1:
2:
3: \documentclass[12pt,preprint]{aastex}
4: \slugcomment{Published in ApJ Letter 617 (2004) L9}
5:
6:
7: \shorttitle{Evolution of Close Galaxy Pairs in DEEP2}
8: \shortauthors{Lin et al.}
9:
10:
11:
12: \begin{document}
13:
14: \title{THE DEEP2 GALAXY REDSHIFT SURVEY: EVOLUTION OF CLOSE GALAXY PAIRS AND MAJOR-MERGER RATES UP TO z $\sim$ 1.2 \altaffilmark{1}}
15:
16:
17: \author{LIHWAI LIN \altaffilmark{2,3}, DAVID C. KOO \altaffilmark{3}, CHRISTOPHER N.
18: A. WILLMER \altaffilmark{3,4}, DAVID R. PATTON \altaffilmark{5},
19: CHRISTOPHER J. CONSELICE \altaffilmark{6}, RENBIN YAN
20: \altaffilmark{7}, ALISON L. COIL \altaffilmark{7}, MICHAEL C.
21: COOPER \altaffilmark{7}, MARC DAVIS \altaffilmark{7,8}, S. M.
22: FABER \altaffilmark{3}, BRIAN F. GERKE \altaffilmark{8}, PURAGRA
23: GUHATHAKURTA \altaffilmark{3}, AND JEFFREY A. NEWMAN
24: \altaffilmark{9}}
25:
26: \altaffiltext{1}{ Some of the data presented herein were obtained
27: at the W. M. Keck Observatory, which is operated as a scientific
28: partnership among the California Institute of Technology, the
29: University of California, and the National Aeronautics and Space
30: Administration. The Observatory was made possible by the generous
31: financial support of the W. M. Keck Foundation.}
32: \altaffiltext{2}{Department of Physics, National Taiwan
33: University, No. 1, Sec. 4, Roosevelt Road, Taipei 106, Taiwan;
34: d90222005@ntu.edu.tw} \altaffiltext{3}{University of California
35: Observatories/Lick Observatory, Department of Astronomy and
36: Astrophysics, University of California at Santa Cruz, 1156 High
37: Street, Santa Cruz, CA 95064.} \altaffiltext{4}{On leave from
38: Observat\'{o}rio Nacional, Brazil.} \altaffiltext{5}{Department of
39: Physics, Trent University, 1600 West Bank Drive, Peterborough, ON
40: K9J 7B8, Canada.} \altaffiltext{6}{Department of Astronomy,
41: California Institute of Technology, Pasadena, CA 91125.}
42: \altaffiltext{7}{Department of Astronomy, University of California
43: at Berkeley, 601 Campbell Hall, Berkeley, CA 94720-3411.}
44: \altaffiltext{8}{Department of Physics, University of California
45: at Berkeley, 366 LeConte Hall, Berkeley, CA 94720-7300.}
46: \altaffiltext{9}{Hubble Fellow; Institute for Nuclear and Particle
47: Astrophysics, Lawrence Berkeley National Laboratory, Berkeley, CA
48: 94720.}
49:
50:
51: \begin{abstract}
52: \hspace{3mm}We derive the close, kinematic pair fraction and
53: merger rate up to redshift $z \sim 1.2$ from the initial data of
54: the DEEP2 Redshift Survey. Assuming a mild luminosity evolution,
55: the number of companions per luminous galaxy is found to evolve as
56: $(1+z)^{m}$, with $m= 0.51\pm0.28$; assuming no evolution, $m =
57: 1.60\pm0.29$. Our results imply that only $9\%$ of present-day
58: $L^{*}$ galaxies have undergone major mergers since $z\sim1.2$ and
59: that the average major merger rate is about $4\times 10^{-4}$
60: $h^{3}$ Mpc$^{-3}$ Gyr$^{-1}$ for $z \sim 0.5 - 1.2$. Most
61: previous studies have yielded higher values.
