Clues on Arp 142: The Spiral-Elliptical merger
Marcelo D. Mora, Sergio Torres-Flores, Veronica Firpo, Jose A. Hernandez-Jimenez, Fernanda Urrutia-Viscarra, Claudia Mendes de Oliveira
MMNRAS , 000–000 (0000) Preprint 5 July 2019 Compiled using MNRAS L A TEX style file v3.0
Clues on Arp 142: The Spiral-Elliptical merger
Marcelo D. Mora, (cid:63) , Sergio Torres-Flores , Ver ´onica Firpo , , Jose A. Hernandez-Jimenez , , Fernanda Urrutia-Viscarra , , Claudia Mendes de Oliveira Instituto de Astrof´ısica. Pontificia Universidad Cat´olica de Chile. Vicu˜na Mackenna 4860, 7820436 Macul Santiago Chile. Departamento de F´ısica y Astronom´ıa, Universidad de La Serena, Av. Cisternas 1200, La Serena, Chile Gemini Observatory, Southern Operations Center, La Serena, Chile Departamento de Ciencias F´ısicas, Universidad Andres Bello, Fernandez Concha 700, Las Condes, Santiago, Chile Departamento de Astronomia, Instituto de Astronomia, Geof´ısica e Ciˆencias Atmosf´ericas da USP, Rua do Mat˜ao 1226,Cidade Universit´aria, 05508-090, S˜ao Paulo, Brazil
ABSTRACT
Nearby merging pairs are unique laboratories in which one can study the gravitational e ff ectson the individual interacting components. In this manuscript, we report the characterization ofselected H ii regions along the peculiar galaxy NGC 2936, member of the galaxy pair Arp 142,an E + S interaction, known as “The Penguin”. Using Gemini South spectroscopy we havederived a high enhancement of the global star formation rate SFR = (cid:12) yr − probablystimulated by the interaction. Star-forming regions on this galaxy display oxygen abundancesthat are consistent with solar metallicities. The current data set does not allow us to con-clude any clear scenario for NGC 2936. Diagnostic diagrams suggest that the central regionof NGC 2936 is ionized by AGN activity and the eastern tidal plume in NGC 2936 is expe-riencing a burst of star formation, which may be triggered by the gas compression due tothe interaction event with its elliptical companion galaxy: NGC 2937. The ionization mech-anism of these sources is consistent with shock models of low-velocities of 200-300 km s − .The isophotal analysis shows tidal features on NGC 2937: at inner radii non-concentric (oro ff -centering) isophotes, and at large radii, a faint excess of the surface brightness profile withrespect to de Vaucouleurs law. By comparing the radial velocity profiles and morphologi-cal characteristics of Arp 142 with a library of numerical simulations, we conclude that thecurrent stage of the system would be about 50 ±
25 Myr after the first pericenter passage.
Key words: galaxies: interactions, ISM: H ii regions Galaxy interaction is a process that not only transforms and de-stroys galaxies, it is also a potential way for internal componentssuch as black holes, star-forming regions, and other features to beassembled. Merging is one of the key predictions of the cold darkmatter model, which is the dominant scenario for galaxy evolution(e.g. White & Rees 1978; Blumenthal et al. 1984; Cole et al. 2000;Conselice et al. 2014), where the merging of dark matter halos leadto the merging of the galaxies associated with the halos (Kau ff mannet al. 1993), process that it is expected to keep occurring as darkmatter haloes continue to merge along cosmic time. The dense envi-ronment plays a role that a ff ects mass growth through gravitationalharassment or stripping, removing a significant part of the gas con-tent of a galaxy (e.g. Moore et al. 1996). These processes may ex-plain the mass distribution observed in the well-defined Hubble se-quence of galaxy types in the nearby universe (Tasca et al. 2014). (cid:63) Contact e-mail: [email protected]
Merging and interacting galaxies are actually ideal places tostudy the e ff ects of the gravitational encounters in the structure ofgalaxies. Di ff erent authors have studied the kinematics of merg-ing / interacting galaxies (e.g. Amram et al. 2003), the star forma-tion rate properties of these systems (e.g. Ellison et al. 2008), andthe morphology of these perturbed objects (e.g. Gallagher & Parker2010). These studies have found peculiar kinematics, enhanced starformation activity, and perturbed morphologies. In the same con-text, Xu et al. (2010) studied a sample of major mergers of galaxypairs. These authors found that spiral-spiral pairs (S + S) displayenhanced star formation rates with respect to a sample of non-interacting systems. The same authors do not find an enhancementin the SFRs of spiral members of spiral-elliptical pairs (S + E). Inthe context of S + E pairs, Weistrop et al. (2012) studied the minormerger NGC 4194, thought to be in an advanced stage of evolutionwhich consists of a gas-rich spiral falling onto an elliptical galaxy(Manthey et al. 2008). These authors found di ff erent stellar pop-ulations in this system (ranging from 10 Myr up to 5 Gyrs), andthey argued that part of this population came from the progenitor c (cid:13) a r X i v : . [ a s t r o - ph . GA ] J u l M. D. Mora et al. galaxies. J¨utte et al. (2010) observed the molecular gas content ofthe advance S + E merger NGC 4441. They found a moderate starformation rate (1-2 M (cid:12) yr − ) with no evidence of cold dense coresof ongoing star formation, thus NGC 4441 was found to be an ex-ample galaxy, in which the starburst has already faded. Authorscited above are one of the few studies that focused on the physicalproperties of star-forming regions located in spiral galaxies belong-ing to S + E pairs in an early close encounter stage. At early stages,gas clouds collide and the star formation is triggered on the spi-ral galaxy forming young objects like star clusters or complexes ofclusters. (e.g. Holtzman et al. 1992; Whitmore et al. 1999; Kauf-man et al. 2012). Then, a detailed study of the physical propertiesof the newly formed regions can give us important information re-garding the interaction process that has taken place during the pairinteraction. Finally, these young regions can be used to trace thegas phase oxygen abundance distribution of these interacting spiralgalaxies. In this work, we analyze the spectacular interacting pairArp 142, focusing on the link between the interaction, the newlyformed regions, the gas, and the star formation feedback.The interacting system Arp 142 (Arp 1966) is located at a dis-tance of ∼
100 Mpc (corrected by the influence of Virgo, Great At-tractor and Shapley, Mould et al. 2000) in the direction of the Hydraconstellation. The system is composed by the spheroidal NGC 2937and the spiral galaxy NGC 2936 that displays a complex morphol-ogy. The latter has been classified as a ring galaxy by Freeman& de Vaucouleurs (1974) and Madore et al. (2009). However, Ro-mano et al. (2008) described it as an arc-like galaxy with no evi-dence of being a twisted ring, suggesting a nearly edge-on view ofthe system. Xu et al. (2010) report the detection of several infraredbright star-forming regions in the tails of NGC 2936, suggesting thepresence of dust. The peculiar morphology of NGC 2936 resemblesthat of a bird head, where the nucleus corresponds to the eye of thebird (Xu et al. 2010). Other authors have suggested that NGC 2936mimics a penguin thus the denomination of penguin galaxy. Visualinspection of Hubble Space Telescope (HST) WFC3-UVIS2 (Hub-ble proposal ID:12812, PI: Levay) imaging shows that NGC 2936and NGC 2937 are distinguishable and not fully mixed yet (see Fig-ure 1), hinting an early stage of the merger (or a going on close en-counter). Furthermore, the high spatial resolution of the HST imageshows clearly that arc-like or ring-like forms described by earlierauthors is actually due to the North-West tidal plume stems fromNGC 2936 and bending towards the elliptical member of Arp 142(NGC 2937). The former does not show visual evidence of shells,but in this work we explore signs of distortion using an isopho-tal analysis, while the spiral component (NGC 2936) shows thepresence of several objects with colors consistent with those ofyoung star-forming complexes located in one of the tidal fields ofNGC 2936, hinting a spiral arm under current disruption. Imagesalso suggest a faint feature seen in projection toward the ellipticalon the direction of the semi-major axis. In addition,
Sloan DigitalSky Survey spectra located in the distorted spiral arm of NGC 2936shows strong emission-lines, which suggests the presence of youngmassive stars ionizing the gas. In this work, we perform a spec-troscopic analysis in a sample of star-forming regions located inNGC 2936, which is complemented with high quality archival datafor the whole interacting system, with the aim of understandinghow the local star formation and the physical conditions of the ion-ized gas were a ff ected -or not- by the interaction. In addition, wediscuss the global picture of the merging scenario for this object,which adds important pieces to understand the localized star for-mation in Arp 142. Of course, a detailed analysis of localized star-forming regions can be a ff ected by sample e ff ects (e.g. sample size, selection bias, etc), however, very fruitful insights can be obtainedfrom a system that has been not studied in detail in the past. Thiswork is structured as follows: in section 2 we present the observa-tions, data reduction procedures, and object selection. In section 3we present the data analysis, in section 4 we discuss our result andfinally in section 5 we draw our conclusions. Observations were acquired in queue mode using the GeminiMulti-Object Spectrograph (GMOS) mounted at Cerro Pach´on,Gemini South Telescope (program ID: GS-2014B-Q77, PI: Mendesde Oliveira). We have used the B
600 disperser with a slit widthof 1 (cid:48)(cid:48) pointed into 4 places (see Figure 1): near the center of theNGC 2936 (Slit 2), across the spiral arms (Slit 1 and Slit 3), andalong several star-forming regions located across the southern tidaltail of NGC 2936 (Slit 4). For each pointing, a set of 3 exposures of900 sec were acquired including a shift of 5Å in the central wave-length aimed to avoid the Hamamatsu CCD gaps. Each science ob-servation was then followed by its corresponding arcs and flats aim-ing at minimizing wavelength di ff erences due to telescope flexures.The spectral range covered by the observations was from 4200Åto 7400 Å. The typical seeing was consistent with the requestedweather conditions ( ∼ (cid:48)(cid:48) ) and it was of the order of 1 (cid:48)(cid:48) . The spec-tral resolution of the observations was 4.5 Å, as measured from thefull width at half-maximum (FWHM) on the 5557 Å sky line.Data reduction was done using Gemini GMOS routines within pyraf / iraf . Raw images were bias and flat corrected. Cosmiccontamination of the CCDs were removed using the La Cosmic routine (van Dokkum 2001) which was applied to individual im-ages. Unfortunately, our images were a ff ected by the hot columnproblem of the Amplifier 5 in the Central CCD Hamamatsu, there-fore we have developed a script to correct it. The code is availableat github . Flats and science were quantum e ffi ciency corrected.Wavelength calibration was done following the Gemini pyraf/iraf package and the typical dispersion was 0.02 Å. Sky was later sub-tracted and images were finally combined into a master scienceframe. The standard star LTT 1020 was observed for flux calibra-tion purposes, with an exposure time of 60 sec and it was reducedfollowing Gemini standard procedures. In addition, HST images(WFC3-UVIS2) used in this work (filters F W , F W , F N ,and F W with exposures times of 2745 sec, 1500 sec, 3600 sec,and 2640 sec respectively) were downloaded from the HST archive. Since star cluster complexes are brighter in blue passband than inthe redder ones, we used the HST images to compose a color im-age using the F W (blue), F W (green) and, F W (red)passband, selecting the blue knots that hints to be located in star-forming regions which maximizes the number of object to be ob-served on each slit. Owing to the seeing limited observations, star-forming regions in NGC 2936 appear as di ff use objects across thedi ff erent slits. Therefore we used the g-filter GMOS-South acqui-sition image to cross identify individual regions on the 2D spectra PYRAF is a product of the Space Telescope Science Institute, which isoperated by AURA for NASA. Image Reduction and Analysis Facility, distributed by NOAO, operatedby AURA, Inc., under agreement with NSF https: // github.com / mmorage / Scripts.- / blob / master / Amp5 v3 END.pyMNRAS000
600 disperser with a slit widthof 1 (cid:48)(cid:48) pointed into 4 places (see Figure 1): near the center of theNGC 2936 (Slit 2), across the spiral arms (Slit 1 and Slit 3), andalong several star-forming regions located across the southern tidaltail of NGC 2936 (Slit 4). For each pointing, a set of 3 exposures of900 sec were acquired including a shift of 5Å in the central wave-length aimed to avoid the Hamamatsu CCD gaps. Each science ob-servation was then followed by its corresponding arcs and flats aim-ing at minimizing wavelength di ff erences due to telescope flexures.The spectral range covered by the observations was from 4200Åto 7400 Å. The typical seeing was consistent with the requestedweather conditions ( ∼ (cid:48)(cid:48) ) and it was of the order of 1 (cid:48)(cid:48) . The spec-tral resolution of the observations was 4.5 Å, as measured from thefull width at half-maximum (FWHM) on the 5557 Å sky line.Data reduction was done using Gemini GMOS routines within pyraf / iraf . Raw images were bias and flat corrected. Cosmiccontamination of the CCDs were removed using the La Cosmic routine (van Dokkum 2001) which was applied to individual im-ages. Unfortunately, our images were a ff ected by the hot columnproblem of the Amplifier 5 in the Central CCD Hamamatsu, there-fore we have developed a script to correct it. The code is availableat github . Flats and science were quantum e ffi ciency corrected.Wavelength calibration was done following the Gemini pyraf/iraf package and the typical dispersion was 0.02 Å. Sky was later sub-tracted and images were finally combined into a master scienceframe. The standard star LTT 1020 was observed for flux calibra-tion purposes, with an exposure time of 60 sec and it was reducedfollowing Gemini standard procedures. In addition, HST images(WFC3-UVIS2) used in this work (filters F W , F W , F N ,and F W with exposures times of 2745 sec, 1500 sec, 3600 sec,and 2640 sec respectively) were downloaded from the HST archive. Since star cluster complexes are brighter in blue passband than inthe redder ones, we used the HST images to compose a color im-age using the F W (blue), F W (green) and, F W (red)passband, selecting the blue knots that hints to be located in star-forming regions which maximizes the number of object to be ob-served on each slit. Owing to the seeing limited observations, star-forming regions in NGC 2936 appear as di ff use objects across thedi ff erent slits. Therefore we used the g-filter GMOS-South acqui-sition image to cross identify individual regions on the 2D spectra PYRAF is a product of the Space Telescope Science Institute, which isoperated by AURA for NASA. Image Reduction and Analysis Facility, distributed by NOAO, operatedby AURA, Inc., under agreement with NSF https: // github.com / mmorage / Scripts.- / blob / master / Amp5 v3 END.pyMNRAS000 , 000–000 (0000) lues on Arp 142 Table 1.
Identification, coordinates and measured radial velocities.
ID RA(J2000) Dec(J2000) V r kms − S1A 9:37:45.519 + ± + ± + ± + ± + ± + ± + ± + ± + ± + ± + ± + ± + ± † + ± + ± † + ± + ± + ± + ± + ± + ± + ± + ± † Same star-forming region observed in two di ff erent slits (S3 and S4). Figure 1.
HST F W image of NGC 2936. Dashed lines represent 1”GMOS slit pointings. Circles correspond to the selected H ii regions ana-lyzed in this manuscript. Labels were assigned according the target positionon their corresponding slits. and we extracted them by considering the area where the H α pro-file, measured in the spatial direction on the 2D spectra, intersectsthe background noise.These criteria yielded a total of 22 extracted regions. We havelabeled our star-forming region candidates following the order thateach region occupies in their respective slit, e.g. S1A is the first re-gion that appearing in the Slit 1, S1B is the second region, and soon. One of 22 regions was extracted in two di ff erent slits (S3E andS4A). In Table 1 we list the identification (ID), RA, Dec, and radialvelocities for the selected regions. In Figure 1 we present the WideField Planetary Camera 3 image acquired on the F W passband,downloaded from the Multimission Archive at the Space TelescopeScience Institute (MAST). Circles represent the identified star-forming regions observed in NGC 2936. In Figure 2 we present theextended spectra (on the left panels) where H α , [N ii ] λ ii ] λ g -bandGMOS-South acquisition images.Finally, spectra were corrected by Galactic extinction assum-ing a color excess of E(B-V) = In this section, we analyze the spectroscopic data for NGC 2936,the spiral galaxy of the Arp 142 pair. We also derive the radialvelocity of the galaxy UGC 05130 NOTES01, located 1.3 arcminnorth of the center of NGC 2631, in projected distance, which isof interest, given that this will give an indication of membership tothe system or not. A photometric analysis of the elliptical galaxyNGC 2937 is also performed, in order to look for isophotal dis-tortions or faint ripples / shells, indicators of interaction. Finally, aGalMer model simulation of the system is attempted and velocitycurves of model and observations are compared. Long-slits placed in NGC 2936 were oriented to maximize thenumber of spectra of the star-forming regions located acrossNGC 2936. Visual inspection of individual spectra show strongnebular emission-lines, the presence of a weak continuum emis-sion which could be associated with the young stellar populationof each star-forming object and the underlying stellar populationof NGC 2936 (except slit S4 where we did not detect continuum).For example, in Figure 3, at the top panel, we show the spec-trum of one source identified in NGC 2936 as S1A (black solidline), and in the bottom panel we show the spectrum of the galaxyUCG 05130 NOTES01. In both cases is clear the presence of strongemission lines, which dominate the spectra.Observed spectra where used to measure the emission-linefluxes by using the iraf splot task and the uncertainties wereestimated using Monte Carlo simulations for 150 runs, given bythe nerrsample parameter of the splot task. Then, for each star-forming region, total extinctions were estimated by comparing theobserved H α / H β ratio with its theoretical value. This correction isspecially important for the case of the star-forming regions whichcoincide with the strong dust lane that crosses this galaxy. We as-sumed a electronic temperature of T e = K and an electron den-sity of N e = cm − , which yields a intrinsic ratio of H α / H β = MNRAS , 000–000 (0000)
M. D. Mora et al.
Figure 2. α ,[N ii ] λ ii ] λ H α emission-line. Rightpanels show g -filter GMOS-South acquisition images with the overlappedslits and the selected regions. starburst extinction law (Calzetti et al. 2000) to correct each indi-vidual object by extinction.The spectral range of our observations allows us to detectthe most intense emission-lines in each star-forming region spec-trum, namely: H γ , H β , [O iii ] λ iii ] λ λ i ] λ ii ] λ α , [N ii ] λ i λ ii ] λ ii ] λ splot task in iraf . Thisprocedure provided us the flux and the associated uncertainty foreach emission-line, which are listed in Table 2. It is well known that the H α luminosity is a good tracer of the cur-rent star formation (Kennicutt & Evans 2012), and its emission canbe associated with the presence of a large number of young andmassive stars whose ultraviolet flux ionize the surrounding gas,where the recombination process is a strong evidence of recent orongoing massive star formation. Hence, the star-forming regionslocated in the interacting system NGC 2936 provide us with a lab-oratory to study the star formation that may have been triggered bygravitational e ff ects or previously triggered by the internal stimula-tion of the spiral arms inherent to a galaxy of this type. In this con-text, a standard procedure would be the estimation of the currentstar formation rate by using H α narrow band imaging. However,the redshift of Arp 142 invalidate this method, given that the H α line is o ff the F657N HST filter. Therefore, H α luminosities werecalculated using the spectroscopic information by measuring thedi ff erent H α fluxes and taking into account the distance to Arp 142.We found H α luminosities to be in the range from log(L H α ) ∼ − , which is consistent with what have been ob-served in giant star-forming regions (E.g. Firpo et al. (2005, 2010,2011), log(L H α ) ∼ − to 40.9 erg s − ) and HII galaxies (e.g.H¨agele et al. (2008), log(L H α ) ∼ − to 41.6 erg s − ). Wethen converted these luminosities into star formation rates (SFRs)using the classical equation given in Kennicutt (1998), which as-sumes a continuous star formation process over the timescale as-sociated with the star-forming tracer. This calibrator considers aSalpeter Initial Mass Function, with masses ranging from 0.1-100Msun. We derive SFR that are in agreement with values found byFerreiro et al. (2008) in the star-forming regions studied in inter-acting galaxies (0.07 to 10 M (cid:12) yr − ). Finally, results are listed inTable 3, where we list the H α luminosities and SFRs measured foreach region belonging to NGC 2936 as well as their values withno reddening corrections (H α uncorr and SFR uncorr ), as well as theirmeasured excess colors [E( B − V )]. We also include the star for-mation rate density for each star-forming region ( Σ S FR ), which wasderived as the SFR divided by the area within which the spectrawere extracted.In the top left panel of Figure 4 we present the spatial dis-tribution of the star formation rates in the regions of NGC 2936.Although this plot does not have the spatial resolution to show lo-calized variations of the quantities, it can give us an idea of trendsand variations seen in our data. From this plot, we can see thatNGC 2936 is actively forming stars in its eastern side. Region S1Bdisplays the highest SFR, which corresponds to 17.9 M (cid:12) yr − . Thissource lies over a dust lane, which produces a high extinction at thislocation (see section 3.2.3). The second most intense star formationburst corresponds to S1A, which is located at the tip of the eastern MNRAS000
Figure 2. α ,[N ii ] λ ii ] λ H α emission-line. Rightpanels show g -filter GMOS-South acquisition images with the overlappedslits and the selected regions. starburst extinction law (Calzetti et al. 2000) to correct each indi-vidual object by extinction.The spectral range of our observations allows us to detectthe most intense emission-lines in each star-forming region spec-trum, namely: H γ , H β , [O iii ] λ iii ] λ λ i ] λ ii ] λ α , [N ii ] λ i λ ii ] λ ii ] λ splot task in iraf . Thisprocedure provided us the flux and the associated uncertainty foreach emission-line, which are listed in Table 2. It is well known that the H α luminosity is a good tracer of the cur-rent star formation (Kennicutt & Evans 2012), and its emission canbe associated with the presence of a large number of young andmassive stars whose ultraviolet flux ionize the surrounding gas,where the recombination process is a strong evidence of recent orongoing massive star formation. Hence, the star-forming regionslocated in the interacting system NGC 2936 provide us with a lab-oratory to study the star formation that may have been triggered bygravitational e ff ects or previously triggered by the internal stimula-tion of the spiral arms inherent to a galaxy of this type. In this con-text, a standard procedure would be the estimation of the currentstar formation rate by using H α narrow band imaging. However,the redshift of Arp 142 invalidate this method, given that the H α line is o ff the F657N HST filter. Therefore, H α luminosities werecalculated using the spectroscopic information by measuring thedi ff erent H α fluxes and taking into account the distance to Arp 142.We found H α luminosities to be in the range from log(L H α ) ∼ − , which is consistent with what have been ob-served in giant star-forming regions (E.g. Firpo et al. (2005, 2010,2011), log(L H α ) ∼ − to 40.9 erg s − ) and HII galaxies (e.g.H¨agele et al. (2008), log(L H α ) ∼ − to 41.6 erg s − ). Wethen converted these luminosities into star formation rates (SFRs)using the classical equation given in Kennicutt (1998), which as-sumes a continuous star formation process over the timescale as-sociated with the star-forming tracer. This calibrator considers aSalpeter Initial Mass Function, with masses ranging from 0.1-100Msun. We derive SFR that are in agreement with values found byFerreiro et al. (2008) in the star-forming regions studied in inter-acting galaxies (0.07 to 10 M (cid:12) yr − ). Finally, results are listed inTable 3, where we list the H α luminosities and SFRs measured foreach region belonging to NGC 2936 as well as their values withno reddening corrections (H α uncorr and SFR uncorr ), as well as theirmeasured excess colors [E( B − V )]. We also include the star for-mation rate density for each star-forming region ( Σ S FR ), which wasderived as the SFR divided by the area within which the spectrawere extracted.In the top left panel of Figure 4 we present the spatial dis-tribution of the star formation rates in the regions of NGC 2936.Although this plot does not have the spatial resolution to show lo-calized variations of the quantities, it can give us an idea of trendsand variations seen in our data. From this plot, we can see thatNGC 2936 is actively forming stars in its eastern side. Region S1Bdisplays the highest SFR, which corresponds to 17.9 M (cid:12) yr − . Thissource lies over a dust lane, which produces a high extinction at thislocation (see section 3.2.3). The second most intense star formationburst corresponds to S1A, which is located at the tip of the eastern MNRAS000 , 000–000 (0000) lues on Arp 142 Figure 3.
Top panel: Typical observed spectra in NGC 2936. The examples correspond to the region namely S1A and S2C (black continuous lines). Mainnebular emission-lines are labeled. Bottom panel: UCG 05130 NOTES01 spectrum including the detected emission-lines. tidal structure in NGC 2936. This object has a SFR of 3.2 M (cid:12) yr − .Both values are extremely high when compared with typical SFRsderived for spiral galaxies (see Mu˜noz-Mateos et al. 2007). Indeed,the star formation rate densities of these sources correspond tolog( Σ S FR , cor ) = (cid:12) yr − kpc − and log( Σ S FR , cor ) = (cid:12) yr − kpc − , which reflects that both objects correspond to a localizedstarburst (Efremov et al. 2004). In the case of the first source, its Σ S FR , cor is consistent with the value shown by star-forming regionslocated in LIRG galaxies (e.g. Piqueras L´opez et al. 2016). How-ever, these estimates could be overestimated. In this analysis wehave used the Calzetti et al. (2000) extinction law, which was de-fined for starburst objects. In order to perform some comparisons,we have dust-corrected the spectrum of region S1B by using theFitzpatrick (1999) extinction law. In this case, we have derived aSFR = (cid:12) yr − for this source, which is much lower than thevalue derived from the spectrum corrected by using the Calzettiet al. (2000) extinction law. This comparison shows how sensitiveare our results depending on the use of a specific extinction law.In any scenario, these results may suggest that the eastern tidalstructure / arm in NGC 2936 is experiencing a burst of star forma- tion. Despite that the current data set does not allow us to concludethat the burst was triggered by gas compression during the inter-action in Arp 142, there are evidence that suggests that interactinggalaxies display strong burst of star-formation. For instance, Fer-reiro et al. (2008) found bright H α knots in a sample of interactinggalaxies. Ellison et al. (2008) found that galaxy pairs (at separa-tion lower than 30-40 h −
70 kpc) display an enhancement in theirSFRs. Therefore, based in our results, we can speculate that thestrong star formation bursts that we are witnessing in NGC 2936are the result of the gravitational encounter between the main mem-bers of Arp 142. This encounter have produced an enhancement inthe SFR of this galaxy pair, which is mainly focused in S1B. Ac-tually, when we take into account global values, we obtain a totalSFR of 35.9 M (cid:12) yr − for NGC 2916 (which could be ∼ (cid:12) yr − if we scale the value for the Fitzpatrick (1999) extinction law).This value decreases 7.35 M (cid:12) yr − , if we consider the fluxes thathave not been corrected by internal extinction. Both values reflecta prominent SFR in this galaxy pair. In addition, regions S3E, S3Fand S4A display high values for the star formation rate, where allthese sources are located in a spiral structure. All the remaining MNRAS , 000–000 (0000)
M. D. Mora et al.
Table 2.
