Research & Scientific Contributions

Uhuru, Ariel V and optical ground-based observations led to the determination of a 3.4-day binary period for the X-ray binary system 4U1700-17 and to the identification of its optical counterpart. Other results include the determination of a 9-day period for 4U0900-40, times of low X-ray intensity in the light curve of Circinus X-1, that were interpreted as eclipses, periods of low optical luminosities in Hercules X-1, lasting as long as 5.5 years, and a 600-day recurrence period for the X-ray transient 4U1630-47.

Uhuru, which means "freedom" in Swahili, was the first satellite devoted to the observation of cosmic X-ray sources. Uhuru was launched from a platform off the coast of Kenya on December 12, 1970, the seventh anniversary of Kenya's independence. (Link)

RESEARCH & SELECTED PUBLICATIONS

Using Uhuru observations, Jones led the discovery of one of the first known X-ray binary systems 4U1700-37, which has a 3.5-day orbital period (Jones, Forman, Tananbaum, Schreier, Gursky, Kellogg, and Giacconi, 1973, ApJ Letters 181, 43).

The top panel in the figure above shows the Uhuru 2-6 keV intensity of 4U1700-37 obtained from May 10 to May 17, 1972. The lower panel shows Uhuru observations of 4U1700-37 obtained between December 1970 and May 1972, folded with the 3.41-day period. In addition to intensity changes during the binary period, 4U1700-37 shows non-periodic intensity variations on time scales of a tenth of a second.

With Tananbaum, Jones analyzed Uhuru observations that showed a transition from an X-ray bright state to an X-ray faint state in Cygnus X-1. The correlation of this X-ray transition with a change in the radio flux of Cygnus X-1, shown in Figure 1) led to its optical identification with a massive O star, and to the high likelihood that Cygnus X-1 was a black hole. (Tananbaum, Gursky, Kellogg, and Jones, 1972, Ap J Letters L5).

Jones also led the analysis to search for X-ray eclipses in the source Circinus X-1 using Uhuru observations (Jones, Giacconi, Forman, and Tananbaum, 1974, ApJ Letters 191, L71), Although apparent X-ray eclipses were found, see Figure 2, it was only possible to determine that the orbital binary period was longer than 15 days. In 1976, Kaluzienski, Holt, Boldt, and Serlemitsos (1976, Ap J Letters 208, L71) determined a period of 16.6 days for Circinus X-1 using Arial 5 All Sky Monitor observations.

From Uhuru and Ariel 5 observations, Jones, Forman, Tananbaum, and Turner (1976 ApJ Letters 210, L9) identified 3U1630-47 as the first recurrent transient X-ray source. As shown in Figure 1, four transient outbursts were observed with a period of about 600 days. In addition, Forman, Jones, and Tananbaum (1976, ApJ Letters 207, L25) identified the first transient X-ray source in a globular cluster (NGC6440).

With her second thesis advisor, William Liller, Jones performed ground based optical observations at CTIO and analyzed the glass plates in the Harvard collection. Since Sco X-1, which is by far the brightest X-ray source, had been identified with a 12th magnitude star (Sandage, Osmer, Giacconi, et al., 1966, Ap. J. 146, 316), it was generally believed that identifying the optical counterparts to galactic X-ray sources that were more than 1000 times fainter in their X-ray emission than Sco X-1 would be extremely difficult. However, Jones and Liller identified one of the first X-ray binary sources, 4U1700-37, with the bright (6.7 magnitude) O star HD153919 (1973), based on 24 consecutive nights of UBV observations in February and March 1973 with a standard photoelectric photometer on a 16-inch telescope at CTIO. The light cure for the optical counterpart is shown below.

Additionally, Jones, Chetin, and Liller (1974, ApJ Letters 190, L1) proposed possible optical counterparts for 92 other Uhuru sources, including three they identified with globular clusters (NGC6624 = 4U1820-30), M15 = 4U2131+11), and NGC6441 = 4U1746-37). None of these possible stellar counterparts showed the large level of optical variability seen in HZ Hercules, the optical counterpart of Hercules X-1. Jones et al. concluded from their observations that the six stars they suggested as the likely optical counterparts to luminous galactic X-ray sources are probably early type stars, as had been found for other galactic sources (e.g. Cyg X-1, 4U1700-37 and 4U0900-40).

The analysis by Jones, Forman, and Liller (1973, ApJ Letters 182, L109) of the Harvard Plates for HZ Herculis, obtained between 1890 and 1972, showed that the optical counterpart of the pulsating binary X-ray source Hercules X-1 does not always exhibit a two-magnitude change (from magnitude 13 to 15) in its luminosity during the 1.7-day binary period. Instead for time intervals in the 1920's and 1930s and again from 1949 to 1956, HZ Hercules remained at magnitude 15 throughout the 1.7-day binary period. These intervals of faint optical emission can be as short as 29 days. Since 1890, HZ Hercules has spent approximately half the time in a low brightness mode. Jones, Forman and Liller suggested that the intervals when the star is optically bright result from times when the star fills its Roche lobe, allowing matter to accrete onto the compact companion, thus producing the X-ray radiation that illuminates the side of the optical star that is facing the compact companion. At other times, when the optical star is faint, the star has contracted to a size smaller than its Roche lobe, so no mass is accreted onto the compact object and the binary enters its inactive X-ray state.

Einstein imaging observations led to 1) the discovery of hot gas halos in early type galaxies and groups that are gravitationally bound by dark matter halos, 2) the identification of multiple mass components in clusters of galaxies ("double" clusters), and 3) showed that 40% of clusters at the present epoch have significant substructures from ongoing mergers and are not relaxed systems. ROSAT observations were used to make the first detection of cavities in the hot gas around NGC1275.
 

Einstein Observatory (LINK)

ROSAT Guest Observer Facility (LINK)

EINSTEIN OBSERVATORY | RESEARCH & SELECTED PUBLICATIONS

Clusters of galaxies have been a focus of X-ray astronomical research since the discovery that they were luminous X-ray sources (Kellogg, Gursky, Leong, Schreier, Tananbaum, Giacconi, 1971, ApJ Letters 165, L49 | Gursky, Kellogg, Leong, Tananbaum, Giacconi, 1971, ApJ Letters L43 | Forman, Kellogg, Gursky, Tananbaum, Giacconi, 1972, ApJ 178, 309). The hot intracluster medium (ICM) is gravitationally the dominant luminous baryonic component in clusters. Thus, the X-ray emission from the ICM provides a sensitive tool for mapping both the mass distribution of the hot gas, as well as the distribution of the total mass, which is primarily dark matter in groups and clusters.

The Einstein Observatory, followed by ROSAT, offered the first opportunities to image the X-ray sky using focusing optics (see Van Speybroeck https://asc.harvard.edu/newsletters/news_10/leon.html), which significantly increased our knowledge of both galactic and extragalactic X-ray sources. The Einstein Observatory provided the first detailed X-ray images of the structure and morphology of clusters of galaxies. The richness and complexity of these images provided an impetus for theoretical investigations, especially numerical simulations of how clusters form and grow, as well as detailed optical and radio studies of clusters.

The paper “X-ray Observations of Galaxies in the Virgo Cluster” by Forman, Schwarz, Jones, Liller, and Fabian (1979, ApJ Letters 234, L27) described the first imaging X-ray observations of the core of the Virgo cluster obtained with the Einstein Observatory. Forman and colleagues detected X-rays from nine Virgo galaxies. The Einstein images revealed complex structures, including extended X-ray emission centered on the elliptical galaxy M86 (see figure 2). The systemic velocity for M86 is -419 km sec-1, indicating that the galaxy must be moving through the Virgo cluster with a velocity of approximately 1500 km sec-1 (Mach 2 for a gas temperature of 3 keV). Due to its infall into Virgo, M86 experiences a head wind that exerts a ram pressure on the galaxy’s hot atmosphere, causing gas to be stripped from the galaxy, resulting in the formation of a tail. Much of the hot X-ray emitting gas that was in the halo of M86 when the galaxy was outside the cluster core, is being ram pressure stripped as the galaxy crosses the cluster core. This stripped gas is eventually incorporated into the hot intracluster medium.

Early X-ray imaging observations from Einstein and ROSAT allowed astronomers to map the density of the hot gas in clusters of galaxies and showed that many clusters were not gravitationally relaxed systems (Forman & Jones, 1982, ARA&A 20, 547). The Einstein X-ray images of clusters also allowed accurate determinations of the gas mass in clusters, which showed that the hot intracluster medium is the dominant luminous baryonic component in clusters.

The Einstein survey of 55 clusters allowed Jones & Forman (1984, ApJ 276, 38-55) to characterize clusters into four classes, shown in Table 2 and Figure 3, based on the cluster structure (e.g. presence of substructure or cool cores, as well as the size of the cluster core radius), and the X-ray luminosity and temperature of the hot intracluster medium.

Based on the analysis of 208 clusters at z<0.2 observed with Einstein and with sufficient source counts to detect subclusters containing 15% of the cluster's luminosity, Jones and Forman (1999, ApJ 511, 1) classified each cluster's morphology, measured the cluster's X-ray luminosity, the surface brightness profile of the X-ray emission, and the central gas density. From this analysis, they found that 40% of clusters exhibit significant substructure, with no significant change in this percentage as a function of X-ray luminosity, and thus no significant change with cluster mass. This result implies that at least 40% of all present epoch clusters are still undergoing significant subcluster mergers and thus are not "relaxed" systems.

