Evidence for Morphology and Luminosity Transformation of Galaxies at High Redshifts
aa r X i v : . [ a s t r o - ph . C O ] M a y Last updated: November 8, 2018
Preprint typeset using L A TEX style emulateapj v. 08/22/09
EVIDENCE FOR MORPHOLOGY AND LUMINOSITY TRANSFORMATION OF GALAXIES AT HIGHREDSHIFTS
Ho Seong Hwang and Changbom Park
School of Physics, Korea Institute for Advanced Study, Seoul 130-722, Korea
Last updated: November 8, 2018
ABSTRACTWe study the galaxy morphology-luminosity-environmental relation and its redshift evolution usinga spectroscopic sample of galaxies in the Great Observatories Origins Deep Survey (GOODS). In theredshift range of 0 . ≤ z ≤ . f E ) strongly increases as it approaches an early-type neighbor, but tendsto decrease as it approaches a late-type neighbor. We find that f E evolves much faster in high densityregions than in low density regions, and that the morphology-density relation becomes significantlyweaker at z ≈
1. This may be because the rate of galaxy-galaxy interactions is higher in high densityregions, and a series of interactions and mergers over the course of galaxy life eventually transform latetypes into early types. We find more isolated galaxies are more luminous, which supports luminositytransformation through mergers at these redshifts. Our results are consistent with those from nearbygalaxies, and demonstrate that galaxy-galaxy interactions have been strongly affecting the galaxyevolution over a long period of time.
Subject headings: galaxies: evolution – galaxies: formation – galaxies: general – galaxies: high-redshift INTRODUCTIONThe role of environment in determining galaxy proper-ties is one of key issues in galaxy formation and evolution.In particular, galaxy morphology is known to depend onenvironment. It was first noted by Hubble & Humason(1931) who found a large population of ellipticalsand lenticulars in galaxy clusters. Later, systematicstudies for the connection between galaxy morphologyand environment suggested the morphology-radius re-lation (Oemler 1974) and the morphology-density re-lation (MDR; Dressler 1980). MDR was detected inthe group environment (Postman & Geller 1984), andwas also found in galaxy clusters at redshifts up to 1(Dressler et al. 1997; Treu et al. 2003; Smith et al. 2005;Postman et al. 2005).Since the Sloan Digital Sky Survey (SDSS; York et al.2000) and Two Degree Field Galaxy Redshift Survey(2dFGRS; Colless et al. 2001) have produced unprece-dentedly large photometric and spectroscopic data ofnearby galaxies, the environmental dependence of galaxyproperties in local universe has been extensively re-visited (e.g., Goto et al. 2003; Balogh et al. 2004a,b;Tanaka et al. 2004; Blanton et al. 2005; Weinmann et al.2006; Park et al. 2007, 2008; Park & Choi 2009;Park & Hwang 2008). Among them, Park et al. (2007)found that the environmental dependence of variousgalaxy properties is almost entirely due to their corre-lation with morphology and luminosity that depend onthe local density. If both morphology and luminosity arefixed, galaxy properties such as color, color gradient, con-centration, size, velocity dispersion, and star formationrate (SFR), are nearly independent of the local density.This study was extended by Park et al. (2008) who took
Electronic address: [email protected], [email protected] into account the effect of the nearest neighbor galaxy.They found that galaxy morphology depends criticallyon the small-scale environment, which is characterizedby the morphology of the nearest neighbor galaxy andthe mass density due to the nearest neighbor galaxy, inaddition to the luminosity and the large-scale density.They suggested a unified scenario that the morphologyand luminosity of a galaxy change through a series ofgalaxy-galaxy interactions and mergers.More recently, Park & Choi (2009) investigated the de-pendence of galaxy properties on both the small- andlarge-scale environments. They found two characteris-tic pair-separation scales where the galaxy propertiesabruptly change: the virial radius r vir , nei of the near-est neighbor galaxy where the effects of galaxy inter-action emerge and ∼ r vir , nei where the galaxies ina pair start to merge. The role of large-scale den-sity is