Emission line fluxes for regions observed in NGC 2936
ID H γ H β [O iii ] λ iii ] λ i λ i ] λ ii ] λ α [N ii ] λ i λ ii ] λ ii ] λ − erg cm − s − S1A 42.37 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± † ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± † ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± † Same star-forming region observed in two di ff erent slits (S3 and S4). sources have considerably low SFRs with respect to the former ob-jects. Recent studies on the distribution of chemical abundances in in-teracting galaxies have provided important clues to understand themetal / gas mixing process in these systems (Kewley et al. 2006;Rupke et al. 2010; Barrera-Ballesteros et al. 2015). Most of theseworks rely on the oxygen abundance to study the chemical con-tent of galaxies. However, due to the di ffi culty to observe weakauroral lines sensitive to temperature (such as [O iii ] λ N method (Pettini &Pagel 2004). This is one of the most used methods involving strongemission-lines such as the H β , [O iii ] λ α , and [N ii ] λ N calibra-tion proposed by Marino et al. (2013). The typical dispersion onthis calibration corresponds to 0.18 dex. Therefore, uncertaintiesin the oxygen abundances are the quadratic sum of the uncertain-ties in the flux and the dispersion of the calibrator. Finally, de-rived oxygen abundances, listed in Table 3, show values between8.44-8.60 ± + log(O / H) (cid:12) = ± ± ± ± + log(O / H) ∼ By using the H α / H β Balmer ratio we obtained the internal extinc-tion for all the sources. The spatial distribution of these values isshown in the bottom left panel of Figure 4. Three knots (S1B, S2Cand, S4D) show high extinction, with values ranging from E(B-V) = = e ) represent the low-excitation zoneof the ionized gas, and these were derived from the parameterR S , which is defined as the ratio between the [S ii ] λ / temden routine of the nebular package in STSDAS /iraf , assuming an electron temperature ofT e = K. The calculated N e values are given in Table 3, fromthem, we can see that in the eastern region of NGC 2936 the esti-mated values of electron densities are in the range of N e = − . In the northern and southern region of this galaxy, we foundvalues ranging between N e = − and N e = − respectively. In six regions (dashed lines in Table 3) we could notestimate electron densities because the sulphur emission-lines havelow S / N. The eastern regions present the lowest densities, with amean electron density of N e =
300 cm − , below the critical value forcollisional de-excitation (500 cm − ). North-Western regions showthe highest values of electron densities with a mean electron den-sity of N e =
600 cm − and the highest value was estimated in regionS4A, where N e =
700 cm − . MNRAS000
700 cm − . MNRAS000 , 000–000 (0000) lues on Arp 142 Figure 4.
Color-coded spatial distribution of the H ii regions on top of the NGC 2936 HST image. On the top left panel, we present the SFRs. On the Topright panel, we present the oxygen abundances. On the bottom left, the extinction, and on the bottom right the H ii velocity measurements. Circle sizes do notrepresent physical quantities and were expanded for best representation in the figure. Finally, in a global view, our estimates of electron densities,which range from 10 cm − to 700 cm − , are consistent with the val-ues estimated by Krabbe et al. (2014) for a sample of local interact-ing galaxies. These authors found values ranging from 24 cm − < Ne < − , which are larger than the values obtained for a sam-ple of isolated galaxies (40 cm − < Ne < − ). In this sense, thestar-forming regions located in NGC 2936 do not display the typi-cal electron densities found in non-interacting systems. The originof this trend is not clear. For example, Krabbe et al. (2014) suggestthat the high electron densities found in H ii regions located in inter-acting galaxies do not seem to be produced by gas shocks. Howeverthere is evidence from (Villar Mart´ın et al. 2014; Arribas et al. 2014,e.g.) that does not allow us to exclude that high values detected nearthe nucleus are due to outflows rather than shock-excited. Deeperstudies on this topic are needed. In the bottom right panel of Figure 4 we show the radial velocitydistribution (listed on Table 1) of the star-forming regions observedin NGC 2836. Radial velocities were derived by using the task em - sao in iraf . The discrete data available does not allow us to studythe whole kinematics of this system. It is clear from the figure thatthe southern tidal structure of NGC 2936 displays lower radial ve-locities with respect to the main disk of this galaxy. By observingthe H α emissions in 2D spectra of the eastern region of NGC 2936( the most intense spots of the left upper panel in Figure 2), wecan see that the kinematics of the region S1A appears to be decou-pled with respect to the kinematics of regions S1B and S1C with adi ff erence in velocity of ∆ v r ∼ − , which may be due to adistorted tidal arm in NGC 2936. Integral field spectroscopy couldbe useful to disentangle the kinematics of region S1A. MNRAS , 000–000 (0000)
M. D. Mora et al. α velocity dispersion Despite the fact that the spectral resolution of our data is low(FWHM ∼
207 km s − ), estimating the velocity dispersion from themost prominent emission line, namely H α , will give us a broaderview of the internal kinematics of each region. Therefore, observedvelocities dispersion ( σ H α ) were corrected for the instrumental ve-locity dispersion σ I and for thermal velocity dispersion σ T to ob-tain the intrinsic velocity dispersion σ int ,using the following equa-tion: σ int = (cid:112) σ H α − σ I − σ T . This exercise allow us to obtainthe corrected velocities dispersion ( σ (H α ) int ). Results are listed inthe last column of Table 3. However, we note that our estimatesshould be read as referential values, considering the medium spec-tral resolution of our observations. Future high spectral resolutionstudies are needed to fully disentangle the internal kinematics ofthe star-forming regions in Arp 142. Most of the sources for which we have spectroscopic informationresemble star-forming objects. However, HII regions can be ei-ther photo-ionized or shock-ionized (e.g. Calzetti et al. 2004; Honget al. 2011, 2013; Rich et al. 2015; Hernadez-Jimenez prep). Inorder to determine and confirm the ionization mechanism of thedi ff erent regions of NGC 2936 we have plotted their [N ii ] / H α ver-sus [O iii ] / H β line ratio in a diagnostic Baldwin-Phillips-Terlevich(BPT) diagram (Baldwin et al. 1981), which is shown in the toppanel of Figure 5. On the BPT diagnostic diagram, we have in-cluded a grid of photoionization models calculated by Kewleyet al. (2001). These models are based on stellar population syn-thesis models STARBURST99 (Leitherer et al. 1999) in conjunc-tion with a photoionization code Mappings III (Binette et al. 1985;Sutherland & Dopita 1993). STARBUST99 generates the spectralenergy distribution of a stellar cluster that serves as an ionizationsource that is used as input for MAPPING III calculating the ion-ization states and the emission line fluxes of the elements of thegas (for further details see Kewley et al. 2001). The photoion-ization model grid (in gray color) shown in the BPT of the Fig-ure 5 diagram has two metallicities (plotted as continuous lines):12 + log(O / H) = + log(O / H) = . × up to 4 . × cm s − (dashedlines). These models are calculated assuming an instantaneous star-burst and a gas density of 10 cm − . As a reference, we also plot-ted the photoionized maximum limits for H ii proposed by Kau ff -mann et al. (2003) (empirically) and Kewley et al. (2001) (theo-retical), which are shown in grey and black lines respectively. Inaddition to the photo-ionization models, we plotted a grid of shock-ionization models from Allen et al. (2008). These authors use theMappings III shock and photo-ionization code to calculate the ion-ization states and the line emission fluxes of the gas. The ioniz-ing radiation field input is produced by the cooling zone behindthe front shock (post-shock region) and it is mostly composed ofthermal bremsstrahlung (free-free) continuum. The main parame-ters of the shock models are: the shock velocity, magnetic field (B),pre-shock gas density, and metallicity. The shock emission of themodels can be divided into two components: the radiative shock(post-shock region) and its photo-ionized precursor (pre-shock re-gion). In the BPT diagrams, we plot a grid with only the radiativeshock for two di ff erent metallicities: 12 + log(O / H) = + log(O / H) = − . The shock velocity (continuous lines) range is 100 −
700 km s − and 200 − − , respectively. The ISMmagnetic field (dashed lines) range is 1 × − − µ G. Becauseof the huge extinction of the nuclear and plume regions, H β line isnot present on the aperture spectra; hence its corresponding pointsare not on BPT diagram. In order to overcome this, we used theWHAN diagram Cid Fernandes et al. (2011): the H α equivalentwidth (EW H α ) vs [N ii ] / H α . In this plot all sources are present (Top-left panel Figure 5). The dash vertical line at [N ii ] / H α = − . HST image of Arp 142, where the circles represent the location ofthe di ff erent spectroscopic sources.Inspecting the top-left panel of Figure 5 we found that all re-gions lying on the spiral arms (dark blue and black squares) fallon the quadrant corresponding to the lower ionizing parameter ofthe photo-ionization models (considering the uncertainties). The re-gions associated to the plume (S4D, S4E, S4F, S4G, and S1E))are located inside a “mixing” zone; they fall near or inside thequadrant of the lower shock velocity (i.e. 200 −
300 km s − ) of theshock-ionization model with a metallicity of 12 + log(O / H) = = = + log(O / H) = β , falling into the AGN sector in the WHAN diagram (top-right panel of Figure 5) indicate that they are also being ionizedby shocks. Moreover in the regions located at the plume, the tidale ff ect is more pronounced (due to they are at the outskirt of thegalaxy), and therefore, a shock excitation is more likely. Finally,from the kinematic point of view, their velocity dispersion mea-sured are super-sonic (see Table 3), therefore, reinforcing the aboveshock scenario. The locus of regions close to the nucleus (S2C andS1D) in WHAN diagram lie on the quadrant of AGN activity, sug-gesting that the ionization of the gas is dominated by nuclear activ-ity and / or shocks.This scenario is consistent with the SDSS nuclearclassification of NGC 2936, which is cataloged as an AGN galaxy.
