molnar et al. (2010) showed using high-resolution cos-mological simulations that the fraction of non-thermalpressure is 20 −
45 % in the core of high-mass relaxedclusters, whose virial mass range is in (1 − × M ⊙ .This is in good agreement with the virial mass of A907( ∼ × M ⊙ , Okabe & Smith 2016). To compare withtheir results, we recalculate the non-thermal fractionusing their definition, i.e., p nth / ( p nth + p therm ), where p nth is non-thermal pressure and p therm is thermal pres-sure. The fraction is then ∼
25 %, which is consis-tent with that suggested by Molnar et al. (2010). Onthe other hand, another cosmological simulations sug-gested that the non-thermal pressure support in thecluster core is about 10 % in relaxed clusters (e.g.,Lau et al. 2009; Nelson et al. 2014). Note that theirmass range is different from that of Molnar et al. (2010).This level is consistent with that in the Perseus cluster(Hitomi Collaboration et al. 2018). To address this ten-sion, it is crucial to reveal a 3D gas motions in clustercores for a large sample of systems, comprehending thenature of galaxy clusters. Since gas sloshing is associ- ine-of-sight gas sloshing in the cool core of Abell 907
Gas dynamics driven by gas sloshing
It is expected that the kinetic energy inferred from thesloshing motion of the ICM is finally converted into heat.If the contribution of such kinetic energy is significantenough to prevent runaway cooling in the cool core fora certain period, the heating by gas sloshing should becomplementary to the AGN feedback.First, we estimate the total mass in the positive andthe negative excess regions, assuming the LOS depthof 140 kpc, the mean molecular weight of 0.6, and theobtained electron number density. The total mass ofthe ICM in both regions is then ∼ . × M ⊙ and ∼ . × M ⊙ , respectively. In this case, the to-tal kinetic energy of each region is estimated to be ∼ . × × ( v/
800 km s − ) erg and ∼ . × × ( v/
800 km s − ) erg, respectively, where v is the LOS ve-locity of the ICM motion. Although the ICM is likelymoving not only along the LOS but also in the plane ofthe sky, it is hard to measure the velocity componenton the plane of the sky. We therefore focus on the LOSvelocity in this estimate. When we assume that suchkinetic energy is released during 1 Gyr constantly, theestimated power in total taking the statistical error intoaccount is 7 . +19 . − . × × ( t/ − erg s − , where t is a time-scale of release of kinetic energy.Next, we assume the bolometric luminosity of theICM as the absorption-corrected X-ray luminosity inthe 0 . −
100 keV band. They are then (1 . ± . × erg s − and (8 . ± . × erg s − , respectively.The total bolometric luminosity in the sloshing core isthus (2 . ± . × erg s − . The ratio of the esti-mated power to the bolometric luminosity is 0 . +0 . − . ,which indicates, albeit with large statistical errors, thatgas sloshing may play a significant role in heating theICM. Note that this is inferred from the LOS velocityalone, so that this value is a lower limit of the power.The remaining required power is however large, whichmeans that an additional heating source(s), such as theAGN feedback is needed, or the time-scale of energyrelease should be less than ∼ . CONCLUSIONSUsing the archival data of the
Chandra X-ray Obser-vatory , we have presented the first detection of LOS gassloshing in the cool core, which appears as a character-istic perturbation in the X-ray surface brightness of thecore of A907. The conclusions of this paper are summa-rized as follows.1. In the residual image of the X-ray surface bright-ness after subtracting the mean profile of A907, wefind the positive and the negative excess regions.The positive excess region is surrounded by thenegative excess region. The height and width ofthe positive excess region are ∼
75 kpc (26 ′′ ) and ∼
30 kpc (10 ′′ ), respectively.2. We analyze the X-ray spectra extracted from thepositive and the negative excess regions and findthat the ICM thermodynamic properties of bothregions are in good agreement with those expectedby gas sloshing. We also find a slight redshift dif-ference between both regions, which correspondsto the velocity shear of ∆ v = 1680 +1310 − km s − .This level of velocity offset is the highest amongthe previously measured systems of gas sloshing.When the ICM in both regions is moving towardinverse-parallel to each other, the shear velocity isexpected to be ∼