From their analysis of this large cluster sample, Jones and Forman (1999) also found that within a fixed metric radius of 1 Mpc, the gas mass fraction in clusters increases as a function of X-ray luminosity, from a gas mass fraction of 10% of the total cluster mass in low X-ray luminosity clusters to 20% in high luminosity massive clusters (see Figure 11).

To further characterize cluster structure, David White, a visiting graduate student from Cambridge University, worked with Jones and Forman to perform a deprojection analysis of the X-ray surface brightness for 207 clusters of galaxies observed with the Einstein Observatory (White, Jones, & Forman, 1997, MNRAS 212, 419). From this large sample of clusters, they confirmed that the ICM temperature as well as the cluster velocity dispersion are strongly correlated with the cluster X-ray luminosity and that approximately 60% of the clusters in the Einstein sample had high central gas densities and thus were expected to have cool gas cores. Figure 2 shows that clusters with cool cores generally have small core radii, while non-cool core clusters generally have large core radii.

In addition to the Einstein imaging surveys, noted above, a wide variety of approaches have been used to characterize and identify substructure in clusters of galaxies (Dressler, 1980, ApJ 236, 351 | Geller & Beers, 1982, PASP 94, 421 | Mohr, Fabricant, Geller, 1993, ApJ 413, 492 | Escalera, Biviano, Girardi, Giuricin, Mardirossian, Mazure, Mezzetti, 1994, ApJ 423, 539). In summary, both optical and X-ray studies have revealed a high fraction of substructure in rich clusters, with typical fractions of clusters with substructure exceeding 40%.

Based on the analysis of 104 clusters observed with the Einstein Monitor Proportional Counter, Ginga or EXOSAT, David, Slyz, Jones, Forman, Vrtilek and Arnaud (1993, ApJ 412, 479) reported a strong correlation between the cluster X-ray luminosity and the gas temperature of the hot intracluster gas. This correlation is shown in Figure 5a.

In addition to the study of clusters of galaxies, Einstein imaging observations of early type galaxies significantly increased our understanding of these systems. A primary example is that hot gaseous coronae are a common and perhaps ubiquitous features of these galaxies, as shown by the analysis of Einstein observations of 55 luminous early-type galaxies, Forman, Jones, and Tucker (1985, ApJ 293, 102). The presence of these hot halos showed that matter, which had been previously thought to be expelled in a galactic wind, is stored in a dark matter halo. The hot coronae provide a unique tracer of the mass in the galaxies dark matter halos. Total masses for these galaxies are up to 5 X 1012 solar masses, yielding mass to light ratios of up to 100, in solar units. This research formed the basis of the first Rossi Prize, presented by the AAS High Energy Astrophysics Division to Forman and Jones.

In the paper "Substructure: Clues to the Formation of Clusters," (1995, ApJ Letters, L5), West, Jones and Forman showed, based on the analysis of the spatial distribution of seven clusters with substructures taken from the sample of clusters observed with Einstein (Jones & Forman, 1984, ApJ 276, 38 | Jones & Forman, 1999, ApJ 511, 65), that the alignment on the sky of the subclusters matches the far larger scale filaments traced by the galaxies. This alignment was interpreted as the result of subclusters falling into clusters along large scale filaments, which then showed that massive clusters form through the merger of smaller clusters along large scale filaments.

ROSAT OBSERVATORY | RESEARCH & SELECTED PUBLICATIONS

In the paper, “Mapping the Dark Matter in the NGC5044 Group with ROSAT,” David, Jones, Forman, and Daines (1994, ApJ 428, 544) used the ROSAT observations of the X-ray luminous, 1 keV gas in NGC5044 to accurately map the total gravitating mass, as well as temperature and abundance profiles of the hot gas to a radius of 250 kpc from the central galaxy. Figure 1 shows the intensity contours from the ROSAT PSPC superposed on the optical image. Figure 5 shows the integrated radial distributions of the hot, X-ray emitting gas, the stellar mass, the dark matter and the total mass in NGC5044. Within a radius of 250 kpc, David et al. measured a total mass of 1.6 x 1013 Mʘ sun and a mass-to-light ratio of 130 (in solar units). 

In the paper "Cosmological Implications of ROSAT Observations of Groups and Clusters," David, Jones and Forman (1995, ApJ 445, 578) found that all seven of the groups and clusters in their sample for which mass-to-light ratios could be measured, have mass-to-light ratios of 100-150 (see Figure 1). This strongly challenged the traditional view that the mass-to-light ratios of rich clusters are significantly greater than that of groups and individual massive early type galaxies. David, Jones and Forman also found that the ratio of total mass to luminous mass (hot gas plus stars) decreases between early type galaxies and clusters, as shown in Figure 6. This decrease is due to the increase in the fraction of the mass that is in the form of hot gas in clusters, compared to the fraction of hot gas in early type galaxies and groups.

NGC1399 is the central massive galaxy in the core of the Fornax cluster. From their analysis of ROSAT PSPC observations, Rangarajan, Fabian, Forman, and Jones (1995, MNRAS, 272, 665), found relatively cool gas (0.85 keV) in the core of NGC1399 and warmer gas (1.1 keV) at radii of 10 to 330 kpc, with the total mass of the hot gas exceeding 1010 solar masses. The predicted mass deposition of the cooling gas is shown in Figure 14. The apparent contradiction between the amount of cool gas detected in X-ray observations in the core of the Fornax cluster and the predicted total amount of cool gas in the core was only resolved through high spatial resolution Chandra observations.

From deep ROSAT observations, Jones, Stern, Forman, Breen, David, Tucker, and Franx (1997, ApJ 482, 143) mapped the X-ray emission in the Fornax cluster of galaxies. The isointensity contour map of the smoothed X-ray emission superposed on the optical emission in Figure 1 shows the extended emission to be brightest on the central galaxy NGC1399, but also asymmetric with more extended emission to the northeast of NGC1399, than to the southwest. Jones et al. used measurements of the hot gas temperature and density to determine the total mass. A comparison to M87/Virgo showed that, while the extended profiles of the optical light are similar, the differences in total mass and gas mass are significant, with M87 having about four times more mass in hot gas and in the gravitating mass than does the Fornax cluster. Approximately 25% of the luminous baryons in M87 are in hot gas, compared to Fornax. Jones et al. also detected X-ray emission from seven other early type galaxies. For NGC1404, shown in Figure 10 (below), Jones et al. found extended X-ray emission with a "tail" toward the southeast, likely the result of ram pressure stripping as NGC1404 falls through the gas in the NGC1399 group. (see papers by Su et al (2017, ApJ 834, 74 | 2017, ApJ 835, 19), based on Chandra observations, for a more extensive analysis of NGC1404). 

In the paper "A Catalog of 200 Galaxy Clusters Serendipitously Detected in the ROSAT PSPC Pointed Observations,” Vikhlinin, McNamara, Forman, Jones, Quintana, and Hornstrup (1998, ApJ 502, 558) reported on their analysis of a large sample of 203 clusters, serendipitously found in 647 ROSAT PSPC observations covering 158 square degrees. X-ray luminosities for these clusters were found to range from those typical of poor gropus to those typical of rich clusters. Cluster redshifts ranged from z = 0.015 to z > 0.5. Figure 3 shows the photon images (upper left), which was used to detect structures on different angular scales. The resulting wavelet image is shown in the upper right. Connected regions are then identified in the wavelet image, as shown in the lower left. The best fit image is shown in the lower right, with four extended sources marked with arrows. All four of these extended sources were later confirmed by optical observations. 

In the paper "Evolution of Cluster Luminosities and Radii: Results from the 160 Square Degree ROSAT Survey," Vikhlinin, McNamara, Forman, Jones, Quintana and Hornstrup (1998, ApJ Letters 498, L21) found from comparing the number of X-ray luminous (> 1044 ergs s-1 clusters at the present epoch, with a high redshift (z>0.3) sample of clusters found in the 160 Square Degree Survey, that the high redshift cluster sample showed a factor of 3 to 4 deficit in the number of clusters, compared to the low redshift sample. This result is consistent with the cluster evolution found in the Einstein Observatory EMSS cluster sample by Gioia, Henry, Maccacaro, Morris, Stocke, Wolter (1990, ApJ Letters, L35) and Henry, Gioia, Maccacaro, Morris, Stocke, Wolter (1992, ApJ 386, 408).

Observations made with the ROSAT HRI took full advantage of the ROSAT mirror assembly. In particular, the ROSAT HRI observations of the Perseus cluster of galaxies showed an X-ray surface brightness peak at the cluster center, which was generally interpreted as due to the pressure induced inflow of gas releasing its thermal energy through radiation. In the paper “Asymmetric Arcminute Scale Structures Around NGC1275” (2000, A&A 356, 788), Churazov, Forman, Jones, and Bohringer argued that complicated structure seen in the X-ray emission in the core of the Perseus cluster were related to past activity of NGC1275. In particular, as shown in Figure 1, two large bubbles, to the north and south of the cluster center are seen in the ROSAT X-ray image and coincide with bright lobes of radio emission. These lobes of relativistic plasma, inflated by jets, would rise buoyantly and mix with the interstellar medium. Churazov et al. also showed that the energy released through the bubbles produced by the SMBH in NGC1275 is on the order of 1045 ergs sec-1, which Is comparable to the energy radiated in X-rays by the central region of the Perseus cluster.

The analysis of Chandra and XMM-Newton observations showed: 1) there is a significantly larger number of massive clusters in the present epoch (z<0.25) than at earlier times (z=0.35-0.90), and 2) from the sensitivity of the cluster mass function to cosmological models, the existence of dark energy is required.

Chandra’s arcsecond angular resolution and broad energy response (0.3 – 10 keV) allowed the density and temperatures of the hot intracluster medium to be spatially determined. In the paper “Chandra Sample of Nearby Relaxed Galaxy Clusters: Mass, Gas Fraction, and Mass-Temperature Relation” (2006, ApJ 640, 691), Vikhlinin, Kravtsov, Forman, Jones, Markevitch, Murray, and Van Speybroeck measured the temperature and density profiles of the hot intracluster medium to r500 or beyond in thirteen low-redshift, relaxed clusters, which allowed the cluster hydrostatic mass, as well as gas mass, to be determined to these radii (see Figure 17). In three clusters, a central excess in the gas mass associated with their cD galaxy was found. Overall, the temperature of the hot intracluster gas within both r500 and r2500 was found to tightly correlate with the total cluster mass as shown in Figure 19.

Total mass within r500 (red)and r2500 (blue), as a function of X-ray spectroscopic temperature, Tspec, and gas-mass-weighted temperature, Tmg.

Figure 19 (Vikhlinin et al., 2006, ApJ 640, 691) 

In the paper “Chandra Cluster Cosmology Project. III,” Cosmological Parameter Constraints (2009, ApJ 692, 1060) Vikhlinin, Burenin, Ebeling, Forman, Hornstrup, Jones, Murray, Nagai, Quintana, and Voevodkin measured the properties of the hot gas for 49 of the X-ray brightest clusters detected in the ROSAT All Sky Survey and for 37 clusters at redshifts <z> = 0.55 identified from the 400 deg2 ROSAT survey. As shown in Figure 1, significant differences in the mass functions of low redshift clusters (z=0.025-0.25) and those at high redshift (z=0.35-0.90) were found. The cluster mass function at z=0.25 is also sensitive to the chosen cosmological model, as shown in Figure 2. The low and high redshift cluster samples also place constraints on σ and Ω . Figure 10 shows the constraints on Dark Energy in a flat Universe from the combination of all cosmological data sets available in 2008. 

One of the important scientific results from Chandra observations of clusters was that the amount of cool gas in the centers of “cooling flow” clusters was less than predicted by classical radiative cooling models (e.g. Churazov, Forman, Jones, Bohringer, 2000, A & A 356, 788 | David, Nulsen, McNamara, Forman, Jones, Ponman, Robertson, Wise, 2001, ApJ 557, 546 | Fabian, Sanders, Allen, Crawford, Iwasawa, Johnstone, Schidt, Taylor, 2003, MNRAS 344, L43 | McNamara, Nulsen, Wise, Rafferty, Carilli, Sarazin, Blanton, 2005, Nature 433, 45 | McNamara, Wise, Nulsen, David, Sarazin, Bautz, Markevitch, Vikhlinin, Forman, Jones, Harris, 2000, ApJ Letters 534, L135 | Fabian 2012 ARA&A 50, 455).

A massive early-type galaxy lies at the center of nearly all cool-core clusters and groups. At the center of this galaxy is a supermassive black hole (SMBH). At very early epochs, these massive black holes grew rapidly through mass accretion and are observed as very luminous quasars. At the present epoch, the supermassive black hole that lies in the center of massive galaxies accretes matter as the hot gas in the galaxy center or in the core of the surrounding group or cluster of galaxies cools. Unlike quasars, these supermassive black holes are generally radiatively faint and are in a state of “maintenance feedback,” in which the SMBHs, although accreting at levels well below the Eddington mass accretion rate, can undergo AGN outbursts. These energetic outbursts produce cavities in the surrounding hot gas in galaxies, groups and clusters. The high spatial resolution of the Chandra observatory has allowed the detection and study of these cavities. By measuring the volumes of these cavities, observers have calculated the energy required to displace the hot gas and have determined that the kinetic energies of the local SMBHs are far larger than their radiative energies. Thus, the energy from the outbursts produced by SMBHs, in most cases, can reheat the cooling gas in the cores of galaxies, groups, and clusters

One of the earliest Chandra observations was of the Hydra A cluster of galaxies. As described in the papers "Chandra X-ray observations of the Hydra A Cluster: An Interaction between the Radio Source and the X-ray Emitting Gas" (McNamara, Wise, Nulsen, David, Sarazin, Bautz, Markevitch, Vikhlinin, Forman, Jones, 2000, ApJ Letters 534, L135), and “A High Resolution Study of the Hydra A Cluster with Chandra: Comparison of the Core Mass Distribution with Theoretical Predictions and Evidence for Feedback in the Cooling Flow” (David, Nulsen, McNamara, Forman, Jones, Ponman, Robertson, Wise, 2001, ApJ 557, 546), Chandra's arcsecond angular resolution showed a remarkable interaction between the radio lobes and the hot X-ray emitting gas. Figure 1 (below) shows the isointensity contours of the X-ray emission superposed on the optical image of the cluster. In the core of Hydra A, the Chandra image shows a bright core, extended X-ray remission and two large cavities in the hot gas, to the northeast and southwest of the core. These two X-ray cavities are filled with radio emission as shown below (images from the Chandra Photo Album). These cavities were created by the expansion of the radio lobes in the hot intracluster medium. The minimum amount of energy required to displace the gas in the cavities is about 1059 ergs. The non-detection of shocks surrounding the cavities implies the gas motions are subsonic. Thus, the estimated time for the lobes to expand to their present size, assuming they expand at the sound speed, is about 20 million years. A surprising result from these Chandra observations is the lack of spectroscopic evidence for multiphase gas within the central region, where the cooling time for the gas is less than 1 Gyr. Based on these results, David et al. proposed a scenario where only a small amount of the hot gas in the cluster core is able to cool and accrete onto the central supermassive black hole, which triggers the formation of a radio jet, which mechanically heats the cooling gas through strong shocks. 

https://chandra.harvard.edu/photo/2009/hydra/ 

While Einstein X-ray observations of NGC4636 (Forman, Jones & Tucker, 1985, ApJ 293, 102) showed the presence of a hot gas halo, Chandra's high angular allowed Jones and her colleagues to study the structure of the halo with a limiting resolution of 50 pc. In the paper "Chandra Observations of NGC4636: An Elliptical Galaxy in Turmoil" (2002, ApJ Letters 567, L115), Jones et al. reported new features in the hot X-ray emitting gas (see Figure 1 below [Jones et al., 2002, ApJ Letters 567, L115]) that can be explained as the results of shocks produced by outbursts from the supermassive black hole in the core of the galaxy. NGC4636 was one of the first examples known of a system where the accumulation of cooling gas from the X-ray halo in the galaxy core is prevented by reheating of the gas by outbursts from the central supermassive black hole. 

From deeper (200 ks) Chandra observations of NGC4636, Baldi, Forman, Jones, Kraft, Nulsen, Churazov, David and Giantucci (2009, ApJ 707, 1034) showed that the “bubble” morphology seen in the X-ray emission is likely produced by shocks which are driven by energy deposited off axis by jets from the central AGN. In addition, the X-ray cavity to the southwest was measured to have a Mach number of 1.7, an age of 2 million years and requires a total energy of 1056 ergs to produce the shock.

From the Chandra observations of M84 (NGC 4374), Finoguenov and Jones (2001, Ap J Letters, 547, L107) found that the structure of the X-ray emission is defined by the radio lobes. In particular, as shown in Figure 1 (below), two regions, north and south of the galactic nucleus were found to have low gas densities. These low-density regions are filled with radio emission and are surrounded by higher gas density X-ray arms, with gas temperatures like that in the galaxy core. Thus, Finoguenov and Jones concluded that the expanding radio lobes created cavities in the hot X-ray-emitting gaseous halo that surrounds M84.

Left: M84 (Chandra X-Ray only), Right: M84 (Composite: Composite: Chandra X-Ray, Optical, and Radio)

As reported in the paper "In-Depth Chandra Study of the AGN Feedback in the Virgo Elliptical Galaxy M84" (Finoguenov, Ruzkowski, Jones, Brugen, Vikhlinin, Mandel, 2008 ApJ 686, 911), M84 hosts two pairs of cavities in the X-ray emitting gas, based on deeper (26 ksec) Chandra observations. The PV work required to produce each of these four cavities ranges from 3 x 1054 ergs for the small cavities to 14 x 1054 ergs for the large cavities. 

As the nearest active galaxy, NGC5128, which hosts the radio jet Centaurus A, was observed with the high-resolution imager (HRI) on both Einstein and ROSAT. The Einstein observations showed the presence of a bright nucleus with an X-ray jet, individual bright sources and diffuse X-ray emission (Schreier, Feigelson, Delvaille, Giacconi, Grindlay, Schwartz, and Fabian, Feigelson, 1979, ApJ Letters 234, 39). Centaurus A has been observed extensively with Chandra. Analysis of these observations has resulted in more than twenty published refereed papers with Jones as an author. Below are highlights from four of these papers.