In Figure 6 we show a color image of Arp 142, which is a composi-tion of NUV / GALEX (blue), F W / HST (green) and 8 µ m / Spitzer images. It is clear from this image that the very strong dust laneseen on the
HST image is evidenced by
Spitzer at 8 µ m.The bluecolor maps the young star-forming regions, and it is in agreementwith the top left panel of the Figure 4, while the magenta color mapsthe embedded / newly formed star-forming regions. In order to quan-tify the star formation rate on this system by using archival data, wehave used NUV / GALEX and
Spitzer already published in the litera-ture. In the case of NUV / GALEX , we have used the magnitude pub-lished by Gil de Paz et al. (2007) and the equation (3) in Iglesias-P´aramo et al. (2006). This exercise provides a SFR
NUV = (cid:12) yr − once the NUV emission has been corrected by dust (equation1 from that paper). If we do not consider the dust correction, thestar formation rate drop to SFR NUV = (cid:12) yr − (using equation3 in Iglesias-P´aramo et al. (2006)). In the case of the IR informa-tion, we can use it to trace the star formation rate associated withdust heated by energetic photons. By using the total IR emission inNGC 2936 (listed in Table 3 of Xu et al. (2010)) and the equation 5
MNRAS000
MNRAS000 , 000–000 (0000) lues on Arp 142 −1.2 −1.0 −0.8 −0.6 −0.4 −0.2 0.0 0.2 0.4 log (N[II]/Hα) −1.0−0.50.00.51.0 l o g ( O [ III ] / H β ) −1.0 −0.8 −0.6 −0.4 −0.2 0.0 0.2 0.4 0.6 log (N[II]/Hα) −1 E W H α ( Å ) S1D S1E S2C S2D S4B1 S4B2 S4C S4D S4E S4F S4G
Star-formingRetired wAGN/LINER AGN/Seyfert
Figure 5.
Diagnostic diagrams to identify the ionization mechanism for the regions observed in NGC 2936. Top-left: BPT diagram. Continuous gray linerepresents the limits suggested by Kau ff mann et al. (2003) while black continuous line represents the limits from Kewley et al. (2001). Grey grid correspondsto photo-ionization model, while red grid correspond to shock-ionization models with a metallicity of 12 + log(O / H) = ii ] over H α abundance distribution used for color coding on the bottom panel. The arrows near B and Vletters indicate the direction in which the magnetic field and shock velocity increment, respectively, in the models. Top-right: WHAN diagram. The circlessymbols come from nuclear and plume regions and are not present in BPT diagram. The color-coded is the same as BPT diagram. The dashed vertical lineat [N ii ] / H α = − . H α = H α = . ii ] over H α abundances. Figure clearly illustrated that both the nuclear and plume regions arebeing ionizing by AGN / Shock activities. of Iglesias-P´aramo et al. (2006), we derived a SFR totalIR = (cid:12) yr − . This value is slightly lower if we consider the corrected IRluminosity displayed in Table 3 of Xu et al. (2010). Finally, H α emission lines from the star-forming regions were also used to getstar formation rates and added up to give a number for the wholesystem. The latter is, of course, a lower limit for the total star for-mation rate of the system. The three star formation rate estimationsobtained from NUV / GALEX (corrected by dust),
Spitzer (total IRemission published by Xu et al. (2010)) and H α (once correctedby internal extinction) display di ff erent results (3.8 M (cid:12) yr − with GALEX , 12.6 M (cid:12) yr − with Spitzer and 35.9 M (cid:12) yr − for H α ). In this context, Iglesias-P´aramo et al. (2006) found that the star for-mation rates traced by IR emission can exceed the values derivedfrom the UV emission (by a factor 2), where this discrepancy couldarise from the dust attenuation correction for dusty galaxies. In thecase of the SFRs derived from H α , we note that dust correctionplays an important role. As previously mentioned, the Fitzpatrick(1999) extinction law allowed us to derive a lower SFR for regionS1B (8.7 M (cid:12) yr − versus 17.9 M (cid:12) yr − once Calzetti extinction lawwas used). If we scale this di ff erence to the total SFR in NGC 2936(derived from H α ), we found a total value of 17.5 M (cid:12) yr − , which isnot so di ff erent with respect to the SFR derived from IR information MNRAS , 000–000 (0000) M. D. Mora et al.
Table 3.
Oxygen abundances and star formation rates (corrected and uncorrected from internal extinction).
ID n([S ii ]) L H α uncor E(B-V) L H α cor 12 + log(O / H) O N SFR H α uncor SFR H α cor Area log Σ SFR , cor FWHM(H α ) σ (H α ) int cm − × erg s − mag × erg s − M (cid:12) yr − M (cid:12) yr − kpc M (cid:12) yr − kpc − Å kms − S1A 70 ±
10 163.0 ± ± < ± ± ± ± ± ± ± < ±
50 205.0 ± ± ± ± ± ± ± ± ± ± ±
70 44.4 ± ± ± ± ± ± ± ± ± ◦ - - - 7.6 ± ± ± ± ± ± ± ± § ± ± ± ± ± ± ± ± ± ± § ± ± ± ± ± ± ± ± ± ± § ± ± < ± ± ± ± ± ± ± ± ◦ - - - 18.3 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± § ± ± ± ± ± ± ± ± ± ± § ± ± ± ± ± ± ± ± ± ± § ± ± ± ± ± ± ± ± ± ± § ± ± < ± ± ± ± ± ± ± ± † ±
90 89.9 ± ± ± ± ± ± ± ± ± ±
50 55.7 ± ± ± ± ± ± ± ± ± † § ± ± ± ± ± ± ± ± ± ± ◦ - - - 2.4 ± ± ± < ± < ± ± ± ◦ - - - 1.2 ± ± ± < ± < ± ± ± ± ± ± ± < ± < ± ± ± ± § ± ± ± ± ± ± ± ± ± § ± ± ± ± ± < ± ± ± ± ± ±
50 24.5 ± ± ± ± ± ± ± ± ± ± § ± ± ± ± ± < ± < ± ± ± ± † Same star-forming region observed in two di ff erent slits (S3 and S4). ◦ Star-forming region uncorrected by extinction. § Due to the very large uncertainties ( ≥ σ (H α ) int : Intrinsic velocity dispersion corrected by instrumental and thermal dispersion. (SFR totalIR = (cid:12) yr − ). In this sense, a precise determination ofthe dust correction is mandatory to compare the di ff erent SFR trac-ers. In any case, the values that we obtain for the spiral memberin Arp 142 show an actively star-forming object, where the interac-tion with NGC 2937 should be playing an important role. Finally,we note that Xu et al. (2010) has also obtained star formation ratesfrom Spitzer , and we (not surprisingly) get the same result, giventhat we also used the same data.
The elliptical galaxy NGC 2937 is only 0.84 arcmin or 24.4 kpcprojected distance from NGC 2936. Therefore it is the natural can-didate responsible for the distorted companion spiral galaxy. How-ever, there is also another possibility, which is a small galaxy northof the system, UGC 05138 NOTES01, at a projected distance of 1.3arcmin north of the center of NGC 2936. We have placed one of theslits on this galaxy and its spectrum is then shown in Figure. 3,bottom panel. From the several emission-lines, we could obtain aradial velocity for this galaxy of 4971 ±
18 km s − , which is ∼ − di ff erent from the Arp 142 system. Therefore we concludethat this galaxy cannot be the disturbing agent that caused the mor-phological and kinematic changes of NGC 2936. Low surface brightness merger indicators (i.e. faint tidal tails,bridges, etc) are often missed on images due to the lack of su ffi cientsensitivity or short exposure times. In particular, when we were se-lecting the interesting regions to be pointed by GMOS, we notedthat there is a faint tidal arm that appears to originate in the centerof the elliptical galaxy NGC 2937 that runs in the direction of thespiral galaxy NGC 2936. This can be seen by eye but it is hard todisplay it in a figure, given that it is a very faint feature. In order Figure 6.
RGB Color composite figure of Arp 142. Blue color correspondsto the NUV from Galex, Green to the HST F W and Red to Spitzer 8microns. to evidence this feature and to eventually highlight other faint andextended features in the image, we mapped the background (i.e. thefaint and extended component) using a Thin Plate Spline Interpola-tion (TPS). Briefly, TPS is a 2D interpolation for arbitrarily spaceddata forming a grid. The TPS analogy is to consider a thin metal MNRAS000
RGB Color composite figure of Arp 142. Blue color correspondsto the NUV from Galex, Green to the HST F W and Red to Spitzer 8microns. to evidence this feature and to eventually highlight other faint andextended features in the image, we mapped the background (i.e. thefaint and extended component) using a Thin Plate Spline Interpola-tion (TPS). Briefly, TPS is a 2D interpolation for arbitrarily spaceddata forming a grid. The TPS analogy is to consider a thin metal MNRAS000 , 000–000 (0000) lues on Arp 142 Figure 7.