weak when morphology and luminosity are fixed.Park & Hwang (2008) reported that the morphologytransformation in massive galaxy clusters is driven byhydrodynamic interactions between galaxies and by re-peated gravitational interactions among galaxies or be-tween galaxies and their host cluster. Surprisingly, thegalaxy morphology does not depend on the local galaxynumber density at fixed luminosity and fixed nearestneighbor separation. They found that the morphology-radius relation exists within the cluster virial radius butthat galaxy morphology is determined almost entirelyby the nearest neighbor distance and morphology out-side the cluster virial radius. The MDR is only an ap-parent phenomenon through the statistical correlation ofthe local galaxy number density with luminosity and thenearest neighbor distance.The MDR beyond the local universe, was investi-gated mostly in high density regions of galaxy clus- Hwang & Parkters (Dressler et al. 1997; Treu et al. 2003; Smith et al.2005; Postman et al. 2005; see also Poggianti et al.2008). However, thanks to the recent large, deep-field surveys such as the Great Observatories Ori-gins Deep Survey (GOODS; Giavalisco et al. 2004),the All-Wavelength Extended Groth Strip InternationalSurvey (AEGIS; Davis et al. 2007), the VIMOS VLTDeep Survey (VVDS; Le F`evre et al. 2005), the Cos-mic Evolution Survey (COSMOS; Scoville et al. 2007),and the Canada-France-Hawaii Telescope Legacy Sur-vey (CFHTLS), environmental dependence of high red-shift ‘field’ galaxy properties has started to be ex-plored: MDR (Nuijten et al. 2005; Capak et al. 2007;van der Wel et al. 2007), color-density relation (CDR)(Cooper et al. 2007; Cucciati et al. 2006; Cassata et al.2007), and SFR-density relation (Elbaz et al. 2007;Cooper et al. 2008). The MDR and CDR observed lo-cally are also found at high redshifts ( z ∼ z ∼
1) is reversed compared to the local one.Namely, the SFR of high redshift galaxies is larger in highdensity regions than in low density regions (Elbaz et al.2007; Cooper et al. 2008).These observational findings raised important ques-tions: (1) whether the correlations between the galaxyproperties and the local density are the consequenceof environmental-driven evolution; (2) which relation isthe most fundamental one (Cooper et al. 2006, 2007;Poggianti et al. 2008). The physical mechanism of envi-ronmental effects on the galaxy properties is also poorlyunderstood (Cooper et al. 2006). Moreover, there arefew studies that focus on the role of the nearest neighborgalaxy, which turned out to be very important in the evo-lution of galaxy morphology and luminosity (Park et al.2008). Since the galaxy interactions and mergers withneighbor galaxies may be more frequent at high redshifts,it is necessary to investigate the role of neighbor galax-ies in determining the properties of high redshift galaxieshoping that the main mechanism for galaxy evolution beunderstood.In this paper, we study the morphology and luminosityof high redshift galaxies adopting the method similar tothat used for nearby SDSS galaxies (Park et al. 2008).Section 2 describes the observational data used in thisstudy. Environmental dependence of galaxy morphologyand luminosity is given in §
3. Discussion and summaryare given in § §
5, respectively. Throughout thispaper we adopt a flat ΛCDM cosmological model withdensity parameters Ω Λ = 0 .
73 and Ω m = 0 . DATA2.1.
Observational Data Set
We used a spectroscopic sample of galaxies in GOODS.GOODS is a deep multiwavelength survey covering twocarefully selected regions including the Hubble DeepField North (HDF-N) and the Chandra Deep FieldSouth (CDF-S). Hereafter, two GOODS fields centered Fig. 1.—
Spectroscopic completeness as a function of i -band ap-parent magnitude ( a ) and observed color and apparent magnitude( b ) of galaxies in GOODS-North. Those for GOODS-South are in( c ) and ( d ). Vertical dashed lines indicate the apparent magnitudelimits used in this study. on HDF-N and CDF-S are called GOODS-North andGOODS-South, respectively. Total observing area isapproximately 300 arcmin and each region was ob-served by NASA’s Great Observatories ( HST , Spitzer and
Chandra ), ESA’s
XMM-Newton , and several ground-based facilities.