The first Chandra observations of Cen A were taken with the High-Resolution Camera (HRC) in September 1999 and published in the paper "A Chandra High Resolution X-ray Image of Centaurus A" (Kraft, Forman, Jones, Kenter, Murray, Aldcroft, Elvis, Evans, Fabbiano, Isobe, Jerius, Karovska, Kim, Prestwich, Primini, Schwarz, Schreier, Vikhlinin, 2000, ApJ Letters 531, L9.) These observations showed that the knots of X-ray emission in the jet were spatially extended, as well as was the X-ray emission from the southwest radio lobe. Small, but significant differences between the X-ray and radio morphologies were found. In addition, 63 X-ray point sources, including four coincident with globular clusters were detected. Results from the analysis of two 36 kilosecond Chandra ACIS-I observations are discussed in "Chandra Observation of the X-ray Jet in Centaurus A" (Kraft, Forman, Jones, Murray, Hardcastle, Worrall, 2002 ApJ 569, 54). This paper reported the detection of diffuse X-ray emission from the jet along with 31 distinct X-ray knots in the jet (see Figure 2). These observations suggest that the X-ray emission is due to synchrotron emission from relativistic particles, while the radio emission originates, at least in part, form a spatially distinct, less energetic population of electrons.

Results from deep Chandra and XMM-Newton observations of Centaurus A were presented in the paper “X-ray Emission from the Hot Interstellar Medium and Southwest Radio Lobe of the Nearby Radio Galaxy Centaurus A" (Kraft, Vazquez, Forman, Jones, Murray, Hardcastle, Worrall, Churazov, 2003, ApJ 592, 129). Analysis of the X-ray observations (Figure 1), showed that the supersonic expansion of this lobe can provide enough energy to reheat the hot interstellar medium.

In the paper "Shocks and Cavities from Multiple Outbursts in the Galaxy Group NGC5813: A Window to Active Galactic Nucleus Feedback," published in 2011 (ApJ 726, 86) Randall, Forman, Giacintucci, Nulsen, Sun, Jones, Churazov, David, Kraft, Donahue, Blanton, Simionescu, and Werner reported on their analysis of Chandra observations of the galaxy group NGC5813. As shown in Figure 1, this group shows three pairs of collinear cavities in the X-ray emission at 1 kpc, 8 kpc, and 20 kpc from the center of NGC5813. These were produced by three distinct outbursts of the central active galactic nucleus that occurred 3 million, 20 million and 90 million years ago. The two more recent outbursts had average powers of 1.5 x 1042 ergs sec-1 and 1043 ergs sec-1, a factor of six difference, indicating that the power of the SMBH changed significantly over a time scales of 107 years.

The initial Chandra observations of NGC5813 were followed by very long (650 ksec) Chandra observations. Results were reported in the paper “A Very Deep Chandra Observation of the Galaxy Group NGC5813: AGN Shocks, Feedback and Outburst History” by Randall, Nulsen, Jones, Forman, Babul, Clarke, Kraft, Blanton, David, Werner, Sun, Donahue, Giancintucci and Simionescu (2015, ApJ 805, 112). The Chandra image of the central region of NGC5813 is shown in Figure 1. A map of the gas temperature with dashed lines indicating the location of the 10 kpc shock fronts is shown in Figure 6. Most importantly, these observations show that the energy from shock heating is sufficient to offset the radiative cooling of the hot gas. In addition, the kinetic energy of the AGN outburst is approximately the same for each of the three outbursts, which implies that the average AGN kinetic energy remains stable over long (~50 Myr) timescales.

M87 is the central elliptical galaxy in the Virgo Cluster and provides a unique laboratory to study the interaction between the energy generated by a supermassive black hole and the surrounding hot intracluster gas. Its proximity allowed the disk surrounding SMBH to be resolved by HST, thus providing a mass determination for the SMBH of 2.4 X 109 solar masses (Harms, Ford, Tsvetanov, Hartig, Dressel, Kriss, Bohlin, Davidson, Margon, Hochhar, 1994, ApJ 435, L35), as well as an image of the SMBH by the Event Horizon Telescope Collaboration (2019, ApJ Letters 875).

Using X-ray observations from Chandra, XMM-Newton and the ROSAT HRI, Forman, Nulsen, Heinz, Owen, Eilek, Vikhlinin, Markevitch, Kraft, Churazov, and Jones found several remarkable features produced by the SMBH outbursts, which were discussed in the paper “Reflections of Active Galactic Nucleus Outbursts in the Gaseous Atmosphere of M87” (2005, ApJ 635, 894). Figure 1c shows the Chandra ACIS-S image of the region around M87. The Chandra images show evidence for buoyant bubbles and energetic outbursts powered by the supermassive black hole in the M87 nucleus. Prominent features include the bright nucleus and jet, cavities in the cluster gas, an X-ray bright core surrounding the 6 cm radio lobes, and four X-ray cavities in the eastern arm. Figure 4 shows the faint features in the hot gas halo of M87 that can be seen when the large-scale radial surface brightness gradient of the X-ray emitting gas in the halo of M87 is removed. These X-ray features include a nearly azimuthally symmetric ring of emission at a radius of 14 kpc, along with brightening of the X-ray “arms,” also at radii of 14 kpc, and a second partial ring of enhanced emission at 17 kpc. The features at 14 and 17 kpc are likely due to weak shocks, produced by outbursts from M87’s SMBH.

As shown in the temperature map of the hot gas around M87 determined from XMM-Newton observations (Figure 6), the gas in the core and along the arms is cooler than the gas in the halo. Contours from the 90 cm radio observation are shown in black.

Using a deep 500 ksec Chandra ACIS-I observation of M87, Forman, Jones, Churazov, Markevitch, Nulsen, Vikhlinin, Begelman, Bohringer, Eilek, Heinz, Kraft, Owen, and Pahre carried out an analysis of the structure of the X-ray emission from the filaments and bubbles in the hot gas in M87 (2007, ApJ 665, 1097). In that paper, “Filaments, Bubbles, and Weak Shocks in the Gaseous Atmosphere of M87,” Forman et al. reported the discovery of a weak shock at a distance of 13 kpc from the center of M87 that was produced by an outburst from the SMBH. The hot X-ray gas surrounding M87 shows reflections of previous outbursts of the supermassive black hole in the form of “bubbles” and their bright rims, shocks and buoyantly uplifted gas structures. Figure 2 shows the “soft” 0.5-1 keV Chandra ACIS-I image of M87. The long southwestern arm (labelled “filament”) appears to be composed of several intertwined filaments, while the eastern arm appears to be a series of bubbles, at different evolutionary stages, as they rise in the hot atmosphere of M87.

The left image of figure 6 shows the 0.5-2.5 keV Chandra ACIS image at full spatial resolution (0.492”x 0.492” pixels), after background subtraction and “flat fielding.” In addition to the bright jet, cavities and the rims of cavities associated with the jets are visible in this Chandra image. The cavity labelled “bud” coincides with the radio emission extending south from the cocoon in the (central) JVLA image. The Chandra “hard” energy band (3.5-7.5 keV) image shows a ring of emission with an outer radius of 12.8 kpc (2.75’).

In the 13 kpc ring, as shown in Figure 10, the gas temperature increases by about 20%. This temperature jump is consistent with the Mach number for the shock of 1.2 derived by Forman et al. (2005). The jump in the gas density yields a Mach number of 1.22 +/- 0.02. Thus, both the density and temperature profiles for the hot gas around M87 show the properties of a classical shock. The age of the outburst that produced this shock can be estimated by dividing the radius of the shock by the shock velocity. For M87, this is 14 Myr. Since the velocity of the shock is expected to be higher in the past, the age of the outburst is somewhat lower.

More recently in their paper “Partitioning the Outburst Energy of a Low Eddington Accretion Rate AGN at the Center of an Elliptical Galaxy: The Recent 12 Myr History of the Supermassive Black Hole in M87” (2017, ApJ 844, 122) Forman, Churazov, Jones, Heinz, Kraft, and Vikhlinin combined constraints from X-ray and radio observations of M87 with a shock model to derive the duration of the AGN outburst that created the shock located at a radius of 13 kpc from M87.

Figure 1 shows the Chandra (left) and the 90 cm VLA (right) on the same scale. The outline of the lowest intensity radio emission is superposed on the Chandra image. At the core, the Chandra image shows the M87 jet extending 20” to the NW. The jet is currently filling the central cavity with relativistic plasma, seen in the dark region of the VLA image. The VLA image also shows a pair of “arms” extending up to 5’ to the east and to the southwest. The radio torus at the end of the eastern arm is a buoyant bubble of plasma that has risen about 20 kpc during the past 40-70 million years (Owen, Eilek, Kassim, 2000, ApJ 543, 611 | Churazov, Bruggen, Kaiser, Bohringer, Forman, 2001, ApJ 554, 261). Only a twisted filamentary radio arm remains of the plasma bubble to the southwest. The Chandra image shows filamentary arms of relatively cool gas, uplifted by the buoyant plasma bubbles. The left image of Figure 2 shows the 3.5-7.5 keV Chandra image, which approximately represents the square of the gas pressure, projected on the sky for cluster gas temperatures in the range from 1 to 3 keV. Thus, this image shows direct evidence for outbursts as over-pressured regions. The image on the right side of Figure 2 shows the data divided by the average radial profile to enhance features in the bright core. Evidence is found for two outbursts, the 13 kpc shock and the central, over-pressured cocoon with an X-ray bright rim, which was initially inflated by the 13 kpc shock, but is now re-pressurized by the current outburst.

To constrain the duration of the outburst, Forman et al. compared the Chandra observations to what would be observed for a short duration (105 years) outburst and for a longer duration (2.2 106 years) outburst. A schematic of the powerful short duration outburst and longer duration, more “gentle” outburst is shown in Figure 10. Finally Figure 12 shows the outburst energy and the duration of the outburst for the various outburst models. Forman et al concluded that the M87 outburst began about 12 billion years ago, had a duration of 1-3 billion years, and had a total energy of 5-6 x 1057 ergs.

Chandra observations of the elliptical/S0 galaxies M86, M60 and NGC4472 in Virgo, NGC1404 in the Fornax cluster and the spiral galaxy ESO137-001 in the cluster A3627 show spectacular tails and other features in the hot gas that is stripped from the galaxies as they move through the intracluster medium.

Prior to the detailed studies of the distributions of galaxies in rich clusters (see Dressler 1980, ApJ 236, 351) and the Einstein and ROSAT X-ray observations of clusters (see Forman & Jones 1982 ARA&A 20, 547 | Jones & Forman, 1984, ApJ 276, 38 | Jones & Forman, 1999, ApJ 511, 65 | Vikhlinin, McNamara, Forman, Jones, Quintana, Hornstrup, 1998, ApJ Letters L21), most researchers thought clusters of galaxies were simple, dynamically relaxed systems. However, the Einstein and ROSAT observations, as well as the mapping of the structure of clusters based on the distribution of galaxies, showed that a large fraction of galaxy clusters are not relaxed systems at the present epoch, but instead are continuing to grow through the infall of galaxies and the merger of subclusters (see examples in published papers described in the Einstein and ROSAT sections). The infall of galaxies and groups into clusters can leave visible imprints on both the hot intracluster medium and the hot corona of the infalling galaxy. Chandra observations have allowed these processes to be studied in far greater detail than was possible in the past. Below are highlights from published papers that Jones contributed to scientifically that describe the infall of the galaxies M86, M49, M60 and M89 into the Virgo Cluster and the infall of NGC1404 into the Fornax Cluster. These highlights include what has been learned about the physical properties of the galaxies and the intracluster medium from these observations.

While the discovery of the ram pressure stripped plume of hot gas in the Virgo galaxy M86 was made using observations from the Einstein Observatory (Forman, Schwarz, Jones, Liller, Fabian, 1979, ApJ Letters 234, L27), a mosaic of nine Chandra observations (see Figure 2) allowed a far more detailed study of M86. As Randall, Nulsen, Forman, Jones, Machacek, Murray, Maughan reported in the paper "Chandra's View of the Ram Pressure Stripped Galaxy M86” (2008, ApJ 688, 208), the true length of the tail is at least 380 kpc, the gas mass is ~109 solar masses and the free-fall velocity of M86 is 1500 km s-1. Thus, M86 is at best only marginally bound to the Virgo cluster. Randall et al. also suggested that the removal of a significant fraction of M86’s hot ISM in a single X-ray emitting “blob” was the result of rapid ram pressure stripping. Figure 15 shows the path of the orbit determined for M86 and the stripped tail. Note that the path of the stripped tail follows a trail of denser gas.

Based on deep Chandra (200 ks) and XMM-Newton observations, the paper "Stripped Elliptical Galaxies as Probes of ICM Physics. III Deep Chandra Observations of NG4552: Measuring the Viscosity of the Intracluster Medium" (2017, ApJ 848, 27) by Kraft, Roediger, Machacek, Foman, Nulsen, Jones, Yu, and Sheardown showed that the viscosity in the hot ICM is negligible and that the plasma stripped from the infalling galaxy is efficiently mixed with the ICM.

Hot gas in the halo of NGC4472 was first observed with Einstein (Forman, Jones, Tucker, 1985, ApJ 293, 102) and further studied with ROSAT (Forman, Jones, David, Franx, Makishima, Ohashi, 1993, ApJ Letters 418, L55 | Irwin & Sarazin, 1996, ApJ 471, 683). Using Chandra and ROSAT observations, Biller, Jones, Forman, Kraft, and Ensslin in their paper “Hot Gas Structure in the Elliptical Galaxy NGC4472” (2004, ApJ 613, 238), showed that the overall morphology of the hot gas in the halo of NGC4472 is due its motion through the Virgo intracluster medium, as earlier suggested by Irwin and Sarazin (1996) based on ROSAT observations (see figure 2). Biller et al. also reported the detection of X-ray cavities associated with the radio lobes and likely produced during an AGN outburst that began approximately 12 million years ago and may still be ongoing (see Figure 3).

From deep (100 ks) XMM-Newton observations of NGC4472, Kraft, Forman, Jones, Nulsen, Hardcastle, Raychaudhury, Evans, Sivikoff, and Sarazin (2011, ApJ 727, 41) found that the NGC4472 group is likely infalling into a filamentary structure in the Virgo cluster that itself is falling toward M87. This paper also presents the effects of outbursts from the central supermassive black hole in NGC4472. In particular, cooler gas in the X-ray “arms” of NGC4472 is likely due to cold gas that was originally in the galaxy core and has been uplifted by outbursts.

During an academic year visit to the CfA, Ryan Wood, a Southampton University student, analyzed Chandra observations of the Virgo galaxy M60. As described in the published paper “The Infall of the Virgo Elliptical Galaxy M60 toward M87 and the Gaseous Structures Produced by Kelvin-Helmholtz Instabilities” (2017, ApJ 847, 79) by Wood, Jones, Machacek, Forman, Bogdan, Andrade-Santos, Kraft, Paggi, Roediger), the Chandra observations show that the gas structures are characteristic of gas-stripping from the hot halo of M60 as the galaxy falls toward M87 with an infall velocity of about 1000 km s-1. The presence of a “tail” and filamentary gaseous “wing” structures that likely arise from Kelvin-Helmholtz instabilities, are due to stripping of the hot halo gas.

In addition to the detailed studies of galaxies that are moving through the Virgo cluster noted above, there has been extensive use of Chandra observations to constrain the dynamical motion and derive physical properties of the massive early-type galaxy NGC1404 as it falls into the Fornax Cluster, the nearest galaxy cluster in the southern sky. As shown by Machacek, Dosaj, Forman, Jones, Markevitch, Vikhlinin, and Warmflash in Figure 1 from the paper “Infall of the Elliptical Galaxy NGC1404 into the Fornax Cluster” (2005, ApJ 621, 663-672), NGC1404 has a sharp leading edge in its hot gas halo, which is produced as the galaxy falls toward the center of the Fornax cluster. Analysis of the temperature and density of the hot gas showed that the sharp edge in the hot gaseous halo around NGC1404 marks a classical “cold front” that separates the cooler and denser gas in the infalling galaxy from the hotter intracluster medium and places constraints on the motion of NGC1404.

The early Chandra observations of NGC1404 noted above were followed by very deep (670 ksec) Chandra observations. The analysis of these observations was led by postdoctoral fellow Yuanyuan Su and presented in three papers. In the first of these papers “Deep Chandra Observations of NGC1404: Cluster Plasma Physics Revealed by an Infalling Early-type Galaxy“ which was published in 2017 (ApJ 834, 74), Su, Kraft, Nulsen, Roediger, Forman, Churazov, Jones, and Machacek reported sub-kiloparsec-scale eddies generated by Kelvin-Helmholtz instabilities, and placed an upper limit of 5% of the Spitzer value on the isotropic viscosity of the hot intracluster gas. Su et al. also found evidence of mixing between the colder gas in the galaxy tail and the hotter gas in the cluster, which indicates the gas has a low viscosity, and placed a limit of 4 microGauss on the magnetic field of the Fornax Cluster ICM. 

In their second paper on the infall of NGC1404 into the Fornax cluster “Capturing the 3D Motion of an Infalling Galaxy via Fluid Dynamics” published in 2017 (ApJ 835, 19), Su, Kraft, Nulsen, Roediger, Forman, Churazov, Randall, Jones, and Machacek found the best model to explain the angular pressure variation along the cold front in NGC1404 had an inclination angle for the infall of NGC1404 of about 33 degrees, which then allowed the determination of a Mach number of 1.3 for NGC1404’s infall velocity. Figure 1 shows the optical image (left) and deep Chandra image of NGC1399 and NGC1404.