Tidal feature derived from HST image. From Right to left Eachstamps corresponds to 36” x 32” Top stamps: From left to right: F W , F W , F N and, F W . Bottom: TPS modeled images. From left toright: F W , F W , F N and, F W sheet that is shaped by a grid, forcing the sheet to not move at thepoints that are modeled on the grid. By applying this technique, thefeature we had noted by eye could then be revealed (see the bottomrow of Figure 7). Since this technique is computationally expen-sive, we focused on the region in between the galaxies. We appliedthe TPS considering a grid separated by 20 x 20 pixels, which isa size large enough to ignore the stars but small enough to mapextended sources. The output is a new image with the backgroundmodeled and that contains the extended sources of the image. In theupper row of Figure 7 the original HST images are shown and re-sults from the TPS are shown in the bottom row of Figure 7 wherethe tidal bridge is clearly seen on the bluest HST filter ( F W )and it is less clearly seen in the red one F W . The narrow-band F N frame does not clearly show the tidal bridge, probably dueto the shallower image, compared with those for the other filters. Unlike to the spiral galaxies, which are “cold” systems, the ellip-ticals response to tidal fields in a subtle fashion because they are“hot” systems. The most striking tidal structures observed in ellip-ticals are the shells and ripples, however, they are relics / debris ofa past interaction rather than tidal structures of an ongoing event(Barnes & Hernquist 1992). The latter events can produce e ff ectssuch as enhancement in surface brightness profile at outer radii(Kormendy 1977), non-centric inner isophotes (e.g. Lauer 1986,1988; Gonz´alez-Serrano & Carballo 2000), and “u-shape” velocityprofiles (e.g. Borne & Hoessel 1988). The first two of these fea-tures we are able to probe by performing a photometry analysis on F W (SDSS’s g -filter) HST image of NGC 2937. The magnitudezero point (ZP) of the image was obtained from the ACS webpagecorresponding to the date of the observation, for ZP ( F W ) = iraf task ellipse . The geometrical parameters; position angle, ellipticity,and center of the ellipses were let free. To avoid the substructuresor bad pixels the clipping algorithm was activated with an allowedmaximum fraction of 20% of flagged pixels. Top panel of Figure 8shows some selected fitted ellipses on outer isophotes. It can beclearly seen that the centers of these outermost isophotes are o ff settoward NGC 2936 at the same direction as the faint tidal featurefound out in Figure 7 (see Section 3.4). One way to quantify thisdisplacement is by the δ/ a ratio (Lauer 1988), where δ is the sep-aration of isophote center with respect to the nucleus and a is thelength of Semi-Major Axis (SMA). The radial profile of this ratio is shown in the middle panel of Figure 8. The δ/ a ratio is < (cid:48)(cid:48) ( ∼ δ = ∼ − ff -centeringisophotes are the signature of m = µ Re = . ± .
04 mag / arcsec and Re = ± .
16 arcsec (3 . (cid:48)(cid:48) ( ∼ . µ = / arcsec of deviation from thede Vaucouleurs law at the largest SMA. These bright envelopes areoften seen in ellipticals with nearby companions (as it is the casefor NGC 2936), whereas in isolated galaxies these envelopes areabsent (Kormendy 1977). Moreover, this e ff ect is observed whenthe dynamical time of the encounter is much shorter than that ofoutermost parts of the galaxy, “feeling” a tidal impulse force ratherthan di ff erential force (Aguilar & White 1985; Binney & Tremaine1987). The above tidal characteristics found out in the ellipticalgalaxy of pair Arp 142 shall be explained in detail in 4.3. In order to build a dynamical scenario for the NGC 2936 / ff erent orbital param-eters (initial orbital energy, angular momentum, direct or progradeorbit, and disc orientation with respect to orbital plane), was vi-sualized using the snapshot preview at di ff erent ages (in steps of50 Myr, as it is provided by the database), changing its projectionand, searching for the view that best resemble the morphology ofArp 142. Finally the orbit that better matches the observed mor-phology of Arp 142 is a prograde orbit with an initial velocity ( v ini )of 300 km s − , a perigalacticum ( r peri ) of 8 kpc with the disk of thespiral galaxy to be perpendicular to the orbit plane. From the sim-ulation, we conclude that the current stage of the system wouldbe 50 ±
25 Myr after the first pericenter passage. The uncertaintycomes from the numerical simulation step of 50 Myr. In Figure 9 atthe top panel, we show a snapshot of the actual stage and at the bot-tom a snapshot of 50 Myr before, at pericenter passage. We can seethat the first snapshot reproduce qualitatively very well the mainmorphologic features of the studied system (c.f. Figure 1). Thelarge plume at North-West direction and the faint tidal bridge to-ward North direction (see Section 3.4). The perpendicular positionof spiral disk with respect to orbital plane explains why the tidalarms are bending toward the elliptical galaxy producing these fea-tures that give the famous ring-like form of the Penguin system. Byobserving the pericenter snapshot, we can conclude that these tidalfeatures are “new”, and therefore a maximum age of 50 ±
25 Myr.One could also attempt to compare the kinematics of the sys-tem in pseudo slits similar to those used for the real observations.Figure 10 shows the output of the GalMer simulation for the ve-locity field of the system. We obtained the radial velocity profiles
MNRAS , 000–000 (0000) M. D. Mora et al.
Figure 8.
Top panel: Selected fitted ellipses (shown in blue) that overlaidthe isophote contours (shown in red). The white circle markers correspondto the ellipse centers, while the red one is the photometric center of thegalaxy. Middle panel: The δ / a ratio profile. Bottom panel: F W bandsurface bright profile (black dot-points) fitted with de Vaucouleurs law (redsolid line). The small box below the figure corresponds to the residuals inmag / arcsec . The blue points in middle and bottom panels correspond tothe selected isophotes showed in top panel. along slits in similar positions to those used in the real observationsand these are plotted in Figure 11, together with the correspond-ing observed velocities. Both velocity profiles show similar shapeswhen the model and the observations are compared. We stress thatthis is just an exercise to show that it is possible that the ellipticalgalaxy NGC 2937 is, in fact, the perturbed of the NGC 2936 spiralgalaxy. However, given the nature of the problem we cannot assurethat this is the only configuration that will describe the interactingscenario for this system.Since GalMer simulations also provides metallicity evolutionof the system, we can further extend this exercise to observe its ex-pected distribution. Since the metallicity estimations from the ob-servations are discrete (i.e. few selected regions), we have producedin Fig. 12 a schematic comparison between the observations and themodel. We stress that this is a rough comparison since GalMer re-sults were normalized in order to match the size of the backgroundimage. Considering these caveats the trend predicted by the modelshow that lowest metallicites are seen on the plunge while the high-est metallicities concentrate in the galaxy nucleus. Due to the un-certainties in our data, it is not possible to conclude if there is ametallicity gradient as the model shows, however the observed val-ues seems to reproduce the relative values according to the positionon the galaxy, i.e., high values close to the nucleus, low values inthe plunge. + E pair of galaxies
Arp 142 is an excellent example of the morphological transforma-tion produced by a gravitational encounter. The southern-east sideof the galaxy NGC 2936 displays an extended structure, which wasmost probably produced by the tidal field of the elliptical galaxyNGC 2937. In this sense, the spiral structure of NGC 2936 seemsto be clearly modified due to the interaction with its companion. Ingeneral, galaxy transformation in pairs of galaxies has been mainlystudied in S + S pairs. However, there are few studies focused onthe properties of S + E pairs (Kojima & Noguchi 1997; Xu et al.2010; J¨utte et al. 2010; Hern´andez-Ibarra et al. 2016), despite theirimportance in the mass assembly scenario. For instance, Xu et al.(2010) studied a sample of 12 elliptical-spiral pairs. On average,these authors found no enhancement in the sSFR of these pairs.This is not the case for the intriguing system Arp 142. Consider-ing their values, (see Table 3) the spiral member of Arp 142 has alog sS FR of ∼ -10.51, implying a SFR of 8.9 M (cid:12) yr − (SFR deter-mined only from IR luminosities). This value is even larger if wetake into account the total IR luminosity listed by these authors (asdescribed in section 3.2.7). This indicates that NGC 2936 has anenhanced sSFR. Moreover, by using the H α luminosity, we founda total SFR of 35.9 M (cid:12) yr − (Calzetti et al. 2000), or 17.5M (cid:12) yr − (Fitzpatrick 1999). Taking into account both values we find thatthey are larger than typical SFRs for spiral galaxies (e.g. 1.17 M (cid:12) yr − from Xu et al. (2010), 2.6 M (cid:12) yr − from Darg et al. 2010). Inthis sense, what is the reason for this enhancement in SFR, con-sidering that this enhancement is not the typical case for spirals inS + E pairs, as shown by Xu et al. (2010), and more recently by Caoet al. (2016). Indeed, in the case of S + E pairs, Hwang et al. (2011)suggest that the X-ray emission associated with an elliptical galaxycould suppress the SFR in its companion spiral, but this seems tonot be the case for NGC 2936, probably due to the mass of the el-liptical companion, which is not high enough to suppress the star
MNRAS000
MNRAS000 , 000–000 (0000) lues on Arp 142 Figure 9.
Snapshot of total mass (star + gas) for the simulated system. Top-panel: actual stage 50Myr after pericenter passage. Bottom-panel: pericen-ter passage. The angles used for visualization were phi -28, theta 37 (formore details see Di Matteo et al. 2008). formation in NGC 2936. Actually, the interaction process betweenNGC 2936 and NGC 2937 could had produced gas flows and gascompression which may be the reasons of the star burst event in theeastern side of NGC 2936. In this context, previous authors havefound that close galaxy pairs display an enhancement in their SFRs(Ellison et al. (2008)). Despite this is not the typical case for S + Epairs, the intriguing system Arp 142 seems to be an exception. New3D spectroscopy (in the optical and submm) can be very useful inorder to disentangle the mechanisms that favour the star formationactivity in this system. ff ects of the interactions on NGC 2936 The interaction that is taking place in Arp 142 is strongly a ff ectingand transforming the spiral NGC 2936. Moreover, NGC 2936 doesnot display the typical metallicity distribution expected for a spiralgalaxy of its mass. Indeed, abundances show no gradient (withinthe error) and we interpreted it as a result of the re-distribution Figure 10.
Velocities map for the simulated system. See text for details.
Figure 11.
Radial velocity profiles for the Arp 142 system - GalMer simula-tions and observations. The red (blue) profiles are the observed (simulated)radial velocities along the four slits. On each panel lower x-axis correspondsto the GMOS pixels while upper x axis corresponds to the simulated pixelsilts from Fig. 10. y axis corresponds to the relative velocities for simula-tions and observations. NGC 2963 velocity curves for the observations werederived by cross-correlating one spectrum (i.e. one row) of the extended 2Dspectra against all the remaining ones of the same slit. of the gas in the galaxy disk due to the inflow from the outskirtto the center of the galaxy. This gas mixing process has been ob-served in interacting galaxies (e.g. Kewley et al. 2006; Rupke et al.2010; Torres-Flores et al. 2014; Rosa et al. 2014; Olave-Rojas et al.2015) Numerical simulation predicted the flattening of the metal-licity gradient to occurs just after first pericenter passage (Rupkeet al. 2010; Perez et al. 2011) hence the age of interaction stage,about 50 ±
25 Myr after pericenter, that we estimated (Sec 3.6)is a good agreement with this scenario. The presence of the AGN
MNRAS , 000–000 (0000) M. D. Mora et al.
Figure 12.
GalMer color-coded spatial distribution of the abundances over-laid and scaled with abundances showed on 4. Big filled dots correspond tothe measured regions while small dots are the output from GalMer simula-tions. For comparison between model and observations colors bar use thesame cmpa, but di ff erent scale. Right bar corresponds to dots, while left barcorresponds to Galmer simulation, same units as right bar.. and the evidence of the central region being ionized in NGC 2936hinted by the BPT diagram agrees with Ellison et al. (2011), andthus with the higher fraction seen in close pairs of composite galax-ies (i. e. galaxies whose central region is ionized by nuclear activityand massive stars) than in isolated systems. A not surprising resultsince a large fraction of AGNs have been found in close pairs com-pared with isolated galaxies (Satyapal et al. 2014). However, ourresults were obtained by using long-slit technique, which preventsus a spatially detailed analysis of the ionization in the center ofthe galaxy. Future IFU observations will elucidate the origin of theionized nuclear activity in NGC 2936. ff ects of the interaction on NGC 2937 The analysis of NGC 2937 (section 3.5) shows that isophotes ap-pear to have a common center until 4 kpc ( ∼ towards theSouth. On the other hand, at large radii (from ∼ t inner = r /σ , islesser or larger than the encounter dynamical time, t encounter = b / v (where r is the radius of inner part, σ is star velocity dispersion, b the parameter of impact, and v is the relative velocity of approxima-tion). When t encounter ≥ t inner the particles inside r have enough timeto react to the perturbation, and therefore the tidal approximationapplies, the inner region is elongated toward the radial direction ofcenter-of-masses, and (as a whole) is pushed away on the same di-rection, presenting an o ff set with respect to the outer envelope (c.f.see panels (b) and (c) of Figure. 1 in Aguilar & White 1986). Hence the inner part of NGC 2937 moved in the North-South direction to-ward closest passage (see bottom-panel in Figure. 9). The innerregion returns to its normal position because of dynamical friction.This displacement lasts around 1 − × yr (Combes et al. 1995).On the other hand, at the outer regions the t encounter ≤ t inner , thereforethe impulsive approximation applies: the particles barely changeits potential energy, thus the energy of the encounter is transferredin the form of kinematic energy to outer particles “heating” them(Aguilar & White 1986), yielding into an expansion of the outerregions and producing an enhancement of the surface brightnessprofile at these radii, as it is seen in Figure. 2 of Aguilar & White(1986). The tidal distension lasts around 0 . × yr after pericenterpassage.One could test if these tidal approximations are valid for theelliptical galaxy of the Arp 142 system by calculating t inner and t encounter . In order to do so, we consider the σ =
273 km s − ob-tained from data release 12 of Sloan Digital Sky Survey (SDSS;York et al. 2000). σ was calculated following the method de-scribed in Bolton et al. (2012). So, by taking r as the radius un-til we observed the o ff -centric isophotes and where begin the en-hancement of surface bright profile, around 4 kpc, we obtain a t inner (cid:39) × /
273 km s − (cid:39) . × yr. An approximation of t encounter can be obtained from numerical simulation (Section 3.6),then t encounter = /
300 km s − = . × yr. The values for t inner and t encounter are quite similar, which means that the both tidalapproximations explained very well the tidal features observed inNGC 2937. Because in this work we are presenting the properties of the dis-torted spiral NGC 2936, triggered by the interaction with the el-liptical NGC 2937, a natural step is to further explore the inter-acting system Arp 142 since spiral-elliptical (gas / dust rich) inter-actions are one of the avenues on which elliptical galaxies mayaccrete dust, yielding to an elliptical (or early type) galaxy withdust lines, with a patchy or filamentary dust distribution (Goud-frooij et al. 1994) as it is the historical case of NGC 5128 (Baade& Minkowski 1954). Moreover, Kaviraj et al. (2012) showed that ∼
65% of ellipticals (early-type galaxies) that contain dust presentdistorted morphologies, concluding that dust is likely to be the re-sult of the recent mergers. They also found that these systems arelocated in low-density environment with 80% of them inhabiting inthe field. In the context of these results, the isolated pair Arp 142seems to be a good candidate for a future dusty early-type galaxy.GalMer simulations predict that the system will merge at an ageof ∼ We have analyzed GMOS spectroscopy of selected star-forming re-gions of NGC 2936, member of the interacting pair Arp 142. Ourresults show an enhanced star formation rate, probably triggered bythe interaction with NGC 2937. Star-forming regions in NGC 2936display oxygen abundances consistent with the solar value and elec-tron densities larger than the typical values found in non-interacting
MNRAS000
MNRAS000 , 000–000 (0000) lues on Arp 142 systems. In addition, our extinction measurements are in agree-ment with the position of the H ii regions on the optical images,like the regions near the dust line seen in NGC 2936. Finally, wealso find evidence suggesting that the nucleus of NGC 2936 is be-ing ionized by AGN activity. Regarding the elliptical member of thepair, isophotal analysis unveils the tidal e ff ects on NGC 2937 show-ing non-concentric isophotes at inner radii, besides that at largeradii the surface brightness profile deviates from de Vaucouleurslaw. Using GalMer simulations we are able to reproduce the ob-served radial velocity profiles and observed galaxy morphologiesonly considering the interaction of an elliptical and a spiral galaxy,discarding, as a first approach a third member on the interaction.The current stage of the system would be about 50 ±
25 Myr afterthe first pericenter passage. The North-West plume of NGC 2936and the North faint tidal bridge between galaxy would be the re-sult of that initial configuration of the system: the perpendicularorientation of the spiral disk with respect to the orbital plane. Themaximum age estimated for these tidal structures is 50 ±
25 Myr.We have proved that the third galaxy UCG 05130 NOTES01 havea di ff erence in radial velocity large enough to do not be part of theinteraction of the Arp 142. To conclude, we believe that Arp 142 isa candidate to become a dusty spiral at the final stage of the merger. ACKNOWLEDGEMENTS
Authors thank the referee, Javier Piqueras L´opez, for his use-ful comments and suggestion that greatly improved this paper.M. D. Mora acknowledges CONICYT, Programa de astronom´ıa,Fondo GEMINI-CONICYT: Este trabajo cont´o con el apoyo deCONICYT, Programa de Astronom´ıa, Cargo de Astr´onomo de So-porte GEMINI-CONICYT 2018. S. Torres-Flores acknowledgesthe financial support of Direcci´on de Investigaci´on y Desar-rollo de la ULS, through a project DIDULS Regular, under con-tract PR16143. V. Firpo acknowledges support from CONICYTAstronomy Program-2015 Research Fellow GEMINI-CONICYT(32RF0002). F. Urrutia-Viscarra acknowledges the financial sup-port of the Chilean agency Conicyt + PAI / Concurso nacional apoyoal retorno de investigadores / as desde el extranjero, convocatoria2014, under de contract 82140065. CMdO acknowledge fund-ing from FAPESP (program 2009 / / / ESA Hubble Space Tele-scope, obtained from the Data Archive at the Space Telescope Sci-ence Institute, which is operated by the Association of Universi-ties for Research in Astronomy, Inc., under NASA contract NAS5-26555. These observations are associated with program
REFERENCES
Aguilar L. A., White S. D. M., 1985, ApJ, 295, 374 Aguilar L. A., White S. D. M., 1986, ApJ, 307, 97Allen M. G., Groves B. A., Dopita M. A., Sutherland R. S., Kewley L. J.,2008, ApJS, 178, 20Allende Prieto C., Lambert D. L., Asplund M., 2001, ApJ, 556, L63Amram P., Plana H., Mendes de Oliveira C., Balkowski C., Boulesteix J.,2003, A&A, 402, 865Arp H., 1966, ApJS, 14, 1Arribas S., Colina L., Bellocchi E., Maiolino R., Villar-Mart´ın M., 2014,A&A, 568, A14Baade W., Minkowski R., 1954, ApJ, 119, 215Baldwin J. A., Phillips M. M., Terlevich R., 1981, PASP, 93, 5Barnes J. E., Hernquist L., 1992, ARA&A, 30, 705Barrera-Ballesteros J. K., et al., 2015, A&A, 579, A45Binette L., Dopita M. A., Tuohy I. R., 1985, ApJ, 297, 476Binney J., Tremaine S., 1987, Galactic dynamicsBlumenthal G. R., Faber S. M., Primack J. R., Rees M. J., 1984, Nature,311, 517Bolton A. S., et al., 2012, AJ, 144, 144Borne K. D., Hoessel J. G., 1988, ApJ, 330, 51Calzetti D., Armus L., Bohlin R. C., Kinney A. L., Koornneef J., Storchi-Bergmann T., 2000, ApJ, 533, 682Calzetti D., Harris J., Gallagher III J. S., Smith D. A., Conselice C. J.,Homeier N., Kewley L., 2004, AJ, 127, 1405Cao C., et al., 2016, ApJS, 222, 16Chien L.-H., Barnes J. E., Kewley L. J., Chambers K. C., 2007, ApJ, 660,L105Chilingarian I. V., Di Matteo P., Combes F., Melchior A.-L., Semelin B.,2010, A&A, 518, A61Cid Fernandes R., Stasi´nska G., Mateus A., Vale Asari N., 2011, MNRAS,413, 1687Cole S., Lacey C. G., Baugh C. M., Frenk C. S., 2000, MNRAS, 319, 168Combes F., Rampazzo R., Bonfanti P. P., Prugniel P., Sulentic J. W., 1995,A&A, 297, 37Conselice C. J., Bluck A. F. L., Mortlock A., Palamara D., Benson A. J.,2014, MNRAS, 444, 1125Darg D. W., et al., 2010, MNRAS, 401, 1552Davoust E., Prugniel P., 1988, A&A, 201, L30Di Matteo P., Combes F., Chilingarian I., Melchior A.-L., Semelin B., 2008,Astronomische Nachrichten, 329, 952Dom´ınguez A., et al., 2013, ApJ, 763, 145Efremov Y. N., Alfaro E., Hodge P., Larsen S., Munoz-Tunon C., 2004,Astronomical and Astrophysical Transactions, 23, 95Ellison S. L., Patton D. R., Simard L., McConnachie A. W., 2008, AJ, 135,1877Ellison S. L., Patton D. R., Mendel J. T., Scudder J. M., 2011, MNRAS,418, 2043Ferreiro D. L., Pastoriza M. G., Rickes M., 2008, A&A, 481, 645Firpo V., Bosch G., Morrell N., 2005, MNRAS, 356, 1357Firpo V., Bosch G., H¨agele G. F., Morrell N., 2010, MNRAS, 406, 1094Firpo V., Bosch G., H¨agele G. F., D´ıaz ´A. I., Morrell N., 2011, MNRAS,414, 3288Fitzpatrick E. L., 1999, PASP, 111, 63Freeman K. C., de Vaucouleurs G., 1974, ApJ, 194, 569Gallagher I. J. S., Parker A., 2010, ApJ, 722, 1962Gil de Paz A., et al., 2007, ApJS, 173, 185Gonz´alez-Serrano J. I., Carballo R., 2000, A&AS, 142, 353Goudfrooij P., Hansen L., Jorgensen H. E., Norgaard-Nielsen H. U., 1994,A&AS, 105H¨agele G. F., D´ıaz ´A. I., Terlevich E., Terlevich R., P´erez-Montero E., Car-daci M. V., 2008, MNRAS, 383, 209Hernadez-Jimenez J. e. a., ”in prep.”Hern´andez-Ibarra F. J., Krongold Y., Dultzin D., del Olmo A., Perea J.,Gonz´alez J., Mendoza-Castrej´on S., Bitsakis T., 2016, MNRAS, 459,291Holtzman J. A., et al., 1992, AJ, 103, 691Hong S., et al., 2011, ApJ, 731, 45Hong S., Calzetti D., Gallagher III J. S., Martin C. L., Conselice C. J., Pel-lerin A., 2013, ApJ, 777, 63MNRAS , 000–000 (0000) M. D. Mora et al.
Hwang H. S., et al., 2011, A&A, 535, A60Iglesias-P´aramo J., et al., 2006, ApJS, 164, 38J¨utte E., Aalto S., H¨uttemeister S., 2010, A&A, 509, A19Kau ff mann G., White S. D. M., Guiderdoni B., 1993, MNRAS, 264, 201Kau ff mann G., et al., 2003, MNRAS, 346, 1055Kaufman M., Grupe D., Elmegreen B. G., Elmegreen D. M., Struck C.,Brinks E., 2012, AJ, 144, 156Kaviraj S., et al., 2012, MNRAS, 423, 49Kennicutt J. R. C., 1998, ApJ, 498, 541Kennicutt R. C., Evans N. J., 2012, ARA&A, 50, 531Kewley L. J., Dopita M. A., Sutherland R. S., Heisler C. A., Trevena J.,2001, ApJ, 556, 121Kewley L. J., Geller M. J., Barton E. J., 2006, AJ, 131, 2004Kojima M., Noguchi M., 1997, ApJ, 481, 132Kormendy J., 1977, ApJ, 217, 406Krabbe A. C., Rosa D. A., Dors O. L., Pastoriza M. G., Winge C., H¨ageleG. F., Cardaci M. V., Rodrigues I., 2014, MNRAS, 437, 1155Lauer T. R., 1986, ApJ, 311, 34Lauer T. R., 1988, ApJ, 325, 49Leitherer C., et al., 1999, ApJS, 123, 3Madore B. F., Nelson E., Petrillo K., 2009, ApJS, 181, 572Manthey E., H¨uttemeister S., Aalto S., Horellou C., Bjerkeli P., 2008, A&A,490, 975Marino R. A., et al., 2013, A&A, 559, A114Moore B., Katz N., Lake G., Dressler A., Oemler A., 1996, Nature, 379,613Mould J. R., et al., 2000, ApJ, 529, 786Mu˜noz-Mateos J. C., Gil de Paz A., Boissier S., Zamorano J., Jarrett T.,Gallego J., Madore B. F., 2007, ApJ, 658, 1006Olave-Rojas D., Torres-Flores S., Carrasco E. R., Mendes de Oliveira C.,de Mello D. F., Scarano S., 2015, MNRAS, 453, 2808Perez J., Michel-Dansac L., Tissera P. B., 2011, MNRAS, 417, 580Pettini M., Pagel B. E. J., 2004, MNRAS, 348, L59Piqueras L´opez J., Colina L., Arribas S., Pereira-Santaella M., Alonso-Herrero A., 2016, A&A, 590, A67Rich J. A., Kewley L. J., Dopita M. A., 2015, ApJS, 221, 28Romano R., Mayya Y. D., Vorobyov E. I., 2008, AJ, 136, 1259Rosa D. A., Dors O. L., Krabbe A. C., H¨agele G. F., Cardaci M. V., PastorizaM. G., Rodrigues I., Winge C., 2014, MNRAS, 444, 2005Rupke D. S. N., Kewley L. J., Barnes J. E., 2010, ApJ, 710, L156Satyapal S., Ellison S. L., McAlpine W., Hickox R. C., Patton D. R., MendelJ. T., 2014, MNRAS, 441, 1297Schlegel D. J., Finkbeiner D. P., Davis M., 1998, ApJ, 500, 525Sutherland R. S., Dopita M. A., 1993, ApJS, 88, 253Tasca L. A. M., et al., 2014, A&A, 565, A10Torres-Flores S., Scarano S., Mendes de Oliveira C., de Mello D. F., AmramP., Plana H., 2014, MNRAS, 438, 1894Villar Mart´ın M., Emonts B., Humphrey A., Cabrera Lavers A., Binette L.,2014, MNRAS, 440, 3202Weistrop D., Nelson C. H., Angione R., Bachilla R., Hancock M., KaiserM. E., 2012, AJ, 143, 98White S. D. M., Rees M. J., 1978, MNRAS, 183, 341Whitmore B. C., Zhang Q., Leitherer C., Fall S. M., Schweizer F., MillerB. W., 1999, AJ, 118, 1551Xu C. K., et al., 2010, ApJ, 713, 330York D. G., et al., 2000, AJ, 120, 1579van Dokkum P. G., 2001, PASP, 113, 1420 MNRAS000