800 km s − .3. All of the observational results indicate that A907is experiencing gas sloshing along the LOS. LOSgas sloshing in the cluster core is found for the firsttime. A907 is important for understanding gas dy-namics driven by gas sloshing and is complemen-tary to galaxy clusters that have gas sloshing inthe plane of the sky.4. The observed shear velocity suggests that the ra-tio of the non-thermal to thermal energy density isabout 0.3. A907 is identified as one of typical mas-sive, relaxed, cool core clusters, while our resultsindicate that the non-thermal fraction of this levelis not negligible at least in the cluster core. Thisindicates that revealing 3D bulk motion inducedby gas sloshing is crucial to measure non-thermalpressure.5. The total kinetic energy inferred from the shear ve-locity and the ICM mass allows us to estimate thetotal power released in gas sloshing. The total esti-mated power is ∼ . × × ( t/ − erg s − Ueda et al. if the energy release rate is constant during gassloshing. This value corresponds to ∼
30 % of thebolometric ICM luminosity, which indicates thatthe contribution of gas sloshing to heat the ICMis non-negligible. To keep the complete balancebetween cooling and heating by only the total ki-netic energy of gas sloshing, the time-scale of en-ergy release should be less than ∼ . Facilities:
CXOREFERENCES
Arnaud, K. A. 1996, in Astronomical Society of the PacificConference Series, Vol. 101, Astronomical Data AnalysisSoftware and Systems V, ed. G. H. Jacoby & J. Barnes,17Ascasibar, Y., & Markevitch, M. 2006, ApJ, 650, 102Blanton, E. L., Randall, S. W., Clarke, T. E., et al. 2011,ApJ, 737, 99B¨ohringer, H., Schuecker, P., Pratt, G. W., et al. 2007,A&A, 469, 363Canning, R. E. A., Sun, M., Sanders, J. S., et al. 2013,MNRAS, 435, 1108Cavagnolo, K. W., McNamara, B. R., Wise, M. W., et al.2011, ApJ, 732, 71Chambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016,ArXiv e-printsChurazov, E., Forman, W., Jones, C., & B¨ohringer, H.2003, ApJ, 590, 225Clarke, T. E., Blanton, E. L., & Sarazin, C. L. 2004, ApJ,616, 178Fabian, A. C. 2012, ARA&A, 50, 455Fruscione, A., McDowell, J. C., Allen, G. E., et al. 2006, inProc. SPIE, Vol. 6270, Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series,62701VFujita, Y., Matsumoto, T., & Wada, K. 2004, ApJL, 612, L9Garmire, G. P., Bautz, M. W., Ford, P. G., Nousek, J. A.,& Ricker, Jr., G. R. 2003, in Society of Photo-OpticalInstrumentation Engineers (SPIE) Conference Series,Vol. 4851, Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series, ed. J. E. Truemper& H. D. Tananbaum, 28–44Ghizzardi, S., De Grandi, S., & Molendi, S. 2014, A&A,570, A117 Hitomi Collaboration, Aharonian, F., Akamatsu, H., et al.2018, PASJ, 70, 9Hlavacek-Larrondo, J., McDonald, M., Benson, B. A., et al.2015, ApJ, 805, 35Ichinohe, Y., Simionescu, A., Werner, N., Fabian, A. C., &Takahashi, T. 2019, MNRAS, 483, 1744Ichinohe, Y., Simionescu, A., Werner, N., & Takahashi, T.2017, MNRAS, 467, 3662Ichinohe, Y., Werner, N., Simionescu, A., et al. 2015,MNRAS, 448, 2971Kalberla, P. M. W., Burton, W. B., Hartmann, D., et al.2005, A&A, 440, 775Komatsu, E., Smith, K. M., Dunkley, J., et al. 2011, ApJS,192, 18Lau, E. T., Kravtsov, A. V., & Nagai, D. 2009, ApJ, 705,1129Leccardi, A., Rossetti, M., & Molendi, S. 2010, A&A, 510,A82Liu, A., Yu, H., Diaferio, A., et al. 2018, ApJ, 863, 102Liu, A., Yu, H., Tozzi, P., & Zhu, Z.-H. 2015, ApJ, 809, 27Lodders, K., & Palme, H. 2009, Meteoritics and PlanetaryScience Supplement, 72, 5154Markevitch, M., & Vikhlinin, A. 2007, PhR, 443, 1McNamara, B. R., & Nulsen, P. E. J. 2007, ARA&A, 45,117McNamara, B. R., Nulsen, P. E. J., Wise, M. W., et al.2005, Nature, 433, 45Molnar, S. M., Chiu, I.-N., Umetsu, K., et al. 2010, ApJL,724, L1Nelson, K., Lau, E. T., & Nagai, D. 2014, ApJ, 792, 25Okabe, N., & Smith, G. P. 2016, MNRAS, 461, 3794O’Sullivan, E., Giacintucci, S., Babul, A., et al. 2012,MNRAS, 424, 2971 ine-of-sight gas sloshing in the cool core of Abell 9079