While the papers above presented examples of the stripping of hot gas from early type galaxies, the paper "A 70 Kiloparsec tail in the Cluster A3627" by Sun, Jones, Forman, Nulsen, Donahue and Voit (ApJ Letters 637, L81 and the paper "Spectacular X-ray Tails, Intracluster Star Formation and ULX's in A3627" by Sun, Donahue, Roediger, Nulsen, Voit, Sarazin, Forman and Jones presented the discovery and study of an X-ray tail associated with a spiral galaxy. Using both Chandra and XMM-Newton observations, the authors argued that the X-ray tail is likely stripped cold material, in particular, HI gas from ESO137-001 that is mixed with the hot cluster gas. A stripping timescale longer than ten years would explain the length of the observed tail, but also be short enough to be consistent with the rarity of long tails in spiral galaxies. Because of its proximity, ESO137-001 is the best example known of this type of gas stripping.

The Bullet Cluster (1E-657-56) is undergoing a major merger of two massive subclusters. HST images (PI Jones) were used to map the distribution of dark matter in the cluster, which showed that dark matter has a different distribution from the hot gas, which comprises the bulk of the visible mass. Chandra and JVLA observations were made of the massive clusters in the HST Frontier Fields. The most massive of these, MACS0717.5+375, has at least four merging subclusters and hosts radio relics and a radio halo. Although the Bullet-like merger of the clusters A3411 and A3412 hosts radio relics, the merger is not sufficiently energetic to produce the relics. Instead a spatially extended population of electrons, likely produced by AGN, was present before the merger, and was accelerated by the merger.

The very early view of galaxy clusters as simple, dynamically relaxed systems changed dramatically four decades ago to understanding that galaxy clusters are still growing, even today, through the infall of galaxies, groups and subclusters. This change in our understanding of clusters was due to both X-ray images obtained with the Einstein and ROSAT observatories which showed significant substructures in clusters (see review: Forman & Jones, 1982, ARAA 20, 547-585) and to studies of the distributions of galaxies in clusters (Dressler, 1980, ApJ 236, 351 | Dressler & Shectman, 1988, AJ 95, 985). These cluster mergers leave long-lasting signatures on both the baryonic and non-baryonic cluster constituents, including shock fronts, cold fronts, radio halos and offsets between the dark matter and the hot intracluster gas.

In this section, the results derived from Chandra observations of the merging clusters A2142, and the Bullet Cluster, as well as the more recent studies of the Frontier Fields clusters and the merging clusters A3411-!3411. Below are selected papers describing clusters undergoing significant mergers, which are based on Chandra, XMM-Newton, and JVLA observations.

The paper “Chandra Observations of Abell 2142: Survival of Dense Subcluster Cores in a Merger” (2000, ApJ 541, 542), Markevitch, Ponman, Nulsen, Bautz, Burke, David, Davis, Donally, Forman, Jones, Kaastra, Kellogg, Kim, Kolodzieiczak, Massotta, Pagliaro, Patel, Van Speybroeck, Vikhlinin, Vrtilek, Wise, Zhao) describes the dense subcluster cores in the rich cluster A2142 that have survived the merger of two clusters. The top image of Figure 2 shows the Chandra image, with a sharp, curved, bright edge in the X-ray surface brightness to the northeast, while near the cluster center, there is an inner surface brightness edge. The bottom image shows a contour plot of the X-ray surface brightness on the optical image. The surface brightness profiles across the NE and SW edges is shown in figure 4c. The shape, sharpness of the edges and lack of large temperature jumps across the edges ruled out shock heating due to a cluster merger as the cause of these edges. Instead, Markevitch et al. suggested that the edges are surfaces where the cluster cores are currently being ram pressured stripped by the surrounding hot gas.

The Chandra ACIS image in the left figure shows a surface brightness edge to the northwest and another, near the cluster core, southeast of the cluster center.

Figure 2 (Markevitch et al., 2000, ApJ 541, 542)

One of the best studied merging clusters is the Bullet cluster (1E 0657-56). The Chandra image in shown in figure below. In addition to Chandra observations, deep HST/ACS images were obtained (PI Jones), along with Magellan and VLT observations to map the distribution of the total cluster mass. As more fully described below, the combination of Chandra and HST observations allowed the spatial distribution of dark matter and well as luminous matter in the cluster to be mapped.

In the third paper on the Bullet cluster, “A Direct Empirical Proof of the Existence of Dark Matter” published in 2006 (ApJ Letters 648, 109), Clowe, Bradac, Gonzalez, Markevitch, Randall, Jones, and Zaritsky used Chandra observations to map the distribution of the hot gas, along with HST/ACS, Magellan and VLT observations to map the distribution of the dark matter through gravitational lensing. As shown in Figure 1 below, from the weak lensing reconstruction, two major peaks are found in the distribution of the dark matter. These lensing peaks require unseen matter concentrations that are more massive than, and offset from the hot plasma. The observed displacement of the bulk of the baryons (the hot gas) from the cluster gravitational potential proves the presence of dark matter, for the most general assumptions regarding the behavior of gravity. 

Left panel: Color image from the Magellan images of the merging cluster 1E 0657-558, with the white bar indicating 200 kpc at the distance of the cluster. Right panel: 500 ks Chandra image of the cluster. Shown in green contours are the weak lensing reconstructions of the distribution of the dark matter.

Figure 1 (Clowe et al., 2006, ApJ Letters 648, 109)

Finally, in the paper “Strong and Weak Lensing United III. Measuring the Mass Distribution in the Merging Galaxy Cluster 1E0567-56,” (2006, ApJ 652,937), Bradac, M, Clowe ,D., Gonzalez, A., Marshall, P., Forman, W., Jones, C., Markevitch, M., Randall, S., Schrabback, T., and Zaritsky, D. again used the HST ACS images to detect significantly more arcs due to gravitational lensing, than in the previous analysis. Within 250 kpc radii, a mass for the main cluster component of 2.8 1014 solar masses was measured along with a mass of 2.3 1014 solar masses for the Bullet subcluster. The majority of the cluster mass is spatially coincident with the galaxies, which implies that the cluster mass is dominated by a relatively collisionless form of dark matter. Figure 4 shows the color composite image of the Bullet cluster with the contours of the surface mass density determined from the combined weak and strong lensing mass reconstruction shown in red and contours from the Chandra X-ray surface brightness map overlaid in white. This work confirmed the result by Clowe et al. 2004 that the total mass in the cluster does not trace the baryonic mass.

The Frontier Fields program was an ambitious HST and Spitzer observing campaign to look deeper into the high redshift Universe than ever before and to set the stage for future JWST studies of the early Universe. In particular, six fields were selected in which a foreground massive cluster of galaxies with a redshift between 0.3 and 0.55 served as a gravitational lens, thus allowing very deep observations (Lotz et al., 2017, ApJ 837, 97). Figures 4, 5 and 6 are prime examples of the HST deep observations. To complement these optical and infrared observations, Jones led observing proposals for deep Chandra, JVLA and for one frontier cluster, XMM-Newton observations. These observations were used to study the lensing clusters, and included the discovery of radio relics, as well as to identify gravitationally lensed X-ray and radio sources. Below are selected highlights from published papers based on these observations. While Jones and Steve Murray were the PI’s for the deep X-ray and radio observations, the first authors on the papers below were very talented CfA post-doctoral fellows, who led the analysis and interpretation of the X-ray and radio observations of the Frontier Fields.

Ogrean, van Weeren, Jones, Forman, Dawson, Golovich, Andrade-Santos, Murray, Nulsen, Roediger, Zitrin, Bulbul, Kraft, Goulding, Umetzu, Mroczkowski, Bonafede, Randall, Sayers, Churazov, David, Merten, Donahue, Mason, Rosati, Vikhlinin, and Ebeling (2016, ApJ 819, 113) authored the paper “Frontier Fields Clusters: Deep Chandra Observations of the Complex Merger MACSJ1149.6+222, based on deep 365 ksec Chandra observations. This cluster is one of the most complex merging systems with four dark matter halos. The cluster, as shown in the Chandra images (Figures 4 and 9) has an elongated structure. A surface brightness edge northeast of the cluster center is likely a merger cold front. Figure 9 shows the HST RGB image with the Chandra surface brightness contours overlaid.

The contours of the 1-4 GHz high resolution radio continuum emission, along with the X-ray emission from Chandra, shown in blue for the A2744 cluster, are superposed on the Subaru image and shown in Figure 1 (Pearce et al., 2017, ApJ 845, 81). A2744 hosts both a giant radio halo with an extent of about 2.1 Mpc, as well as a radio relic east of the core with an extent of 1.5 Mpc and a diffuse, radio source south of the cluster core, with an elongation of 1.15 Mpc. Since radio relics likely trace particles that have been re-accelerated by cluster mergers (e.g. Brunetti, Setti, Feretti, Giovannini, 2001, MNRAS 320, 365 | Petrosian, 2001, ApJ 557, 560), by mapping radio relics, one can better determine a cluster’s merger history. For A2744, the measured spectral index of the relic corresponds to a Mach number for the shock of about 2. A gas density jump at the position of another radio relics corresponds to a shock front Mach number of 1.3. The detection of two additional large relics shows there is even more complexity in this cluster than previously known. Creating the relic R1 (in Figure 1), requires the reacceleration by the merger shock of a previously existing pool of mildly relativistic fossil electrons.

The paper “Chandra and JVLA Observations of the HST Frontier Fields Cluster MACS J0717.5+3745” by van Weeren, Ogrean, Jones, Forman, Andrade-Santos, Pearce, Bonafede, Bruggan, Babul, Clarke, Churazov, David, Dawson, Donahue, Goulding, Kraft, Mason, Merton, Mroczkowski, Nulsen, Rosati, Roediger, Randall, Sayers, Umetzu, Vikhlinin, Zitrin (2017, ApJ 835, 197) used both Chandra and JVLA observations to show that this very massive cluster has at least four merging subclusters and hosts a complex radio relic and radio halo, in addition to several filamentary radio sources with sizes of 100-300 kpc. A narrow-angled tailed radio galaxy is proposed to provide the fossil electrons that have been reaccelerated by a merger shock to produce the radio relic. 

The HST Frontier Fields cluster MACS J0416.1-2403, at a redshift of 0.396, is a massive merging system with two main subclusters separated by 250 kpc on the sky and two additional smaller mass subclusters. Results from the analysis of deep (324 ks) Chandra observations, along with radio observations from the JVLA and GMRT and HST observations were reported in the paper “Frontier Fields Clusters: Chandra and JVLA View of the Pre-merging Cluster MACS J0416.1-2403” by Ogrean, van Weeren, Jones, Clarke, Sayers, Mroczkowski, Nulsen, Forman, Murray, Pandey-Pommier, Randall, Churazov, Bonafede, Kraft, Forman, Dawson, Golovich, Andrade-Santos, Murray, Nulsen, Roediger, Zitrin, Bulbul, Kraft, Goulding, Umetzu, Mroczkowsku, Bonafede, Randall, Sayers, Churazov, David, Merten, Donahue, Mason, Rosati, Vikhlinin, Ebeling (2015, ApJ 819, 113). Since the lensing map shows two less massive structures, in addition to the two main substructures (Jauzac, Richard, Jullo et al., 2015, MNRAS 446, 4132), Ogrean and her colleagues compared the positions of the dark matter components with the hot ICM and found, that, unlike in the Bullet cluster, the peaks of the hot ICM are co-located with the dark matter components. This can be explained if the subclusters have not yet interacted, or if the dark matter in the subclusters overlaps only in projection, or if the subclusters have experienced a core passage and the gas in the SW cluster has been “slingshotted” and has caught up with the DM halo.

In addition to the X-ray observations of the Frontier Fields Clusters described above, Jones and her colleagues also obtained JVLA radio observations for the Frontier Fields Clusters A370, A2744, and AS1063. Since radio relics likely trace particles that have been re-accelerated by cluster mergers (e.g. Brunetti, Setti, Feretti, Giovannini, 2001, MNRAS 320, 365 | Petrosian, 2001, ApJ 557, 560), by mapping radio relics, one can better determine a cluster’s merger history.

The recent paper “The Discovery of Radio Halos in the Frontier Fields Clusters Abell S1063 and Abell 370” by Xie, van Weeren, Lovisari, Andrade-Santos, Botteon, Bruggen, Bubal, Churazov, Clarke, Forman, Intema, Jones, Kraft, Lal, Mroczkoqski, and Zitrin (2020, A&A 636, 3) used JVLA and GMRT observations, along with Chandra and XMM-Newton observations to discover and characterize radio halos in two Frontier Fields clusters Abell S1063 and Abell 370. Figure 9 shows contours of the diffuse radio emission found at 3.0 GHz for A370 on the XMM-Newton 0.3-7 keV image, along with dashed blue lines indicating the X-ray surface brightness edges found by Botteon et al (2018, MNRAS 476, 5591). As found for other clusters, the halo likely results from a cluster merger.

Planck was the first European mission to study the Cosmic Microwave Background (CMB) over the full sky. As the photons from the CMB traverse the Universe and encounter a cluster of galaxies, their spectral energy distribution is modified. In particular, the hot gas in a cluster scatters the radiation from the very early Universe. This effect, first predicted by Sunyaev and Zeldovich (1972, Comments on Astrophysics and Space Science 4, 173), allowed the Planck mission to detect large numbers of clusters of galaxies. Chandra and XMM-Newton observations were used to map the density and temperature of the hot intracluster medium for a sample of 164 clusters detected by Planck, within a redshift of 0.35. Along with new X-ray observations (PI Jones), JVLA radio observations were obtained. Selected scientific highlights from published papers based on these observations are presented below.

Although many of the clusters detected by Planck are undergoing mergers, one of the most remarkable is the A3411-A3412 system (aka PLCKESZ G241.97+14.98), which provides compelling evidence for electron re-acceleration at the locations of cluster shocks. The cluster pair A3411-A3412 was observed with Chandra ACIS-I for a total exposure time of 211 ks. The first paper on these merging clusters (“Complex Diffuse Radio Emission in the Merging Planck ESZ Cluster A3411” by van Weeren, Fogarty, Jones, Forman, Clarke, Bruggen, Kraft, Lal, Murray, Rottgering (2013, ApJ 769, 101) was followed by the paper “Chandra Observations of the Spectacular A3411-12 Merger Event” (2019, ApJ 887, 31). Using optical and radio observations in addition to the deep Chandra observations, Andrade-Santos, van Weeren, Di Gennaro, Whitman, Ryu, Lal, Placco, Fogarty, Ge, Stroe, Sobral, Forman, Jones, Kraft, Murray, Bruggan, Kang, Santucci, Golovich, and Dawson reported compelling evidence for electron re-acceleration at the locations of cluster shocks. In particular, they showed that the relatively small (<1.15) Mach number of the merging “bullet-like” subcluster in the A3411-A3412 system is not sufficient to produce the observed radio relics that span several Mpc. Instead the existence of a population of energetic elections in extended regions of the cluster must have been present prior to the cluster merger, over extended regions of the cluster.

The left panel of Figure 3 shows the relation of the total cluster X-ray luminosity with total mass for relaxed clusters in blue and disturbed clusters in red, while the right panel of Figure 3 shows the same relations with the X-ray emission from the cluster cores excluded.

The source of this proposed population of energetic electrons in the merging cluster A3411-12 was identified in the paper “The Case for Electron Re-acceleration at Galaxy Cluster Shocks” van Weeren, Andrade-Santos, Dawson, Golovich, Lal, Kang, Ryu, Bruggen, Ogrean, Forman, Jones, Placco, Santucci, Wittman, Jee, Kraft, Sobral, Stroe, Fogarty (2017, Nature Astronomy 1E, 5). This discovery showed that “fossil” relativistic electrons from active galactic nuclei can be re-accelerated by small Mach number cluster shocks, thus solving the puzzle of how radio relics are produced in minor cluster mergers.

In addition to individual clusters that have been studied using the Chandra and XMM-Newton observations, two papers have compared the fractions of morphologically disturbed clusters in X-ray samples compared to clusters in the Planck sample. In particular the paper “The Fraction of Cool-Core Clusters in X-ray verses SZ Samples Using Chandra Observations” by Andrade-Santos, Jones, Forman, Lovisari, Vikhlinin, van Weeren, Murray, Arnaud, Pratt, Democles, Kraft, Mozzotta, Bohringer, Chon, Giacintuicci, Clarke, Borgani, David, Douspis, Pointecouteu, Dahle, Brown, Aghanim, Rasia (2017, ApJ 843, 76) examined the fractions of cool-core clusters in the Chandra-Planck survey with an X-ray flux limited sample of 100 clusters with redshifts < 0.3. They found that the sample of X-ray selected clusters contained a significantly larger fraction of cool core clusters than does the sample of Planck SZ selected clusters, and concluded this is due to the fact that cool core clusters are more X-ray luminous for a fixed cluster mass than are SZ selected clusters. Thus, if cluster masses are estimated through the scaling relation between X-ray luminosity and mass, as will often be the case for clusters in the eROSITA sky survey, the masses for cool core clusters will be biased high, which will cause the determinations of Ω and σ to be shifted to higher values.

In the paper “X-ray Morphological Analysis of Planck ESZ Clusters” (2017, ApJ 846, 51), Lovisari, Forman, Jones, Ettori, Andrade-Santos, Arnaud, Democles, Pratt, Randall and Kraft used XMM-Newton observations to characterize the dynamical state of 189 clusters that comprise the Planck ESZ sample (Planck Collaboration et al. 2011, A&A 536, A9) and found that samples of SZ selected clusters are more dynamically disturbed than X-ray selected cluster samples. The XMM-Newton images of these clusters are shown in Figure 17.

In the recent paper “X-ray Scaling Relations for a Representative Sample of Planck-selected Clusters Observed with XMM-Newton” (2020, ApJ 892, 102), Lovisari, Schellenberger, Sereno, Ettori, Pratt, Forman, Jones, Andrade-Santos, Randall, and Kraft (2020, ApJ 892, 102) reported on their analysis of XMM-Newton observations for 120 galaxy clusters from the Planck ESZ sample in the redshift range from 0.059 to 0.546. They found that relaxed clusters have X-ray luminosities that are on average 30% higher that disturbed clusters with the same mass, but that excising the cluster core before measuring the X-ray luminosity reduces the scatter and brings into better agreement the luminosity-mass and luminosity-gas temperature relations determined for different samples. More importantly, this paper reported, for the first time, a significant evolution with redshift of the relation between the total cluster mass and the temperature of the hot intracluster gas. This is consistent with higher redshift clusters being, on average, more disturbed than lower redshift clusters. The left panel of Figure 3 shows the relation of the total cluster X-ray luminosity with total mass for relaxed clusters in blue and disturbed clusters in red, while the right panel of Figure 3 shows the same relations with the X-ray emission from the cluster cores excluded. Using these core-excised luminosities, the relaxed and disturbed systems have very similar relations. Figure 5 shows the tight relation between the total cluster mass and the gas mass. 

In the early 1990’s, the best studied and probably best understood gravitational lens system was 0957+561, which is comprised of a quasar at redshift 1.41 that is lensed by an intervening galaxy and cluster of galaxies at a redshift of 0.36. In the paper “Discovery of a New Component in the Gravitationally Lensed Quasar 0957+561” (1993, ApJ 410, 21), Jones, Stern, Falco, Forman, David, Shapiro, and Fabian reported the detection of point-like X-ray emission from the southern lensed image of the quasar and a semi-circular arc of extended X-ray emission between the two images of the quasar, based on Einstein observations, as shown in Figure 1.

The Einstein HRI observations of 0957+561 were followed by ROSAT High Resolution Camera (HRC) observations. Jones was the ROSAT PI and George Chartas was the first author of two papers based on these observations. In the paper “ROSAT Observations of the Gravitationally Lensed System 0957+561,” Chartas, Falco, Forman, Jones, Schild, and Shapiro (1995, ApJ 445, 140) reported X-ray intensity increases by a factor of five for one of the lensed images between the time of the Einstein (late 1970’s) and the ROSAT observations (early 1990’s).

In the second paper, based on the ROSAT HRI observations of Q0957+561 “X-ray Detection of the Primary Lens Galaxy Cluster of the Gravitational Lens System Q0957+561” published in 1998 (ApJ 504, 661), Chartas, Chuss, Forman, Jones, and Shapiro detected X-ray emission from the lensing cluster of galaxies at z=0.36, determined the total cluster mass to be in the range from 1.5 to 3.2 X 1014  solar masses, and for a time delay of 1.1 years, constrained the Hubble constant to be in the range from 67 to 82 km per second per Mpc.

In 2000, Jones was the PI for a deep 46 ksec Chandra ACIS observation of the lensed QSO 0957+561. Figure 2 from the paper “Constraining H0 from Chandra observations of Q0957+561” by Chartas, Gupta, Garmire, Jones, Falco, Shapiro, and Avecchio (2002, Ap J. 565, 96), shows the X-ray intensity map with the outline of the lensed radio emission from the AGN core and jet. Three X-ray knots are detected in the jet, which extends 8” to the northeast of image “A.” When this paper was written, this jet had the highest known redshift (z=1.41). The gas temperature, core radius and shape of the lensing cluster were also determined from the Chandra observation.

The Frontier Fields clusters were chosen for deep HST observations, primarily because these massive clusters would be strong gravitational lenses, allowing the detection of very high redshift galaxies, prior to the launch of JWST. The paper “The Discovery of Lensed Radio and X-ray Sources Behind the Frontier Fields Cluster MACS J0717.5+3745 with the JVLA and Chandra” by van Weeren, Ogrean, Jones, Forman, Andrade-Santos, Bonagede, Bruggen, Bulbul, Clarke, Churazov, David, Dawson, Donahue, Goulding, Kraft, Mason, Merten, Mroczkowsky, Murray, Nulsen, Rosati, Roediger, Tandal Sayers, Umetsu, Vikhlinn, and Zitrin (2016, ApJ 817, 98) reported the detection of 51 compact high redshift radio sources in the region of the HST image. The majority of these sources are likely star-forming galaxies in the redshift range 0.6 – 2.0. Sixteen of these are located behind the cluster and 7 of these 16 have amplification factors greater than 2. From the derived radio luminosity function, van Weeren and colleagues found evidence for an increase in the number density of radio sources in this redshift range, compared to that measured at z<0.3. Figure 2 shows radio intensity contours of these compact lensed radio sources.

Currently, Jones is a member of the scientific team for the HST Treasury Program RELICS (Reionization Lensing Cluster Survey, PI D. Coe). This project was conceived during the 2015 workshop “Science from the Frontier Fields” held in Sesto, Italy, where Jones was invited to present results from Chandra and radio observations of the Frontier Field clusters. The Frontier Fields have been extremely rewarding scientifically (see for example the papers on mergers in Frontier Fields clusters in section “Chandra & XMM-Newton: Cluster Mergers and the Production of Radio Relics”). Each Frontier Field was observed with HST for 140 orbits. Although numerous gravitationally lensed objects were found, many were extremely faint optically, severely limiting follow-up observations. Instead of very deep observations of a few clusters, in the RELICS program, 41 massive clusters were each observed with HST for five orbits, in seven filters. (28 of the 41 RELICS clusters had some previous HST observations.) RELICS clusters were selected to allow the detection of more than 100 high redshift galaxies (z=9-12), and approximately 170 z=8 galaxies, that would be bright enough for JWST follow-up observations. In addition to the HST observations, Spitzer IRAC programs, for a total of 945 hours were obtained for these clusters (PI’s Bradac and Soifer) in addition to the ~100 hours of archival Spitzer data for 18 RELICS clusters. Thirteen papers have been published or submitted to date based on the RELICS program. Except for the paper “RELICS: Reionization Lensing Cluster Survey” which was led by RELICS HST PI Dan Coe, the additional twelve published or submitted RELICS papers are led by more junior scientists.

The overview paper “RELICS: Reionization Lensing Cluster Survey” by Coe, Salmon, Bradac, Bradley, Sharon, Zitrin, Acebron, Cerney, Cibirka, Strait, Paterno-Mahler, Mahler, Avela, Ogaz, Juang, Pelliccia, Stark, Mainall, Oesch, Trenti, Carrasco, Dawson, Rodney, Strolber, Reiss, Jones, Frye, Czakon, Umetsu, Vulcani, Graur, Jha, Graham, Molino, Nonino, Hjorth, Selsing, Christensen, Kikuchihara, Ouchi, Oguri, Welch, Lemaux, Andrade-Santos, Hoag, Johnson, Peterson, Past, Fox, Agull, Livermore, Ryan, Lam, Sendra-Server, Toft, Lovisari, and Su was published in 2019 (ApJ 884, 85). Figure 2 shows the RELICS clusters as red squares on a plot of cluster mass vs redshift for all 1094 clusters in the Planck PSZ2 catalog. A horizontal dashed line at a mass of 8.7 1014 Msun separates the 21 RELICS clusters selected solely based on their mass from the 20 RELICS clusters selected based on other criteria.

One of the strong lensing models that were derived for clusters in the RELICS HST program is shown in Figure 2 (Acebron, Cibirka, Zitrin, Coe, Agulli, Sharon, Frye, Livermore, Mahler, Salmon, Umetzu, Bradley, Andrade-Santos, Abila, Carrasco, Cerny, CZakon, Dawson, Hoag, Huang, Johnson, Jones, Kikuchihara, Lam, Lovisari, Mainali, Oesch, Ogaz, Ouchi, Past, Paterno-Mahler, Peterson, Ryan, Sendra-Server, Stark, Strait, Toft, Trenti, Vulcani, 2018, ApJ 858, 42). Three lensed objects with redshifts of ~6 were found in this field.

The HST images of a possible redshift 10 galaxy, that is gravitationally lensed into an arc stretching about 2.5” on the sky is shown in Figure 1 (Salmon, Coe, Bradley, Bradac, Strait, Paterno-Mahler, Huang, Oesch, Zitrin, Acebron, Cibirka, Kikuchihara, Oguri, Brammer, Sharon, Trenti, Avila, Ogaz, Andrade-Santos, Carrasco, Cerny, Dawson, Grye, Hoag, Jones, Mainali, Ouchi, Rodney, Stark, Umetsu (2018, ApJ Letters 864, 22).

The recent paper, “RELICS: The Reionization Lensing Cluster Survey and the Brightest High-z Galaxies” by Salmon, Coe, Bradley, Bouwens, Bradac, Huang, Oesch, Stark, Sharon, Trenti, Avila, Ogaz, Andrade-Santos, Carrasco, Cerny, Dawson, Grye, Hoag, Johnson, Jones, Lam, Lovisari, Mainali, Past, Paterno-Mahler, Perterson, Riess, Rodney, Ryan, Sendra-Server, Strolger, Umetsu, Vulcani, Zitrin (2020, ApJ 889, 189), describes the 321 candidate galaxies with redshifts between z~6 and z~8 from the RELICS survey. From the total of more than 76,000 sources in the RELICS catalogs, 1337 were identified at likely being at redshifts z>5.5. The high redshift sample was further refined to a sample of 841 objects by selecting galaxies with an F160W detection greater than 3 σ. After further examination to remove galactic and dwarf stars and image artifacts, the final RELICS sample contained 255 galaxies at z~6, 56 at z~7 and 8 at z~8.

Figure 3 shows HST images for nine of the ten cluster fields with the highest numbers of lensed galaxies with redshifts greater than 5.5. The cyan, magenta, and yellow circles mark redshift 6, 7, and 8 candidate galaxies. Figure 12 shows all 8 galaxy candidates at redshift ~8. In addition, one candidate redshift 10 galaxy has been found in the RELICS program.