HST observations with Advanced Cam-era for Surveys (ACS) were conducted in four bands: B (F435W, 7200s), V (F606W, 5000s), i (F775W,5000s), and z (F850LP, 10,660s). The drizzled im-ages have a pixel scale of 0 . ′′ pixel − and point-spread function FWHM of ∼ ′′ . Spectroscopic datafor GOODS sources are enormous in the literature:GOODS-North (Cohen et al. 2000; Cowie et al. 2004;Wirth et al. 2004; Reddy et al. 2006) and GOODS-South(Szokoly et al. 2004; Le F`evre et al. 2004; Mignoli et al.2005; Vanzella et al. 2005, 2006, 2008; Ravikumar et al.2007; Popesso et al. 2009). Among the sources in theACS photometric catalog, we used 4443 galaxies whosereliable redshifts are available in the literature for furtheranalysis.Since we are going to investigate the effects of the near-est neighbor galaxies, it is important to identify genuineneighbor galaxies. In addition, since we use a spectro-scopic catalog of galaxies combined from various red-shift surveys with diverse selection criteria, our sampleis heterogeneous. Therefore, it is necessary to know thecompleteness of our spectroscopic sample accurately. Tocompute the spectroscopic completeness, we used a sam-ple from the ACS photometric catalog with the objectshaving SExtractor stellarity class greater than 0 .
79. Thevalue was chosen from the stellarity distribution of gen-uine stars confirmed by the spectroscopic observation. InFigure 1, we plot the completeness of each survey as afunction of the observed magnitude (SExtractor ‘BEST’magnitude) and color. It is seen that the completenessdecreases significantly near i W,BEST ∼ .
5. Themean completeness at i W,BEST ≤ . Fig. 2.—
Evolution corrected, rest frame M B vs. redshift forthe spectroscopic sample of galaxies in GOODS-North ( a ) andGOODS-South ( b ). Solid lines define the volume-limited sampleused in this study. noted that the completeness below the magnitude limitchanges significantly with color, which can cause a biasin the studies of galaxy morphology.The rest frame B -band absolute magnitude M B ofgalaxies is computed based on the ACS photometry withGalactic reddening corrections (Schlegel et al. 1998) and K -corrections (Blanton & Roweis 2007). The evolutioncorrection (an increase of 1 . M B per unit redshift) wasapplied to compute the final rest frame M B (Faber et al.2007). In Figure 2, we show the evolution corrected,rest frame M B as a function of redshift for the sam-ple of galaxies with i W,BEST ≤ .
5. We finallydefine a volume-limited sample of 1332 galaxies with M B ≤ − . . ≤ z ≤ . Morphology Classification
For the volume-limited sample of galaxies shown inFigure 2, we visually inspected the images in individual(
Bviz ) bands and
Bvi pseudo-color images. We dividedthe galaxies into two morphological types: early types(E/S0) and late types (S/Irr). Early-type galaxies arethose with little fluctuation in the surface brightness andcolor and with good symmetry, while late-type galaxiesshow internal structures and/or color variations in thepseudo-color images. However, the total color itself isnot used as a classification criterion. We checked our re-sults by comparing with the morphological classificationof Bundy et al. (2005) who used the same ACS imagesas ours, and found that 98% of our classifications agreewith those of Bundy et al. Some galaxies that were clas-sified as early types in Bundy et al. (2005) are found tobe late types in this study because of color variations inthe pseudo-color images.To see how often we classify galaxies differently becausewe classified galaxies in different wavelength bands acrossredshift, we made the following experiment. We classi-fied galaxies at 0 . < z < . M B = − . v - or i -band images separately, and checked ifthe morphological types of each galaxy agree with eachother. At z = 0 . i -band is centered at ∼ A in the Fig. 3.—
Velocity difference between the target galaxies with − . ≥ M B > − . M B + 0 . a ) and ( b ), and those of GOODS-Southin ( c ) and ( d ). rest frame, and this wavelength roughly corresponds tothe v -band at z ≈ .
5. We found no galaxy was assigneddifferent morphology. Therefore, our morphology clas-sification is not expected to be affected by the redshifteffects. 2.3.
Galaxy Environment
We consider two kinds of environmental factors: a sur-face galaxy number density estimated from five nearestneighbor galaxies (Σ ) as a large-scale environmental pa-rameter, and the distance to the nearest neighbor galaxy( r p ) as a small-scale environmental parameter.The background density, Σ , is defined by Σ =5( πD p, ) − . D p, is the projected proper distance to the5th-nearest neighbor. The 5th-nearest neighbor of eachtarget galaxy was identified among the neighbor galaxieswith M B ≤ − . − to exclude foregroundand background galaxies.To define the small-scale environmental parameter at-tributed to the nearest neighbor, we first find the nearestneighbor of a target galaxy that is closest to the targetgalaxy on the projected sky and satisfies the conditionsof magnitude and relative velocity. We searched for thenearest neighbor galaxy among galaxies that have mag-nitudes brighter than M B = M B, target + 0 . v = | v neighbors − v target | =600 km s − for early-type target galaxy and less than∆ v = 400 km s − for late-type target galaxy. Thesevalues are the same as those used for selecting the near-est neighbor galaxy in the SDSS data (Park et al. 2008).Since we use the volume-limited sample of galaxies with M B ≤ − .
0, we study only the target galaxies brighterthan M B, target = − . − . To fit the velocity distribution, we used the Gaus- Hwang & Parksian plus constant model for each type of target galaxiesby combining the data in two surveys. We obtained, forthe galaxies at projected proper distance r p < h − kpc, the best-fits values σ ∆ v = 351 ±
48 and 216 ±
103 kms − for early- and late-type target galaxies, respectively.It indicates that the velocity limit used for selecting thenearest neighbor is large enough not to miss the neighborgalaxies.The spectroscopic completeness can affect the identi-fication of the genuine nearest neighbor. The complete-ness that depends on the apparent magnitude and coloras shown in Figure 1, can also depend on the distancebetween galaxies due to the difficulty in observing galax-ies close to each other using multi-object spectrograph(MOS). We checked the completeness as a function ofthe projected distance to the target galaxy, and foundthat it does not change with the projected distance. Itmight be because we combined spectroscopic data fromnumerous references, therefore, the difficulty in observingnearby galaxies using MOS is significantly reduced.The virial radius of a galaxy is defined as the radiuswithin which the mean mass density is 200 times thecritical density of the universe ( ρ c ), and is given by r vir = (3 γL/ π/ ρ c ) / , (1)where L is the galaxy luminosity, and γ is the mass-to-light ratio. We assume that the mass-to-light ratio ofearly-type galaxies is on average twice as large as that oflate-type galaxies at the same absolute magnitude M B ,which means γ (early)= 2 γ (late) [see § § ρ c is a function of redshift z andΩ m ( z ) = ρ c ( z ) /ρ b ( z ) = ρ c ( z ) /ρ (1 + z ) , where ρ b and ρ are the mean matter densities in proper and comovingspaces, respectively. Then, the virial radius of a galaxyat redshift z in proper space can be rewritten by r vir ( z ) = [3 γL Ω m, / πρ/ { Ω m, (1 + z ) + Ω Λ , } ] / . (2)Ω m, and Ω Λ , are the density parameters at the presentepoch. We compute the mean mass density ρ using thegalaxies at z = 0 . − . M B = − . − .
0. We foundthat the mean mass density appears to converge when themagnitude cut is fainter than M B = − .
0, which meansthat the contribution of faint galaxies is not significantbecause of their small masses. In this calculation, weweigh each galaxy by an inverse of completeness accord-ing to its apparent magnitude and color (see Fig. 1). Weobtain ρ = 0 . γL ) − ( h − Mpc) − where ( γL ) − isthe mass of a late-type galaxy with M B = −
20. This fi-nal value we adopted is computed using all galaxies with i W,BEST ≤ . . ≤ z ≤ . m, = 0 .
27 and Ω Λ , = 0 .
73, the virialradii of galaxies with M B = − . − . z = 0are 220 and 350 h − kpc for early types, and 180 and280 h − kpc for late types, respectively. The proper-space virial radii of galaxies with the same luminositiesas above, but located at z = 1, are 160 and 250 h − kpcfor early types, and 120 and 200 h − kpc for late types,respectively. RESULTS
Fig. 4.—
Fraction of early-type galaxies in fixed absolute mag-nitude ranges as a function of Σ . Shown are ( a ) GOODS-Northplus South, ( b ) GOODS-North, and ( c ) GOODS-South samples. Background Density Dependence of GalaxyMorphology
In Figure 4, we plot the fraction of early-type galaxiesas a function of the background density, Σ . To accountfor the incompleteness shown in Figure 1, the early-typefraction is computed by weighing each galaxy by theinverse of completeness corresponding to its apparentmagnitude and color. Each of the volume-limited sam-ples is divided into brighter ( − . ≥ M B > − . − . ≥ M B > − .
25) subsamples. Thesolid and dotted lines are early type-fractions ( f E ) forbrighter and fainter subsamples, respectively. The un-certainties of the fraction represent 68% (1 σ ) confidenceintervals that are determined by the bootstrap resam-pling method. Figure 4 clearly shows that the early-type fraction increases along with Σ (i.e., MDR), whichis already known in similar high redshift surveys (e.g.,Capak et al. 2007). We note that the overall early-typefraction is higher for the brighter subsamples in all sur-veys. It implies that the morphology-luminosity rela-tion is already well-established at the redshifts understudy. The background density dependence of morphol-ogy seems stronger for the brighter galaxies. In partic-ular, the early-type fraction of the brighter sample risessharply at high densities of Σ & h Mpc − .3.2. Effects of the Nearest Neighbor
To measure the effects of the nearest neighbor galaxyon the galaxy morphology, we plot, in Figure 5, the frac-tion of early-type galaxies as a function of the distanceto the nearest neighbor. We can see the probability thata galaxy is found to be an early type, strongly dependson the projected distance to the nearest neighbor galaxy( r p ) as well as neighbor’s morphology. When a galaxyis located farther than the virial radius from its nearestneighbor galaxy ( r p & r vir , nei ), the early-type fractionslowly increases as the distance to the neighbor decreases,but its dependence on neighbor’s morphology is weak.On the other hand, when r p . r vir , nei , the early-typefraction increases as the target galaxy approaches anearly-type neighbor, but decreases as it approaches avolution of Galaxy Morphology 5 Fig. 5.—
Early-type fraction as a function of the distance to thenearest neighbor galaxy. The distance is normalized with respectto the virial radius of the nearest neighbor. Galaxies in the com-bined GOODS-North plus South sample are used in ( a ), those ofGOODS-North sample in ( b ), and those of GOODS-South in ( c ). late-type neighbor. It is important to note that the bifur-cation of the early-type fraction, occurs at r p ∼ r vir , nei .In the case of the cosmology we adopt the radius r p = r vir , nei corresponds to the local mass density due to theneighbor of ρ n = 740 ρ . The bifurcation of the early-typefraction at r p ∼ r vir , nei is similarly found by Park et al.(2008) using nearby ( z < .
1) SDSS galaxies.To investigate the effects of large-scale environment onthe morphology in company with the effects of the near-est neighbor galaxy, we study the early-type fraction inthe two-dimensional environmental parameter space asshown in Figure 6. In this case the combined sampleof GOODS-North plus South is used. Galaxies are dis-tributed along the diagonal in this figure due to the sta-tistical correlation between r p and Σ . But there is asignificant dispersion in r p at fixed Σ . It is noted thatgalaxies are located in wide ranges of Σ and r p for bothearly- and late-type neighbor cases even though the earlytypes tend to have relatively larger Σ and smaller r p . Inthis figure one can study the dependence of f E on r p ateach fixed value of Σ by moving along a vertical line.Figure 6 clearly shows that, when r p . r vir , nei , the de-pendence of morphology on the background density andthe nearest neighbor distance becomes completely differ-ent when the neighbor’s morphology changes. When theneighbor is a late type, f E of target galaxies is about 0.2and is almost independent of r p or Σ . But when theneighbor is an early type, f E can rise up to 0.7 whenthe target galaxy enters a region with the backgrounddensity exceeding Σ ≫ h Mpc − . There is also adiscontinous rise in f E when the galaxy approaches theneighbor closer than about 0 . r vir , nei , which correspondsto 50 ∼ h − kpc. When r p ≫ r vir , nei , galaxy morphol-ogy is independent of all environmental factors and f E is about 0.2. It should be emphasized that the differencebetween the left and right panels of Figure 6 is mani-fest only when r p is less than about 2 r vir , nei . As notedabove, a direct effect of the existence of the neighbor on f E is evident at the neighbor distance of r p . . r vir , nei Fig. 6.—
Morphology-environment relation when the nearestneighbor galaxy is an early type ( upper ) or a late type ( lower ).Contours show constant early-type galaxy fraction f E . Galaxiesused here are brighter than M B = − . in the case of early-type nearest neighbor. This confirmsthe net effects of the nearest neighbor on morphologyat fixed large-scale environment for the GOODS galax-ies. However, the effects of late-type neighbors do notappear as strong as what is seen for the nearby SDSSgalaxies (Park & Choi 2009). More data with a highercompleteness are needed to confirm it.In Figure 7, we plot the evolution-corrected, rest frame M B as a function of the distance to the nearest neighborfor the combined sample of GOODS-North plus Southwith M B ≤ − .
5. The lines show the median value ineach projected separation bin. Top panels show that theearly-type galaxies having r p > r vir , nei are significantlybrighter than those at r p < r vir , nei , while the magni-tude difference for late-type galaxies in the two regionsis smaller. If we divide the galaxies into those in rela-tively higher density region (Σ > h Mpc − ) and thosein lower density region (Σ ≤ h Mpc − ), the increaseof galaxy luminosity with increasing r p is manifest in thehigh density region. DISCUSSION4.1.
Morphology and Luminosity Transformation
We found in Figure 5 that the morphological types ofhigh redshift galaxies depend critically on the small-scaleenvironment, which is characterized by the morphologyof the nearest neighbor galaxy and the distance to thenearest neighbor galaxy. If our results are affected bythe trend that the early-type fraction ( f E ) is higher whenthe local density is higher (i.e., MDR), f E should be a Hwang & Park Fig. 7.—
Absolute magnitude of early-( left ) and late-type ( right ) galaxies with M B ≤ − . . ≤ z ≤ . a ) and ( b ), those in high density regions are in ( c ) and ( d ), and those in relatively low density regions are in ( e )and ( f ). Number in parenthesis is the number of galaxies in the combined sample or local density subsample. function of r p which is independent of the neighbor mor-phology, and the two curves in the top panel of Figure5 have the same amplitude at each neighbor separation.The fact that f E is independent of neighbor morphologyat r p > r vir , nei but starts to show a significant depen-dence at r p < r vir , nei , demonstrates that the neighboreffects are the dominating factor of the change in f E andthe large-scale background density is not.Increase of f E with decreasing the neighbor distanceat r p & r vir , nei , can be explained by the tidal effects ofneighbor galaxy. Park et al. (2008) showed that tidal en-ergy deposit relative to the binding energy for the darkhalos of equal mass galaxies, is not negligible at the sep-aration of virial radius. The tidal effects can acceleratethe consumption of cold gas, changing late types to earlytypes. In fact, Park & Choi (2009) showed that the cen-ter of late-type galaxies becomes bluer by the existence ofneighbor galaxies even when r p > r vir , nei independentlyof the morphological type of the neighbor. Their surfacebrightness and central velocity dispersion increase as theneighbor distance decreases, implying the growth of thebulge component.The bifurcation of f E at r p ∼ r vir , nei was interpretedas due to the hydrodynamic effects of the nearest neigh-bor (Park et al. 2008). If a target galaxy approaches alate-type neighbor within one virial radius of the neigh-bor, the cold gas of the neighbor can flow into the targetgalaxy and the target galaxy tends to become a late type.On the other hand, if the target galaxy approaches anearly-type neighbor within one virial radius of the neigh-bor, the hot gas and the tidal force of the neighbor ac-celerate the consumption of cold gas so that the targetgalaxy tends to evolve to an early type.Moreover, it is noted in Figure 6, that the galaxies tendto be early types when they have close early-type neigh-bors even though they are in low density regions, and the galaxies are likely to be late types when they have closelate-type neighbors even though they are in high densityregions. The galaxy morphology appears to depend onthe large-scale background density. But it may be the re-sult of the cumulative effects of neighbor interaction thatis stronger in high density regions (Park et al. 2008).We also found in Figure 7 that isolated galaxies arebrighter than less isolated ones. These results can pro-vide important hints for the evolution of galaxies. Previ-ously, Park et al. (2008) obtained results similar to oursusing the SDSS data (see their Fig. 6), and suggested aunified scenario as follows. Once the separation of twogalaxies becomes small enough, namely r p . . r vir , nei (Park & Choi 2009), they undergo a merger and themerger product will be more massive. The merger prod-uct will typically find itself isolated from its neighborsof comparable mass. As galaxies experience a series ofmerger events, the cold gas will be exhausted due to themerger-induced star formation. Thus the massive galax-ies are likely to be early types. As a supporting evidencefor this scenario, they showed that, at fixed backgrounddensity, post-merger features such as large displacementof the galaxy nucleus from the center, turmoil features,and/or very close double cores, are more frequently seenin the isolated galaxies compared to the less isolated ones.We extended our analysis to another multiwavelengthsurvey, AEGIS, of which survey area ( ∼
710 arcmin ) islarger than GOODS ( ∼
300 arcmin ). Using AEGIS data,we measured the early-type fraction of target galaxies asa function of the nearest neighbor distance (e.g., Fig. 5),and found that the early-type fraction does not changesignificantly with the neighbor distance even though itshowed dependence on the large-scale background den-sity. Furthermore, it does not show the bifurcation in ac-cordance with neighbor’s morphology. These results areinconsistent with those from GOODS and SDSS samplesvolution of Galaxy Morphology 7 Fig. 8.—
Early-type fraction in the combined GOODS-Northplus South sample as a function of redshift. Filled and open circlesindicate the early-type fraction in high and low density regions,respectively, located at the median redshift in each redshift bin. (e.g., Park et al. 2008; Park & Choi 2009). The resultsfrom AEGIS may have been significantly affected by thelow completeness of DEEP2 survey. Since the complete-ness of DEEP2 spectroscopic survey is only about 0.5, itis expected that the nearest neighbor is seriously misiden-tified compared to the case using the GOODS or SDSSdata. Our Monte Carlo experiment shows that the frac-tion of the misidentified nearest neighbor reaches about50% when the sample completeness is 50%. We there-fore conclude that it is not appropriate to study galaxyinteractions using the DEEP2/AEGIS data.4.2.
Redshift Evolution of Galaxy Morphology
If galaxies transform their morphology through a se-ries of interactions and mergers, it is expected that thehigh-redshift galaxies, on average, are less massive andricher in cold gas. Successive consumption of cold gas ingalaxies through interaction and merger events is likelyto transform late types into early types. And thenthe early-type fraction at high redshifts is expected tobe lower compared to that at low redshifts, and thefraction of blue early types among early types to behigher at higher redshifts (Park et al. 2008; Capak et al.2007; see also Conselice et al. 2005, 2008; Lotz et al.2008; Menanteau et al. 2004; Lee et al. 2006; Puzia et al.2007). Moreover, the early-type fraction is expected toevolve differently depending on the background density(Park et al. 2008). Since the mean separation betweengalaxies is relatively smaller and the merger/interactionrate is higher in high density regions, galaxies are ex-pected to show a stronger time evolution in morphologyas well as luminosity, color, and SFR in high densityregions. On the contrary, in low density regions wherethe mean separation between galaxies is larger, the timeevolution of galaxies is expected to be slower becausethe interactions and mergers among galaxies are less fre-quent.There have been several studies of the evolution of mor-phology of high redshift galaxies focusing on the role ofenvironment. For example, Nuijten et al. (2005) usedS´ersic index n and the u ∗ − g ′ color of galaxies as proxiesfor galaxy morphology to study galaxy evolution. Theyfound that the number fraction of “bulge-dominated” Fig. 9.—
Dependence of the early-type fraction on the morphol-ogy of the nearest neighbor galaxy for the galaxies in the combinedGOODS-North plus South sample. Filled and open star symbolsindicate the early-type fraction for the early- and late-type neigh-bor cases in the GOODS-North plus South sample, respectively.The redshift bins are 0 . ≤ z ≤ . . < z ≤ .
0, and pointsare at the median redshifts. Circles denote the early-type fractionfor the nearby galaxies in the SDSS (Park et al. 2008). galaxies with n > u ∗ − g ′ >
1, the fraction of red galaxiesincreases as redshift decreases in all environments. How-ever, since they used photometric redshifts to estimatethe local density, the environment was not accuratelydetermined. In fact, Cooper et al. (2007) found no timeevolution of the fraction of red galaxies in low density re-gions at 0 . < z < .
35 using spectroscopic redshift dataof DEEP2 galaxy redshift survey, though the number ofred galaxies in their sample may be too small to draw astatistically significant conclusion.Recently, Capak et al. (2007) used Gini parameter toselect early-type galaxies in COSMOS data. They foundthat the early-type fraction in all environments growsas redshift decreases, and the growth rate is larger inhigh density regions compared to that in low density re-gions. However, in low density regions with <
100 galax-ies Mpc − , the early-type fraction at z > . n and bumpiness of galaxiesas a proxy of galaxy morphology in SDSS and GOODS-South data. They found no time evolution of the early-type fraction at z < . r p < r vir , nei ).For the early-type fraction at low redshifts, we used theSDSS galaxies with − . > M r > − . z < . M B magni-tude range (Choi et al. 2007; Park et al. 2008). Figure 9clearly shows that the early-type fraction monotonicallyincreases as redshift decreases, and the early-type frac-tion for the case of early-type neighbor is always largerthan that for the late-type neighbor case. Primordial ori-gin is not likely to be the reason for this systematic differ-ence because the neighbor effects are limited within rel-atively tiny volume of the universe associated with eachgalaxy (see Fig. 5). As a comparison, the RMS displace-ment of matter in the flat ΛCDM universe is 7.7 h − Mpcby z=0.5, more than an order magnitude larger thanthe virial radius of typical galaxies (Park & Kim 2009).This result demonstrates an important role of the near-est neighbor in the evolution of galaxy morphology, andis consistent with the prediction that the early-type frac-tion from high to low redshifts will increase because aseries of interactions and mergers tend to transform latetypes into early types.In summary, the early-type fraction in high density re-gions increases significantly as redshift decreases, whilethat in low density regions increases much more gently.The evolution of galaxy morphology is also found to de-pend critically on the small-scale environment, which ischaracterized by the morphology of the nearest neighborgalaxy and the distance to the nearest neighbor galaxy,in addition to the large-scale background density. A largesample of high redshift galaxies is needed to separate be-tween the roles of small- and large-scale environments.We also found that isolated galaxies are brighter thanless isolated ones. All these results are consistent withthe predictions of Park et al. (2008), and confirm theirunified scenario of transformation of galaxy morphologyand luminosity class at high redshifts. It implies thatgalaxy-galaxy interactions play an important role in theevolution of morphology and luminosity classes of galax-ies over a long period of time. CONCLUSIONSUsing the spectroscopic sample of galaxies in theGOODS, we presented evidence for morphology and lu-minosity transformation of galaxies at 0 . ≤ z ≤ .
0. Wedetermined the morphological types of all high redshiftgalaxies by visual inspection, and used spectroscopic red-shifts of galaxies to determine the environmental param- eters. We examined the effects of the nearest neighborgalaxy and the local galaxy number density on the galaxymorphology. Our main results are as follows:1. The early-type fraction increases with the surfacegalaxy number density estimated from 5th-nearestneighbor galaxies (Σ ). This confirms the MDRfollowed by high redshift galaxies (0 . ≤ z ≤ . f E ) de-creases with increasing distance, and is indepen-dent of morphological type of the nearest neighbor.3. When the separation with the nearest neighborgalaxy is smaller than the virial radius of the neigh-bor, f E increases as the target galaxy approachesan early-type neighbor, but tends to stay constantas it approaches a late-type neighbor. Conformityin morphology between neighboring galaxies is con-firmed at high redshifts. The realm of conformityis confined within the virialzed region associatedwith each galaxy plus dark halo system.4. We find that more isolated galaxies are more lumi-nous. It can be explained by the luminosity evolu-tion of galaxies through a series of mergers.5. The early-type fraction f E increases very rapidlyas redshift decreases in high density regions, butincreases only mildly in low density regions. At z > REFERENCESBalogh, M. L., Baldry, I. K., Nichol, R., Miller, C., Bower, R., &Glazebrook, K. 2004a, ApJ, 615, L101Balogh, M., et al. 2004b, MNRAS, 348, 1355 Blanton, M. R., Eisenstein, D., Hogg, D. W., Schlegel, D. J., &Brinkmann, J. 2005, ApJ, 629, 143Blanton, M. R., & Roweis, S. 2007, AJ, 133, 734Bundy, K., Ellis, R. S., & Conselice, C. J. 2005, ApJ, 625, 621 volution of Galaxy Morphology 9volution of Galaxy Morphology 9