62: \end{abstract}
63:
64:
65:
66: \keywords{galaxies:interactions - galaxies:evolution - large-scale
67: structure of Universe}
68:
69: \section{INTRODUCTION}
70: Galaxy interactions and mergers are an integral part of our
71: current paradigm of the hierarchical formation and evolution of
72: galaxies. Such processes are expected to affect the morphologies,
73: gas distributions, and stellar populations of galaxies (e.g.,
74: Mihos \& Henquist 1994, 1996; Dubinski, Mihos, \& Henquist 1996).
75: Although mergers are rare today (Patton et al. 2000, hereafter
76: P2000), cold dark matter $N$-body simulations show merger rates of
77: halos increasing with redshift as $(1+z)^{m}$, with $2.5\leq
78: m\leq3.5$ \citep{gov99,got01}.
79:
80: Observations, however, have yielded results with $0\leq m\leq4$
81: \citep{zep89,bur94,car94,woo95,yee95,
82: pat97,neu97,le00,car00,pat02,bun04}. The diverse results are
83: likely due to different pair criteria, observational techniques,
84: selection effects, and cosmic variance (Abraham 1999; P2000). To
85: identify close pairs, the most secure method is via spectroscopic
86: redshifts for both galaxies to find kinematic pairs \citep[
87: hereafter P2002]{pat02}. This Letter adopts the approach of P2002
88: and uses the early data of the DEEP2 (Deep Extragalactic
89: Evolutionary Probe 2) Redshift Survey (Davis et al. 2003; S. M.
90: Faber et al. 2004, in preparation) to derive the pair fraction and
91: merger rates up to $z=1.2$. In $\S$2, we describe the sample and
92: selection functions. In $\S$3, we use both the projected
93: separation and the relative velocity to select close galaxy pairs
94: and to determine their evolution. Major merger rates out to
95: $z\sim1.2$ are computed in $\S$4. The results are discussed in
96: $\S$5. Throughout this Letter, we adopt a cosmology of $H_{0}=70$
97: km s$^{-1}$ Mpc$^{-1}$, $\Omega_{m}=0.3$, and
98: $\Omega_{\Lambda}=0.7$.
99:
100: \section{DATA AND SELECTION FUNCTIONS}
101: The DEEP2 Redshift Survey (DEEP2 for short) will measure redshifts
102: for $\sim50,000$ galaxies at $z\sim 1$ \citep{dav03} using DEIMOS
103: (DEep Imaging Multi-Object Spectrograph; Faber et al. 2003) on the
104: 10 m Keck II telescope. The survey convers four fields, with Field
105: 1 (Extended Groth Strip) being a strip of 0.25 $\times$ 2
106: deg$^{2}$ and Fields 2, 3, and 4 each being 0.5 $\times$ 2
107: deg$^{2}$. The photometry is based on $BRI$ images taken with the
108: 12k $\times$ 8k camera on the Canada-France-Hawaii Telescope (see
109: Coil et al. 2004 for details). Galaxies are selected for
110: spectroscopy using a limit of $R_{AB}=24.1$ mag. Except in Field
111: 1, a two-color cut is also applied to exclude galaxies with
112: redshifts $z < 0.75$. A 1200 line mm$^{-1}$ grating (R $\sim$
113: 5000) is used with a spectral range of $6400 \-- 9000$ \AA, where
114: the [O II] $\lambda$3727 doublet would be visible at $z\sim0.7 \--
115: 1.4$. The data are reduced using an IDL pipeline developed at
116: UC-Berkeley (J. A. Newman et al. 2004, in preparation). The
117: $K$-correction is derived from spectra of local galaxies (C. N. A.
118: Willmer et al. 2004, in preparation). The data used here are from
119: Fields 1 and 4, cover $\sim 0.4$ deg$^{2}$, and contain $\sim5000$
120: galaxies.
121:
122: To measure the spectroscopic selection functions (see L. Lin et
123: al. 2004, in preparation, for details), we compared the sample
124: with successful redshifts to galaxies in the full photometric
125: catalog that satisfy the limiting magnitude and two-color cuts.
126: Following analogous approaches by \citet{yee96} and P2002, we
127: calculate the spectroscopic weight $w$ of each galaxy as $1/S$,
128: where $S$ is the spectroscopic selection function derived from the
129: $R$ magnitude, $B-R$ and $R-I$ colors, $R-$band surface brightness
130: $\mu_{R}$, and local galaxy density of the galaxy itself. To
131: correct for bias due to slit collisions \citep{dav04}, we also
132: compute the angular weight, $w(\theta)$, as a function of angular
133: separation $\theta$ (P2002). For $\theta \leq 20"$, we find $0.95
134: \leq w(\theta) \leq1.5$, a result confirmed with the DEEP2 mock
135: catalog of \citet{yan04}.
136:
137: \section{PAIR STATISTICS}
138: A major limitation on the direct measurement of merger fractions
139: for galaxies is the difficulty in identifying on-going mergers,
140: especially for distant galaxies. One alternative is to count only
141: the pairs with projected separations $\vartriangle$$r$ and
142: relative line-of-sight heliocentric velocities $\vartriangle$$v$
143: less than $r_{max}$ and $v_{max}$, respectively. A large fraction
144: of pairs with physical separations less than 20 $h^{-1}$ kpc and
145: velocity difference less than 500 km\ s$^{-1}$ appear to have
146: disturbed morphologies or signs of interactions, and these
147: galaxies are expected to merge within 0.5 Gyr (P2000). In our
148: work, close pairs are defined such that their projected
149: separations satisfy 10 $h^{-1}$ kpc $ \leq $ $\vartriangle r$
150: $\leq r_{max}$ and their rest-frame relative velocities
151: $\vartriangle$$v$ are less than 500 km\ s$^{-1}$. We adopt an
152: inner cutoff with a projected distance 10 $h^{-1}$ so as to avoid
153: the ambiguity between very close pairs and single galaxies with
154: multiple star-forming knots. We choose values of $r_{max}$ = 30,
155: 50, and 100 $h^{-1}$ kpc, where 30 $h^{-1}$ kpc is most likely to
156: include genuine merger pairs, while the two larger separations
157: provide larger samples and thus better statistics. To ensure the
158: selection of the same types of galaxies at different redshifts in
159: the presence of luminosity evolution, P2002 adopted a specific
160: range in evolution-corrected absolute magnitude $M_{B}^{e}$,
161: defined as $M_{B}+Qz$, where the evolution is parameterized as
162: $M(z)=M(z=0)-Qz$. Following P2002, we adopt $Q=1$ as a primary
163: model, and we also study the effect of different models on the
164: pair statistics. We restrict the analysis to galaxies with
165: luminosities $-21\leq M_{B}^{e}\leq -19$ for $z=0.45 - 1.2$. Since
166: there is only a 2 mag range in our sample, and assuming a constant
167: $M/L$ ratio, most observed pairs are thus major mergers, i.e.,
168: mass ratios between 1 : 1 and 6 : 1. Using data from the DEEP2
169: Fields 1 and 4 helps us reduce the effects of cosmic variance.
170: Following P2000 and P2002, we compute the average number of
171: companions per galaxy
172: \begin{equation}
173: N_{c}=\frac{\sum^{N_{tot}}_{i=1}\sum_{j}w_{j}w(\theta)_{ij}}{N_{tot}},
174: \end{equation}
175: where $N_{tot}$ is the total number of galaxies within the chosen
176: magnitude range, $w_{j}$ is the spectroscopic weight for the
177: $j$th companion belonging to the $i$th galaxy, and
178: $w(\theta)_{ij}$ is the angular weight for each pair as described
179: in Section 2. Details of the weighting scheme can be found in
180: Section 5 of P2002 and references therein. This quantity $N_{c}$
181: is similar to the pair fraction when there are few triplets or
182: higher order $N$-tuples in the sample, which is the case here. In
183: this work, $N_{c}$ will sometimes simply be referred to as the
184: pair fraction.
185:
186: Figure 1 shows $N_{c}$ versus redshift $z$ for a
187: sample with $v_{max} = 500$ km\ s$^{-1}$ and 10 $h^{-1}$ kpc
188: $\leq$
189: $\vartriangle$$r \leq r_{max}$ with $r_{max}= $ 30, 50, and 100 $h^{-1}$ kpc, respectively. In the case of using
190: $r_{max}=$ 50 $h^{-1}$ kpc, we find 79 paired galaxies out of 2547
191: galaxies. The derived $N_{c}$ is $\sim8\%$ at $z\sim0.6$ and
192: increases to 10\% at $z\sim1.1$. Figure 1 also shows results from
193: the SSRS2 (Southern Sky Redshift Survey; P2000) at $z\sim 0.015$
194: and CNOC2 (Canadian Network for Observational Cosmology; P2002) at
195: $z\sim 0.3$, after corrections that adopt the same cosmology and
196: the same luminosity range.
197:
198: The DEEP2 sample is $R$-band-selected and hence, for redshifts
199: greater than 0.8, is biased toward galaxies that are bright in the
200: rest-frame UV, i.e., against faint red galaxies, especially at
201: redshifts $z
202: >1.0$.
203: To avoid this bias, we divide
204: our sample into blue and red galaxies using the color bimodality feature
205: and then repeat the
206: calculation of $N_{c}$ using $r_{max}=$ 50 $h^{-1}$ kpc for blue
207: galaxies at $z$ = 0.3 (CNOC2), 0.6, and 0.85 bins (DEEP2). When
208: using only blue galaxies, the values of the pair fraction
209: $N_{c}^{blue}$ decrease at $z$ = 0.3 and 0.85 but increase at $z$
210: = 0.6 compared to the results using the original sample. The
211: changes in the measured value of $N_{c}$ could be up to a factor
212: of 2, depending on the field and redshift. To mitigate the
213: underestimation of the pair fraction in the redshift bin $1.0<z<1.2$,
214: we calculate a corrected $N_{c}$ at $z=1.1$ as
215: \begin{equation}
216: N_{c}^{cor}=\max[N_{c},2N_{c}^{blue}],
217: \end{equation}
218: shown as open triangles in Figure 1. Parameterizing the evolution
219: of $N_{c}$ as $Nc(0)(1+z)^{m}$, we find that the best fit of
220: $(N_{c}(0),m)$ is $(0.029\pm0.005,1.08\pm0.40)$ with
221: $\chi^{2}=0.27$ for $r_{max}=30$ $h^{-1}$ kpc,
222: $(0.068\pm0.008,0.51\pm0.28)$ with $\chi^{2}=0.89$ for
223: $r_{max}=50$ $h^{-1}$ kpc, and $(0.177\pm0.014,0.47\pm0.18)$ with
224: $\chi^{2}=1.2$ for $r_{max}=100$ $h^{-1}$ kpc. For the highest
225: redshift bin, $N_{c}^{cor}$ instead of $N_{c}$ is used for the
226: above fitting procedure. The best fits are shown as long-dashed
227: curves in Figure 1. For three different choices of $r_{max}$, we
228: find only a small increase of $N_{c}$ from redshifts $z=0$ to
229: $z=1.2$. In figure 1, we also list the upper limits of errors from
230: the cosmic variance for each redshift bin calculated by assuming a
231: spherical geometry of the survey volume. Incorporating the cosmic
232: variance into the fitting procedure slightly raises $m$ to
233: $0.62\pm0.49$ for $r_{max}=50$ $h^{-1}$ kpc.
234:
235: The adopted choice of $Q = 1$ in the luminosity evolution model
236: sets the luminosity range of the sample at each redshift and thus
237: affects our results. To assess this effect, we repeat the pair
238: analysis for other choices of $Q$. The luminosity ranges are
239: chosen such that they are locked to $-22\leq M_{B}\leq -20$ at
240: $z=1$. Using $r_{max}=$ 50 $h^{-1}$ kpc and an upper redshift
241: limit of $z = 1$, we find that $m$ varies from $1.60\pm0.29$ for
242: $Q = 0$, $0.86\pm0.29$ for $Q = 0.5$, $0.41\pm0.30$ for $Q = 1$,
243: to $-0.24\pm0.35$ for $Q = 2$. In figure 2, we plot $N_{c}$ as a
244: function of absolute $B$-band magnitude ($M_{B}$) for three
245: redshift samples. A clear dependence of $N_{c}$ on $M_{B}$ is
246: evident. The trend that $m$ decreases with $Q$ can be understood
247: as the result of including fainter galaxies and thus having larger
248: $N_{c}$ at lower redshifts when adopting higher $Q$ values.
249: Nevertheless, even with the lowest choice of $Q = 0$ to maximize
250: $m$, models with $m>3.5$ are ruled out at a 3 $\sigma$ level of
251: confidence for all three choices of separation limits.
252:
253: \section{MAJOR MERGER RATES}
254: The comoving merger rate, usually defined as the number of mergers per
255: unit time per comoving volume, can be
256: estimated as
257: \begin{equation}
258: N_{mg}=0.5n(z)N_{c}(z)C_{mg}T^{-1}_{mg},
259: \end{equation}
260: where $T_{mg}$ is the time-scale for physically associated pairs
261: to merge, $C_{mg}$ denotes the fraction of galaxies in close pairs
262: that will merge in $T_{mg}$, and $n(z)$ is the comoving number
263: density of galaxies. The factor 0.5 is to convert the number of
264: galaxies into the number of merger events. The merger rate
265: calculated above, however, is not suitable for comparison to the
266: merger rate derived from morphological approaches. The reason is
267: that while we have restricted the luminosity range of companions
268: to compute $N_{c}$, the morphological approach does not. To
269: correct for this restriction, equation (3) can be modified :
270: \begin{equation}
271: N_{mg}=(0.5+G)n(z)N_{c}(z)C_{mg}T^{-1}_{mg},
272: \end{equation}
273: where the added parameter $G$ accounts for the excess number of
274: companions failing to fall into our sample. Assuming that the
275: maximum mass ratio needed to yield significant morphology
276: distortions is 4 : 1, we calculate the value of $G$ as 1.24 for
277: $-21\leq M_{B}^{e}\leq -19$ using the local luminosity function
278: \citep{dri03}. $T_{mg}$ depends on the relative mass ratio,
279: dynamical orbit, and detailed structure of the two merging
280: galaxies, and thus it varies from case to case. Here we adopt
281: $T_{mg}$ = 0.5 Gyr, a value suggested by $N$-body simulations and
282: simplified models (Mihos 1995; P2000). The value $C_{mg}$ is
283: approximately 0.5 for close pairs with 5 $h^{-1}$ kpc $\leq$
284: $\vartriangle$$r$ $\leq 20$ $h^{-1}$ kpc and $v_{max} = 500$ km\
285: s$^{-1}$; this estimate is based on morphological studies of local
286: close pairs (P2000). Nevertheless, both $T_{mg}$ and $C_{mg}$
287: remain uncertain.
288:
289: The derived merger rates are displayed in Figure 3. Here we have
290: applied the pair fraction evolution of $m=0.51$ derived for the
291: $r_{max}=50$ $h^{-1}$ kpc case to the $z\sim0$ (SSRS2) $N_{C}$
292: result for pairs with 5 $h^{-1}$ kpc $\leq$ $\vartriangle$$r$
293: $\leq 20$ $h^{-1}$ kpc. The parameter $n(z)$ is calculated as the
294: sum of the number of galaxies in the adopted magnitude range, each
295: weighted by its spectroscopic weight and divided by the comoving
296: volume occupied by the included sources. The errors shown for
297: DEEP2 measurements represent $40\%$ variations that are typical
298: for close pair counts in our sample, and they do not include the
299: uncertainties of $T_{mg}$ and $C_{mg}$. Also shown are the results
300: from \citet{con03}, who relied on morphologically identified
301: mergers for $M_{B}<-19$ using \emph{Hubble Space Telescope}
302: (\emph{HST}) data in the Hubble Deep Field-North (\emph{filled
303: triangles}). Clearly, the comoving merger rate in our data changes
304: little with redshift. From $z\sim0.5$ to $z\sim1.2$, the average
305: is $N_{mg} \sim4\times 10^{-4}$ $h^{3}$ Mpc$^{-3}$ Gyr$^{-1}$ for
306: $-21\leq M_{B}^{e} \leq -19$. This value is about 1 order of
307: magnitude lower than the average merger rate for galaxies with
308: $M_{B} \leq -19$ derived by Conselice et al. (2003).
309:
310: Finally, we calculate the merger remnant fraction, $f_{rem}$,
311: defined as the fraction of present-day galaxies that have
312: undergone major mergers (P2000). Adopting the merger fraction
313: $C_{mg}$ to be 0.5 and $T_{mg}$ to be 0.5 Gyr, we estimate, using
314: equation (32) in P2000, that about $9\%$ of present $L^{*}$
315: luminous galaxies have undergone major mergers since $z\sim1.2$.
316:
317: \section{DISCUSSION}
318: Studies up to $z\sim 1$ using pair counts have found a wide range
319: in the evolution of merger rates. For example, based on roughly
320: 300 Canada-France Redshift Survey galaxies measured with
321: \emph{HST}, Le F\`{e}vre et al. (2000) concluded that the pair
322: fraction evolves with $m=2.7\pm0.6$ while the fraction of merger
323: candidates evolves with $m=3.4\pm0.6$. In contrast, Carlberg et
324: al. (2000) measured the mean fractional pair luminosity from $z$ =
325: 0.2 to 1 and found no evolution with $m~\sim0\pm1.4$, consistent
326: with our result. Since the pair fraction depends on adopted
327: luminosity limits, differences among studies may depend on the
328: choice of luminosity evolution models. \citet{le00}, e.g., applied
329: no corrections while \citet{car00} adopted the Q=1 model, as we do
330: here. The key advantages of our pair sample over previous surveys
331: include having measurements at $z > 1$, a larger sample, and more
332: restrictive pair criteria than those adopted by \citet{le00} and
333: \citet{car00}.
334:
335: The discrepancy in the derived merger rates at $z\sim1$ between
336: our estimates and those from \citet{con03} is more difficult to
337: reconcile. Besides the choice of luminosity ranges and the
338: uncertainties in $C_{mg}$ and $T_{mg}$, other factors may explain
339: the discrepancy. First, cosmic variance can always come into play,
340: since most morphological studies have been forced to use the few
341: small fields covered deeply by \emph{HST} imaging, whereas
342: pair-count surveys cover much larger areas. Second, the two
343: approaches may sample interacting and merging galaxies at
344: different stages, with morphological approaches identifying very
345: advanced mergers and merger remnants, while the pairs detect some
346: of the same systems before distortions are discernible. Matching
347: the derived merger rates from these two approaches requires
348: knowledge of the precise timescales of close pairs and of the
349: duration time for the appearance of distorted morphologies.
350: However, neither is well understood yet. Third, while our close
351: pairs identify only major mergers, morphological criteria may be
352: sensitive to minor mergers as well. Finally, both the pair-count
353: and morphology methods are subject to systematic uncertainties
354: that require detail modeling (Bell 2004). Given the current
355: discrepancy in results, it behooves us to study the connection
356: between kinematic pairs and morphologically disturbed galaxies at
357: various redshifts, both via observational approaches (P2000; J.
358: Lotz et al. 2004, in preparation) and through numerical
359: simulations (T. J. Cox et al. 2004, in preparation).
360:
361: The merger rates from pair counts are also an excellent test for
362: those estimated from $N$-body simulations, since the former reveal
363: the behavior of luminous baryons while the latter reflect the
364: behavior of dark halos. Gottl\"{o}ber et al. (2001) defined merger
365: rates in $N$-body simulations as equivalent to the major merger
366: fraction per gigayear by tracing the formation and evolutionary
367: history of each halo. These fractions will differ from the pair
368: fractions roughly by a constant, and therefore their redshift
369: evolution is amenable to direct comparisons. They found merger
370: rates evolving as $(1+z)^{3}$ up to $z\sim2$. This theoretical
371: value of $m = 3$ is significantly higher than the $m \sim 0.5$ we
372: find from observed pair counts. We should, however, be wary about
373: this comparison for the following reasons. First, definitions of
374: merger rates in observations and simulations may not be consistent
375: with each other, since the halos and visible galaxies span
376: different size scales and since $N$-body simulations also suffer
377: limitations in resolution. Second, the merger rates/fractions of
378: halos are likely to be a function of halo mass, which is suggested
379: by our finding that the galaxy pair fraction is a function of
380: luminosity. Pair count works using $K$-band-selected samples
381: (e.g., Bundy et al. 2004) can also provide another avenue for
382: testing the mass dependence of merger rates since the $K$-band
383: luminosity is more representative of any underlying stellar mass
384: and suffers less from evolution corrections. Although preliminary,
385: mock catalog simulations by E. Van Kampen et al. (2004, in
386: preparation) predict flat slopes for the pair-count fractions (see
387: Fig. 3 of Bell 2004), just as seen in our observations. We expect
388: a dramatic improvement in our understanding of merger histories
389: via pair counts after more realistic comparisons to simulations
390: are possible and especially after a 10-fold increase in sample
391: size when DEEP2 is complete in 2005.
392:
393: \acknowledgments We thank the referee for helpful comments, J.
394: Primack, J. Lotz, and B. Allgood for useful discussions, and the
395: Keck staff for dedicated assistance. L. L. acknowledges support
396: from Taiwan via the COSPA project and NSC grant
397: NSC92-2112-M-002-021, D. R. P. from NSERC of Canada, and the DEEP2
398: team from NSF grants AST00-71048, AST00-71198, and KDI-9872979. We
399: close with thanks to the Hawaiian people for allowing us to use
400: their sacred mountain.
401:
402:
403: \begin{thebibliography}{}
404: \bibitem[Abraham(1999)]{abr99} Abraham, R. G. 1999, IAU Symp. 186, Galaxy Interactions at Low and High Redshifts (astro-ph/9802033)
405: \bibitem[Bundy et al.(2004)]{bun04} Bundy, K., Fukugita, M.,Ellis, R. S., Kodama, T., \& Conselice, C. J. 2004, \apj, 601, L123
406: \bibitem[Bell(2004)]{bel04} Bell, E. F. 2004, preprint (astro-ph/0408023)
407: \bibitem[Burkey, Keel, \& Windhorst(1994)]{bur94} Burkey, J. M., Keel, W. C., \& Windhorst, R. A. 1994, \apj, 429, L13
408: \bibitem[Carlberg et al.(1994)]{car94} Carlberg, R. G., Pritchet, C. J., \& Infante, L. 1994, \apj, 435, 540
409: \bibitem[Carlberg et al.(2000)]{car00} Carlberg, R. G., et al. 2000, \apj, 532, L1
410: \bibitem[Coil et al.(2004)]{coi04} Coil, A. L., et al. 2004, ApJ, submitted (astro-ph/0403423)
411: \bibitem[Conselice et al.(2003)]{con03} Conselice, C. J., Bershady, M. A., Dickinson, M., \& Papovich, C. 2003, \aj, 126, 1183
412: \bibitem[Davis et al.(2003)]{dav03} Davis, M., et al. 2003, SPIE, 4834, 161 (astro-ph/0209419)
413: \bibitem[Davis, Gerke, \& Newman(2004)]{dav04} Davis, M., Gerke, B., \& Newman, J. A. 2004, preprint(astro-ph/0408344)
414: \bibitem[Driver \& De Propris(2003)]{dri03} Driver, S., \& DePropris, R. 2003, Ap\&SS, 285, 175
415: \bibitem[Dubinski, Mihos, \& Hernquist(1996)]{dub96} Dubinski, J., Mihos, J. C., \& Hernquist, L. 1996, \apj, 462, 576
416: \bibitem[Faber et al.(2003)]{fab03} Faber, S. M., et al. 2003, SPIE, 4841, 1657
417: \bibitem[Gottl\"{o}ber, Klypin, \& Kravtsov(2001)]{got01} Gottl\"{o}ber, S., Klypin, A., \& Kravtsov, A. V. 2001, \apj, 546, 223
418: \bibitem[Governato et al.(1999)]{gov99} Governato, F., Gardner, J. P., Stadel, J., Quinn, T., \& Lake, G. 1999, \aj, 117, 1651
419: \bibitem[Le F\`{e}vre et al.(2000)]{le00} Le F\`{e}vre, O., et al. 2000, \mnras, 311, 565
420: \bibitem[Mihos \& Hernquist(1994)]{mih94} Mihos, J. C., \& Hernquist, L. 1994, \apj, 425, L13
421: \bibitem[Mihos (1995)]{mih95} Mihos, J. C. 1995, \apj, 438, L75
422: \bibitem[Mihos \& Hernquist(1996)]{mih96} Mihos, J. C., \& Hernquist, L. 1996, \apj, 464, 641
423: \bibitem[Neuschaefer et al.(1997)]{neu97} Neuschaefer, L. W., Im, M., Ratnatunga, K. U., Griffiths., R. E., \& Casertano, S. 1997, \apj, 480, 59
424: \bibitem[Patton et al.(1997)]{pat97} Patton, D. R., Pritchet., C. J., Yee, H. K. C., Ellingson, E., \& Carlberg, R. G. 1997, \apj, 475, 29
425: \bibitem[Patton et al.(2000)]{pat00} Patton, D. R., Carlberg, R. G., Marzke, R. O., Pritchet., C. J., da Costa, L. N., \& Pellegrini, P. S. 2000, \apj, 536, 153
426: \bibitem[Patton et al.(2002)]{pat02} Patton, D. R., et al. 2002, \apj, 565, 208
427: \bibitem[Woods, Fahlman, \& Richer(1995)]{woo95} Woods, D., Fahlman, G. G., \& Richer, H. B. 1995, \apj, 454, 32
428: \bibitem[Yan, White, \& Coil(2004)]{yan04} Yan, R., White, M., Coil, A. L. 2004, \apj, 607, 739
429: \bibitem[Yee \& Ellingson(1995)]{yee95} Yee, H. K. C., \& Ellingson, E. 1995, \apj, 445, 37
430: \bibitem[Yee, Ellingson, \& Carlberg(1996)]{yee96} Yee, H. K. C., Ellingson, E., \& Carlberg, R. G. 1996, \apjs, 102, 269
431: \bibitem[Zepf \& Koo(1989)]{zep89} Zepf, S. E., \& Koo, D. C. 1989, \apj, 337, 34
432: \end{thebibliography}
433:
434: \clearpage
435:
436: \begin{figure}[t]
437: \plotone{f1.eps} \caption{Pair fraction, $N_c$, as a function of
438: redshift $z$. The filled squares represent measurements from DEEP2
439: Field 1, and the open squares represent those from Field 4. The
440: crosses mark the results from SSRS2 and CNOC2. The open triangles
441: show the corrected pair fraction $N_{c}^{cor}$ for the highest
442: redshift bin. The long-dashed curves are best fits by the form
443: $(1+z)^{m}$, using DEEP2, SSRS2, and CNOC2 data (see text); the
444: dot-dashed curves represent evolution with $m=2.5$ for reference.
445: From the top to the bottom, $r_{max}=100$ $h^{-1}$ kpc,
446: $r_{max}=50$ $h^{-1}$ kpc, and $r_{max}=30$ $h^{-1}$ kpc. Note
447: that the bottom panel has a different vertical scale. The error
448: bars shown in the plot and used for fitting are calculated by
449: bootstrapping. The numbers appearing in the bottom panel indicate
450: the upper limits of errors (in units of percentage) from the
451: cosmic variance for each redshift bin. Those in the parentheses
452: are for open symbols. \label{fig1}}
453: \end{figure}
454:
455:
456: \clearpage
457:
458: \begin{figure}[t]
459: \plotone{f2.eps} \caption{Pair fraction, $N_c$, as a function of
460: absolute $B$-band magnitude ($M_{B}$) using $r_{max}=50$ $h^{-1}$
461: kpc for different redshift samples as indicated. For each data
462: point, $N_c$ is calculated using galaxies with absolute $B$-band
463: magnitude between $M_{B}-1$ and $M_{B}+1$. \label{fig2}}
464: \end{figure}
465:
466: \clearpage
467:
468: \begin{figure}[t]
469: \plotone{f3.eps} \caption{Comoving volume merger rate, $N_{mg}$,
470: as a function of redshift $z$. Estimates from the DEEP2 fields
471: using equation (4) are marked with symbols as indicated; the data
472: from Conselice et al. (2003) are represented by the filled
473: triangles. \label{fig3}}
474: \end{figure}
475:
476:
477: \end{document}
478: