1: \documentclass[preprint]{aastex}
2: \usepackage{psfig}
3: \slugcomment{KSUPT-02/1 \hspace{0.5truecm} July 2002}
4:
5: \newcommand{\omegal}{\Omega_{\Lambda}}
6: \newcommand{\kmsmpc}{{\rm \, km\, s}^{-1}{\rm Mpc}^{-1}}
7:
8: \begin{document}
9:
10: \title{Cosmological Constraints from Compact Radio Source Angular Size
11: versus Redshift Data}
12:
13: \author{Gang Chen and Bharat Ratra}
14:
15: \affil{Department of Physics, Kansas State University, 116 Cardwell Hall,
16: Manhattan, KS 66506.}
17:
18: \begin{abstract}
19: We use the Gurvits, Kellermann, \& Frey compact radio source
20: angular size versus redshift data to place constraints on cosmological
21: model parameters in models with and without a constant or time-variable
22: cosmological constant. The resulting constraints are consistent with
23: but weaker than those determined using current supernova apparent magnitude
24: versus redshift data.
25: \end{abstract}
26:
27: \keywords{cosmology: cosmological parameters---cosmology:
28: observation---large-scale structure of the universe}
29:
30: \section{Introduction}
31:
32: Cosmological models now under consideration have a number of adjustable
33: parameters. A simple way to determine whether a model provides a
34: useful approximation to reality is to use many different cosmological tests
35: to set constraints on cosmological-model-parameter values and to
36: check if these constraints are mutually consistent (see, e.g., Maor et al.
37: 2002; Wasserman 2002).
38:
39: During the past few years much attention has been focussed on the
40: Type Ia supernova apparent magnitude versus redshift test (see, e.g.,
41: Riess et al. 1998; Perlmutter et al. 1999; Podariu \& Ratra 2000;
42: Waga \& Frieman 2000; Gott et al. 2001; Leibundgut 2001).\footnote{
43: The proposed SNAP satellite should provide much tighter constraints
44: on cosmological parameters from this test (see, e.g., Podariu, Nugent,
45: \& Ratra 2001a; Weller \& Albrecht 2002; Wang \& Lovelace 2001; Gerke
46: \& Efstathiou 2002; Eriksson \& Amanullah 2002).}
47: This cosmological test indicates that the energy density of the current
48: universe is dominated by a cosmological constant, $\Lambda$, or a term
49: in the material stress-energy tensor that only varies slowly with time and
50: space and so behaves like $\Lambda$.
51:
52: In conjunction with dynamical estimates which indicate a low non-relativistic
53: matter density parameter $\Omega_0$ (see, e.g., Peebles 1993), cosmic
54: microwave background anisotropy measurements also suggest the presence of
55: $\Lambda$ or a $\Lambda$-like term (see, e.g., Podariu et al. 2001b; Wang,
56: Tegmark, \& Zaldarriaga 2002; Baccigalupi et al. 2002; Durrer,
57: Novosyadlyj, \& Apunevych 2001; Scott et al. 2002; Mason et al. 2002).
58:
59: However, the observed rate of multiple images of radio sources or quasars,
60: produced by gravitational lensing by foreground galaxies, appears to favor
61: a smaller value for $\Lambda$ (see, e.g., Ratra \& Quillen 1992;
62: Helbig et al. 1999; Waga \& Frieman 2000; Ng \& Wiltshire 2001)
63: than is indicated by the observations mentioned above. It is therefore
64: of interest to examine the entrails of other cosmological tests.
65:
66: In this paper we consider the redshift-angular size test, using the
67: Gurvits, Kellermann, \& Frey (1999) compact radio source measurements.
68: The redshift-angular size relation is measured, for structures a few orders
69: of magnitude larger than those considered by Gurvits et al. (1999), by
70: Buchalter et al. (1998) for quasars, and by Guerra, Daly, \& Wan (2000)
71: for radio galaxies; we do not use these data sets in our analysis here.
72: Vishwakarma (2001), Lima \& Alcaniz (2002), and references therein, use
73: the Gurvits et al.(1999) data to set constraints on cosmological parameters;
74: our results are consistent with, but as discussed next extend, their
75: analyzes.\footnote{
76: The Gurvits et al. (1999) data augments that of Kellermann (1993).
77: Stelmach (1994), Stepanas \& Saha (1995), Jackson \& Dodgson (1996),
78: and Kayser (1995) discuss the Kellermann (1993) data.}
79:
80: Cosmological applications of the redshift-angular size test require
81: knowledge of the linear size of the ``standard rod" used. Some earlier
82: analyzes of the Gurvits et al. (1999) data appear to assume that this
83: linear size will
84: be determined using additional data, and so quote limits on cosmological
85: parameters (such as $\Omega_0$ or the cosmological constant density
86: parameter $\omegal$) for a range of values of this linear size. Here
87: we note that it is best to treat this linear size as a
88: ``nuisance" parameter (for the cosmologically relevant part of this
89: test), that is also determined by the redshift-angular size data,
90: and so marginalize over it (using a prior to incorporate other information
91: about it, if needed).\footnote{
92: The situation here is similar to that for the redshift-magnitude test
93: (e.g., Riess et al. 1998; Perlmutter et al. 1999) where one must
94: marginalize over the magnitude of the standard candle used, treating
95: it as a nuisance parameter. In fact, Gurvits et al. (1999) determine the
96: linear size from the redshift-angular size data by using the model of
97: Gurvits (1994).}
98:
99: In $\S$ 2 we summarize our computation. Results are
100: presented and discussed in $\S$ 3. We conclude in $\S$ 4.
101:
102:
103: \section{Computation}
104:
105: For our analyzes here we use the redshift-angular size data of Fig.
106: 10 of Gurvits et al. (1999), which are binned redshift-angular size
107: data derived from measurements of 145 sources. These measurements
108: are combined in twelve redshift bins, with about the same number of
109: sources per bin, with the lowest and highest redshift bins centered
110: at redshifts $z = 0.52$ and $z = 3.6$.
111:
112: We consider two cosmological models as well as a currently popular
113: parametrization of dark energy. These are low-density cold dark
114: matter (CDM) dominated cases, consistent with current observational
115: indications. The first model is parametrized by two ``cosmological"
116: parameters, $\Omega_0$ and $\omegal$ (in addition to all the other usual
117: parameters). This model includes, as special cases, two ``one parameter"
118: models: the currently popular $\Lambda$CDM case with flat spatial
119: hypersurfaces and $\Lambda > 0$ (see, e.g., Peebles 1984; Efstathiou,
120: Sutherland, \& Maddox 1990; Stompor, G\'orski, \& Banday 1995; Ratra
121: et al. 1997; Sahni \& Starobinsky 2000) and a model with open spatial
122: hypersurfaces and no $\Lambda$ (see, e.g., Gott 1982; Ratra \& Peebles
123: 1995; Cole et al. 1997).
124:
125: We also derive constraints on the parameters of a spatially-flat model
126: with a dark energy scalar field ($\phi$) with scalar field potential
127: energy density $V(\phi)$ that at low $z$ is $\propto \phi^{-\alpha}$,
128: $\alpha > 0$. The energy density of the scalar field decreases with
129: time, behaving like a time-variable $\Lambda$ (see, e.g., Peebles \&
130: Ratra 1988; Ratra \& Peebles 1988; Steinhardt 1999; Brax, Martin,
131: \& Riazuelo 2000; Carroll 2001).
132:
133: In linear perturbation theory, a scalar field is
134: mathematically equivalent to a fluid with time-dependent
135: equation of state parameter $w = p/\rho$ and speed of sound squared
136: $c_s^2 = dp/d\rho$, where $p$ is the pressure and $\rho$ the
137: energy density (see, e.g., Ratra 1991). The XCDM parametrization for
138: dark energy approximates $w$ as a constant (see, e.g., Steinhardt 1999;
139: Sahni \& Starobinsky 2000; Carroll 2001; Huterer \& Turner 2001), which
140: is accurate in the radiation and matter dominated epochs but not in
141: the current, dark energy scalar
142: field dominated epoch. Nevertheless the XCDM parametrization is
143: recommended by its simplicity, so we also determine redshift-angular size
144: constraints on its parameters.
145:
146: We want to determine how well the Gurvits et al. (1999) redshift-angular
147: size measurements distinguish between different
148: cosmological-model-parameter values. To do this we pick one of the
149: above models or the XCDM parametrization, and a range of model-parameter
150: values and compute the angular size distance $r (z)$ (the coordinate
151: position for the object considered at redshift $z$ with the observer
152: at the origin) for a grid of model-parameter values that
153: span this range. An object of physical or proper length $l$ transverse to the
154: line of sight subtends an angle $\theta (z) = l (1 + z) / [a_0 r(z)]$,
155: where $a_0$ is the current value of the scale factor (Peebles 1993,
156: $\S$ 13). To determine the probability distribution of the
157: cosmological model parameters ($P$), we compute
158: \begin{equation}
159: \chi^2 (l, P) = \sum_{i = 1}^{12} \left[{{\theta(l, P, z_i) -
160: \theta_{\rm obs} (z_i)} \over \sigma(z_i)} \right]^2 ,
161: \end{equation}
162: where $\theta_{\rm obs} (z_i)$ and $\sigma (z_i)$ are the observed
163: angles and errors for each of the twelve redshift bins centered
164: at redshifts $z_i$ of the Gurvits et al. (1999) data. $P$ represents the
165: model parameters, for instance $\Omega_0$ and $\omegal$ in the general
166: two-dimensional constant $\Lambda$ case. This representation (eq. [1])
167: is exact for the case where the correlated errors between redshift bins
168: are negligible.
169:
170: With a uniform prior for the physical length $l$, the probability
171: distribution (likelihood) of the cosmological model parameters is
172: \begin{equation}
173: L(P) = \int dl \, e^{- \chi^2 (l, P)/2} ,
174: \end{equation}
175: where the integral is over a large enough range of $l$ to include
176: almost all of the probability. We typically integrate $l$ over the range
177: 1 $h^{-1}$ to 60 $h^{-1}$ pc,\footnote{
178: Here the Hubble constant $H_0 = 100 h$ km s$^{-1}$ Mpc$^{-1}$.}
179: sampled at 60 points, although the range depends somewhat on the model
180: under analysis. For the constant $\Lambda$ model we
181: compute over the ranges $ 0 \leq \Omega_0 \leq 1$ and $ -1 \leq \omegal
182: \leq 1$, both sampled at 201 points; for the scalar field dark energy model
183: we compute over the ranges $ 0.005 \leq \Omega_0 \leq 0.995$ and $ 0 \leq
184: \alpha \leq 8$ sampled at 199 and 161 points, respectively; while for the
185: XCDM parametrization we compute over the ranges $ 0 \leq \Omega_0 \leq 1$
186: and $ -1 \leq w \leq 0$, both sampled at 201 points. The confidence limits
187: are computed from the distribution of eq. (2), and we typically show 1, 2,
188: and 3 $\sigma$ confidence contours which correspond to enclosed probabilities
189: of 68.27, 95.45, and 99.73 $\%$, respectively.
190:
191: Since $\Omega_0$ is a positive quantity, we also consider the
192: non-informative or logarithmic prior $p(\Omega_0) \propto 1/\Omega_0$
193: (Berger 1985; Gott et al. 2001) and compute confidence regions for the
194: probability distribution $L(P)/\Omega_0$, where $L(P)$ is given in eq. (2).
195:
196:
197: \section{Results and Discussion}
198:
199: Figure 1 shows the Gurvits et al. (1999) redshift-angular size
200: constraints on the general two-dimensional constant $\Lambda$ case.
201: Apparently these have not previously been published. These constraints
202: are consistent with, but mostly not as constraining as those from Type Ia
203: supernova redshift-magnitude data, except at larger values of $\omegal$ and
204: $\Omega_0$ (see, e.g., Podariu \& Ratra 2000, Fig. 5).
205:
206: Figure 2 shows the redshift-angular size constraints on the XCDM parameters.
207: Lima \& Alcaniz (2002, Fig. 2) show related constraints computed at
208: fixed physical length $l$ for the same Gurvits et al. (1999) data.
209: While the shapes are similar, the Lima \& Alcaniz (2002) contours are much
210: more constraining than those found here, largely because our procedure of
211: marginalizing over the physical length $l$ also accounts for the uncertainty
212: in the determination of $l$. Podariu \& Ratra (2000, Fig. 2) show
213: corresponding constraints on the XCDM parameters from the Type Ia
214: supernova redshift-magnitude data, which are significantly more
215: constraining that those shown in Fig. 2 here.
216:
217: Figure 3 shows the constraints on the dark energy scalar field model
218: with potential energy density $V(\phi) \propto \phi^{-\alpha}$,
219: $\alpha > 0$ (Peebles \& Ratra 1988). They are consistent with, but not as
220: constraining as those from the Type Ia supernova redshift-magnitude data
221: (Podariu \& Ratra 2000; Waga \& Frieman 2000).
222:
223:
224: \section{Conclusion}
225:
226: Constraints on cosmological model parameters derived from the redshift-angular
227: size compact radio source data of Gurvits et al. (1999) are consistent with
228: but less constraining than those derived from the redshift-magnitude Type
229: Ia supernova data of Riess et al. (1998) and Perlmutter et al. (1999).
230:
231: Higher quality redshift-angular size data will more significantly constrain
232: cosmological models, and in combination with high quality redshift-magnitude
233: data will provide a check of conventional general relativity on
234: cosmological length scales.
235:
236: \bigskip
237:
238: We are indebted to L. Gurvits for providing the binned redshift-angular
239: size data. We acknowledge helpful discussions with J. Alcaniz, R. Daly,
240: J. Lima, and J. Peebles, and support from NSF CAREER grant AST-9875031.
241:
242: %\clearpage
243:
244: \begin{thebibliography}{}
245:
246: \bibitem[Baccigalupi et al. (2002)]{bacigalupi02}
247: Baccigalupi, C., Balbi, A., Matarrese, S., Perrotta, F., \& Vittorio, N.
248: 2002, \prd, 65, 063520
249:
250: \bibitem[Berger (1985)]{berger85}
251: Berger, J. O. 1985, Statistical Decision Theory and Bayesian Analysis
252: (New York: Springer-Verlag), 82
253:
254: \bibitem[Brax, Martin, \& Riazuelo (2000)]{brax00b}
255: Brax, P., Martin, J., \& Riazuelo, A. 2000, \prd, 62, 103505
256:
257: \bibitem[Buchalter et al. (1998)]{buchalter98}
258: Buchalter, A., Helfand, D. J., Becker, R. H., \& White, R. L. 1998,
259: \apj, 494, 503
260:
261: \bibitem[Carroll (2001)]{carrolll01}
262: Carroll, S.M. 2001, Living Rev. Relativity, 4, 1
263:
264: \bibitem[Cole et al. (1997)]{cole97}
265: Cole, S., Weinberg, D. H., Frenk, C. S., \& Ratra, B. 1997, \mnras, 289, 37
266:
267: \bibitem[Durrer et al (2001)]{durrer01}
268: Durrer, R., Novosyadlyj, B., \& Apunevych, S. 2001, astro-ph/0111594
269:
270: \bibitem[Efstathiou et al. (1990)]{efstathiou90}
271: Efstathiou, G., Sutherland, W. J., \& Maddox, S. J. 1990, Nature, 348, 705
272:
273: \bibitem[Eriksson \& Amanullah (2002)]{eriksson02}
274: Eriksson, M., \& Amanullah, R. 2002, \prd, 66, 023530
275:
276: \bibitem[Gerke \& Efstathiou (2002)]{gerke02}
277: Gerke, B., \& Efstathiou, G. 2002, \mnras, in press
278:
279: \bibitem[Gott (1982)]{gott82}
280: Gott, J.R. 1982, Nature, 295, 304
281:
282: \bibitem[Gott et al. (2001)]{gott01}
283: Gott, J.R., Vogeley, M.S., Podariu, S., \& Ratra, B. 2001, \apj, 549, 1
284:
285: \bibitem[Guerra et al. (2000)]{guerra00}
286: Guerra, E. J., Daly, R. A., \& Wan, L. 2000, \apj, 544, 659
287:
288: \bibitem[Gurvits (1994)]{gurvits94}
289: Gurvits, L. I. 1994, \apj, 425, 442
290:
291: \bibitem[Gurvits et al. (1999)]{gurvits99}
292: Gurvits, L. I., Kellermann, K. I., \& Frey, S. 1999, A{\&}A, 342, 378
293:
294: \bibitem[Helbig et al. (1999)]{helbig99}
295: Helbig, P., Marlow, D., Quast, R.,
296: Wilkinson, P. N., Browne, I. W. A., \& Koopmans, L. V. E. 1999,
297: A{\&}AS, 136, 297
298:
299: \bibitem[Huterer \& Turner (2001)]{huterer01}
300: Huterer, D., \& Turner, M.S. 2001, \prd, 64, 123527
301:
302: \bibitem[Jackson \& Dodgson (1996)]{jackson96}
303: Jackson, J. C., \& Dodgson, M. 1996, MNRAS, 278, 603
304:
305: \bibitem[Kayser (1995)]{kayser95}
306: Kayser, R. 1995, A{\&}A, 294, L21
307:
308: \bibitem[Kellermann (1993)]{kellerman93}
309: Kellerman, K. I. 1993, Nature, 361, 134
310:
311: \bibitem[Leibundgut (2001)]{leibundgut01}
312: Leibundgut, B. 2001, ARAA, 39, 67
313:
314: \bibitem[Lima \& Alcaniz (2002)]{lima02}
315: Lima, J. A. S., \& Alcaniz, J. S. 2002, \apj, 566, 15
316:
317: \bibitem[Maor et al. (2002)]{maor02}
318: Maor, I., Brustein, R., McMahon, J., \& Steinhardt, P.J. 2002, \prd, 65,
319: 123003
320:
321: \bibitem[Mason et al. (2002)]{mason02}
322: Mason, B. S., et al. 2002, astro-ph/0205384
323:
324: \bibitem[Ng \& Wiltshire (2001)]{ng01}
325: Ng, S.C.C., \& Wiltshire, D.L. 2001, \prd, 63, 023503
326:
327: \bibitem[Peebles (1984)]{peebles84}
328: Peebles, P.J.E. 1984, \apj, 284, 439
329:
330: \bibitem[Peebles (1993)]{peebles93}
331: Peebles, P.J.E. 1993, Principles of Physical Cosmology (Princeton:
332: Princeton University Press)
333:
334: \bibitem[Peebles \& Ratra (1988)]{peebles88}
335: Peebles, P.J.E., \&\ Ratra, B. 1988, ApJ, 325, L17
336:
337: \bibitem[Perlmutter et al. (1999)]{perlmutter99}
338: Perlmutter, S., et al. 1999, \apj, 517, 565
339:
340: \bibitem[Podariu et al. (2001a)]{podariu01a}
341: Podariu, S., Nugent, P., \& Ratra, B. 2001a, \apj, 553, 39
342:
343: \bibitem[Podariu \& Ratra (2000)]{podariu00}
344: Podariu, S., \& Ratra, B. 2000, \apj, 532, 109
345:
346: \bibitem[Podariu et al. (2001b)]{podariu01b}
347: Podariu, S., Souradeep, T., Gott, J. R., Ratra, B., \& Vogeley, M. S.
348: 2001b, \apj, 559, 9
349:
350: \bibitem[Ratra (1991)]{ratra91}
351: Ratra, B. 1991, \prd, 43, 3802
352:
353: \bibitem[Ratra \& Peebles (1988)]{ratra88}
354: Ratra, B., \& Peebles, P.J.E. 1988, \prd, 37, 3406
355:
356: \bibitem[Ratra \& Peebles (1995)]{ratra95}
357: Ratra, B., \& Peebles, P.J.E. 1995, Phys. Rev. D, 52, 1837
358:
359: \bibitem[Ratra \& Quillen (1992)]{ratra92}
360: Ratra, B., \& Quillen, A. 1992, MNRAS, 259, 738
361:
362: \bibitem[Ratra et al. (1997)]{ratra97}
363: Ratra, B., Sugiyama, N., Banday, A. J., \&\ G\'orski, K.M. 1997, ApJ,
364: 481, 22
365:
366: \bibitem[Riess et al. (1998)]{riess98}
367: Riess, A.G., et al. 1998, \aj, 116, 1009
368:
369: \bibitem[Sahni \& Starobinsky (2000)]{sahni00}
370: Sahni, V., \& Starobinsky, A. 2000, Int. J. Mod. Phys. D, 9, 373
371:
372: \bibitem[Scott et al. (2002)]{scott02}
373: Scott, P. F., et al. 2002, astro-ph/0205380
374:
375: \bibitem[Steinhardt (1999)]{steinhardt99}
376: Steinhardt, P.J. 1999, in Proceedings of the Pritzker Symposium
377: on the Status of Inflationary Cosmology, in press
378:
379: \bibitem[Stelmach (1994)]{stelmach94}
380: Stelmach, J. 1994, \apj, 428, 61
381:
382: \bibitem[Stepanas \& Saha (1995)]{stepanas95}
383: Stepanas, P. G., \& Saha, P. 1995, MNRAS, 272, L13
384:
385: \bibitem[Stompor et al. (1995)]{stompor95}
386: Stompor, R., G\'orski, K.M., \& Banday, A. J. 1995, \mnras, 277, 1225
387:
388: \bibitem[Vishwakarma (2001)]{vishwakarma01}
389: Vishwakarma, R.G. 2001, Class. Quant. Grav., 18, 1159
390:
391: \bibitem[Waga \& Frieman (2000)]{waga00}
392: Waga, I., \& Frieman, J.A. 2000, \prd, 62, 043521
393:
394: \bibitem[Wang et al. (2002)]{wang02}
395: Wang, X., Tegmark, M., \& Zaldarriaga, M. 2002, \prd, 65, 123001
396:
397: \bibitem[Wang \& Lovelace (2001)]{wang01}
398: Wang, Y., \& Lovelace, G. 2001, \apj, 562, L115
399:
400: \bibitem[Wasserman (2002)]{wasserman02}
401: Wasserman, I. 2002, astro-ph/0203137
402:
403: \bibitem[Weller \& Albrecht (2002)]{weller02}
404: Weller, J., \& Albrecht, A. 2002, \prd, 65, 103512
405:
406: \end{thebibliography}
407:
408: \begin{figure}[p]
409: \psfig{file=f1.eps,height=7.0in,width=6.7in,angle=0}
410: \caption{Contours of 1, 2, and 3 $\sigma$ confidence for the constant
411: $\Lambda$ model. Solid lines are contours computed using the uniform
412: prior $p(\Omega_0) = 1$, while short dashed lines show the case using
413: the logarithmic prior $p(\Omega_0) = 1/\Omega_0$ (with three contours
414: lying on each other at the left edge). The horizontal dot-dashed
415: line demarcates models with a vanishing cosmological constant, $\Lambda
416: = 0$, the dot-dashed line running from the point $\Omega_0 = 0$,
417: $\omegal = 1$ to the point $\Omega_0 = 2$, $\omegal = -1$ indicates the
418: spatially-flat $\Omega_0 + \omegal = 1$ case, and models lying in the
419: upper left hand corner beyond the dotted line do not have a big bang.}
420: \end{figure}
421:
422: \begin{figure}[p]
423: \psfig{file=f2.eps,height=7.0in,width=6.7in,angle=0}
424: \caption{Contours of 1, 2, and 3 $\sigma$ confidence for the XCDM
425: parametrization of dark energy, $p = w \rho$. Solid lines are contours
426: computed using the uniform prior $p(\Omega_0) = 1$ (with three contours
427: bunched together in the lower left corner), and short dashed lines show
428: the case using the logarithmic prior $p(\Omega_0) = 1/\Omega_0$ (with
429: three contours lying on each other at the left edge).}
430: \end{figure}
431:
432: \begin{figure}[p]
433: \psfig{file=f3.eps,height=7.0in,width=6.7in,angle=0}
434: \caption{Contours of 1, 2, and 3 $\sigma$ confidence for the dark
435: energy scalar field model with inverse power-law potential energy
436: density $V(\phi) \propto \phi^{-\alpha}$. Solid lines are contours
437: computed using the uniform prior $p(\Omega_0) = 1$ (with three contours
438: bunched together in the lower left corner), and short dashed lines show
439: the case using the logarithmic prior $p(\Omega_0) = 1/\Omega_0$ (with
440: three contours lying on each other at the left edge, at the boundary of
441: parameter space).}
442: \end{figure}
443:
444:
445: \end{document}
446:
447:
448:
449: