astro-ph0011176/ms.tex
1: %MS# by R. Kantowski and R. C. Thomas
2: 
3: \documentclass[preprint]{aastex}
4:  
5: \newcommand{\be}{\begin{equation}} 
6: \newcommand{\ee}{\end{equation}}
7: 
8: \newcommand{\bea}{\begin{eqnarray}} 
9: \newcommand{\eea}{\end{eqnarray}} 
10: \newcommand{\eg}{e.g., } 
11: \newcommand{\ie}{i.e., } 
12: \newcommand{\OM}{\Omega_m} 
13: \newcommand{\OO}{\Omega_0}
14: \newcommand{\OL}{\Omega_{\Lambda}} 
15: 
16: \newcommand{\nn}{\nonumber\\} 
17: 
18: 
19: \newcommand{\bO}{b_{\Omega}}
20: \newcommand{\bb}{\hat b}
21: \newcommand{\mz}{$m$-$z$\ } 
22: 
23: \begin{document}
24: \title{Distance-Redshift in Inhomogeneous $\OO=1$ 
25: Friedmann-Lema\^\i tre-Robertson-Walker Cosmology} 
26: \author{ R.  Kantowski }
27: \affil{ University of Oklahoma, Department of Physics and
28: Astronomy,\\ Norman, OK 73019, USA }
29: \email{kantowski@mail.nhn.ou.edu}
30: 
31: \author{ R. C. Thomas }
32: \affil{ University of Oklahoma, Department of Physics and
33: Astronomy,\\ Norman, OK 73019, USA }
34: \email{thomas@mail.nhn.ou.edu}
35: 
36: \begin{abstract} \vskip .2 truein 
37: Distance--redshift relations 
38: are given in terms of associated Legendre functions 
39: for partially filled beam observations in
40: spatially flat Friedmann-Lema\^\i tre-Robertson-Walker (FLRW) 
41: cosmologies. These models are dynamically pressure-free, flat FLRW on large scales
42: but, due to mass inhomogeneities, differ in their optical properties. The 
43: partially filled beam area-redshift equation is  a Lame$^{\prime}$
44: equation for arbitrary FLRW and is shown to simplify to the associated Legendre
45: equation for the spatially flat, \ie $\OO=1$ case.
46: We fit these new analytic Hubble curves to recent supernovae (SNe) data in an attempt to determine 
47: both the mass parameter $\OM$ and the beam filling parameter $\nu$. We find that current data 
48: are inadequate to limit $\nu$. However, we are able to estimate what limits 
49: are possible when the number of observed SNe is increased by factor of 10 or 100,
50: sample sizes achievable in the near future with the proposed SuperNova Acceleration 
51: Probe satellite.
52: \end{abstract}
53: 
54: \keywords{cosmology:  theory -- large-scale structure of universe}
55: 
56: \section{INTRODUCTION} \label{sec-intro} 
57: Distance-redshift or equivalently the Hubble curve is critical in determining
58: current values of the cosmological parameters $H_0, \OM$, and $\OL$. Conversely,
59: current values of these three parameters determine the large scale 
60: dynamics of the Universe into the distant past. A complication occurs when 
61: attempting to determine these parameters from high $z$ comparisons to the 
62: standard Hubble curve.  The standard Hubble curve is a theoretical quantity 
63: computed assuming all 
64: gravitating matter is homogeneously distributed; whereas, observational data 
65: is taken in the real inhomogeneous Universe. In an inhomogeneous universe
66: an observing light beam is lensed by inhomogeneities located external to,
67: but near the light beam, and defocused (relative to the standard Hubble curve) 
68: by the less than average matter density within the beam. 
69: The simplest way to take into account these effects is to
70: correct all beams for the missing homogeneous matter but correct 
71: for lensing only when necessary. This procedure requires the introduction of
72: one additional parameter, \eg a 
73: filling parameter 
74: $\nu, \ 0\le\nu\le2$ defined by the fraction of inhomogeneous 
75: matter $\rho_I/\rho_0 \equiv\nu(\nu+1)/6 
76: \le 1$ 
77: excluded from observing beams ($\nu=0$ is the standard 100\% filled beam FLRW
78: case and  $\nu=2$ is the empty beam case).
79: When observing high $z$ objects ($z\sim 1$) the reader can think of the parameter $\nu$  as
80: representing matter that exists in galaxies but not in the intergalactic medium. 
81: To find the theoretical Hubble curve for observations in such a universe one must solve the 
82: geometrical optics equation [see \cite{KR98}] given as equation (\ref{Area}) in the next section. 
83: This equation is actually equivalent to the Lame$^{\prime}$ equation 
84: for general FLRW but as pointed out by \cite{KKT} reduces to the 
85: associated Legendre equation (\ref{Legendre}) for the special 
86: case considered here, $\OO=1$.
87: In \S\,\ref{sec-lumdist} we solve this equation using appropriate boundary conditions
88: and give the Hubble curve in terms
89: of associated Legendre functions (eq.[\ref{Pans}]) as well as  
90: in terms of hypergeometric functions (eq.[\ref{2F1ans}]). 
91: In \S\,\ref{sec-fit} we fit this new Hubble curve to data for 60 supernovae (SNe)
92: from the Supernova Cosmological Project (SCP) and from the Cala$^{\prime}$n/Tololo
93: Supernova Survey (CTSS) in an attempt to determine 
94: the mass parameter $\OM$ and the filling parameter $\nu$. 
95: In \S\,\ref{sec-conclusions} we give some concluding remarks.
96: 
97: 
98: \section{The Luminosity Distance-redshift Relation} \label{sec-lumdist} 
99: 
100: For models being discussed here (and for most cosmological models), 
101: angular or apparent size distance 
102: is related to luminosity distance by $D_<(z)=D_{\ell}(z)/(1+z)^2$. 
103: We choose to give luminosity distances in this paper. 
104: The $D_{\ell}(z)$ which accounts
105: for a partially depleted mass density in the observing beam but neglects 
106: lensing by external masses
107: is found by integrating the second order differential equation for the 
108: cross sectional area $A(z)$
109:  of an observing beam from source ($z=z_s$) to observer ($z=0$), see \cite{KR98}
110: for some history of this equation,\footnote{This equation follows from 
111: applying Sach's optics equations [\cite{SR}] to an inhomogeneous FLRW
112: universe and neglecting  shear (external lensing) \cite{KR}. The first version
113: of equation (\ref{Area}) was 
114: given by \cite{Zel} and later \cite{DR74} included the cosmological term.}:
115: \bea 
116: &&(1+z)^3\sqrt{1+\OM z+\OL[(1+z)^{-2}-1]}\times\nonumber\\ &&\hskip 1 in {d\ \over
117: dz}(1+z)^3\sqrt{1+\OM z+ \OL[(1+z)^{-2}-1]}\,{d\ \over dz}\sqrt{A(z)}\nonumber\\ &&\hskip 2.0 in
118: + {(3+\nu)(2-\nu)\over 4}\OM(1+z)^5\sqrt{A(z)}=0.   \label{Area} 
119: \eea
120: The required  boundary conditions are
121: \bea
122: \sqrt{A}|_s&=&0,\nonumber\\ {d\sqrt{A }\over dz}\Big|_s&=& -\sqrt{\delta\Omega} {c\over
123: H_s(1+z_s)}, \label{Aboundary} 
124: \eea 
125:  where $\delta\Omega$ is the solid angle of the beam at the source
126: and 
127: the FLRW value of the Hubble parameter at $z_s$ is
128: related to the current value $H_0$ at $z=0$ by 
129: \be
130: H_s=H_0(1+z_s)\ \sqrt{1+\OM
131: z_s+\OL[(1+z_s)^{-2}-1]}.  \label{Hs} 
132: \ee
133: The luminosity distance is then simply related to the area $A\big|_0$ of the beam 
134: at the observer by
135: \be
136: D_{\ell}^2\equiv {A\big|_0 \over \delta\Omega}(1+z_s)^2.
137: \label{Dl} 
138: \ee
139: Equation (\ref{Area}) can be put into the form of a Heun  equation
140: and its solution has been given in terms of Heun 
141: functions in \cite{KR98}.  Even though the Heun equation is only slightly more 
142: complicated than the hypergeometric
143: equation, \eg it has 4 regular singular points rather than 3, 
144: Heun functions are not yet available in standard libraries. 
145: Consequently, such expressions are not particularly 
146: useful for comparison with data, at this time. 
147: Because the exponents of three of the singular points of the 
148: area equation (in standard Heun form) are 0 and 1/2 
149: [see eq.\,(13) in \cite{KR98}], 
150: equation (\ref{Area}) is actually equivalent to the doubly periodic Lame$^{\prime}$
151: equation. We now show that it reduces to the associated Legendre equation
152: for the spatially flat universe, $\OO=1$. 
153: The required change of dependent and
154: independent variables
155: are respectively
156: \bea
157: P(A,z)&\equiv&(1+z)^{5/4}\sqrt{{A\over\delta\Omega}}\,,\\
158: \label{P}
159: \eta(z)&=&\sqrt{ {1+\OM z(3+3z+z^2)\over \OM(1+z)^3} }\,.
160: \label{eta}
161: \eea
162: The resulting associated Legendre equation is 
163: \be
164: (1-\eta^2){d^2P\over d\eta^2} -2\eta {dP\over d\eta} + 
165: \left(-\left[\frac{1}{6}\right]\left[\frac5{6}\right]-{[(1+2\nu)/6]^2
166: \over 1-\eta^2}\right)\ P = 0\ , 
167: \label{Legendre} 
168: \ee
169: with initial conditions: 
170: \bea
171: P|_s&=&0,\nonumber\\ 
172: {dP \over d\eta}\Big|_s&=&  {c\over
173: H_0}\frac23\ {\sqrt{\OM}\over 1-\OM}\ (1+z_s)^{11/4}. 
174: \label{Pboundary} 
175: \eea 
176: The resulting luminosity distance is then given by
177: \be
178: D_{\ell}(z_s)=(1+z_s)P(\eta(0)).
179: \label{DlP}
180: \ee
181: Expressed as associated Legendre functions equation (\ref{DlP}) becomes
182: \bea
183: &&D_{\ell}(\OM,\OL=1-\OM,\nu;z)= {c\over H_0}{
184: 2\ \Gamma\left(\frac{5-2\nu}6\right)\Gamma\left(\frac{7+2\nu}6\right)(1+z)^{3/4}
185: \over 
186: (1+2\nu)\sqrt{\OM}
187: }\nonumber\\
188: &&
189: \times\Biggl[
190: {\rm P}^{\ (1+2\nu)/6}_{\ -1/6}\left(\sqrt{ {1+\OM z(3+3z+z^2)\over \OM(1+z)^3} }\right)
191: {\rm P}^{-(1+2\nu)/6}_{\ -1/6}\left({1\over \sqrt{\OM}}\right)
192: \nonumber\\
193: &&
194: -
195: {\rm P}^{\ (1+2\nu)/6}_{\ -1/6}\left({1\over \sqrt{\OM}}\right)
196: {\rm P}^{-(1+2\nu)/6}_{\ -1/6}\left(\sqrt{ {1+\OM z(3+3z+z^2)\over \OM(1+z)^3} }\right)
197: \Biggr].
198: \label{Pans}
199: \eea
200: In this expression the associated Legendre functions take on their
201: their analytically 
202: continued values (\ie the arguments are on the real axis and $> 1$).
203: When the filling parameter has values $\nu=0,1,$ or $2$, equation 
204: (\ref{Pans}) reduces respectively to
205: equations (22), (39) and (54), of \cite{KKT}. 
206: Because the associated Legendre equation is a special type of the hypergeometric equation 
207: it is always possible to write associated Legendre functions in terms of hypergeometric functions.
208: And because hypergeometric functions are the more universally available, these results 
209: are the more useful for most parameter values. That 
210: the hypergeometric result existed has independently been seen by \cite{KKT} and \cite{DM}.   
211: That the area equation  reduces to the
212: associated Legendre equation  for another special case ($\Lambda=0$) has been known 
213: for some time, see \cite{KVB} and \cite{SS}. 
214: The appropriate change of variables is
215: \bea
216: h(A,z)&\equiv&(1+z)\sqrt{{A\over\delta\Omega}}=(1+z)^{-1/4}P,\\
217: \label{h}
218: \zeta(z)&=& {\OM \over 1-\OM}(1+z)^3+1 = {\eta^2\over \eta^2-1}.
219: \label{zeta}
220: \eea
221: The resulting hypergeometric equation is 
222: \be
223: (1-\zeta)\zeta\ {d^2h\over d\zeta^2} +\left(\frac12-\frac76\zeta\right) {dh\over d\zeta} + 
224: {(\nu)(\nu+1)\over 36}\ h = 0\ , 
225: \label{hypergeometric} 
226: \ee
227: with initial conditions 
228: \bea
229: h_s&=&0,\nonumber\\ 
230: {dh \over d\zeta}\Big|_s&=&  -{c\over
231: H_s}\ {1-\OM\over 3\OM}\ (1+z_s)^{-2}.
232: \label{Fboundary} 
233: \eea 
234: The resulting luminosity distance is then given by
235: \be
236: D_{\ell}(z_s)=(1+z_s)h(\zeta(0)).
237: \label{DlF}
238: \ee
239: Expressed in terms
240: of hypergeometric functions equation (\ref{DlF}) becomes
241: \bea 
242: &&D_{\ell}(\OM,\OL=1-\OM,\nu;z)= 
243: {c\over H_0}{(1+z)\ 2\over (1+2\nu)\OM^{1/3}}\left[1+\OM z(3+3z+z^2)\right]^{\frac{\nu}6}
244: \nonumber\\
245: &&
246: \times
247: \Biggl\{
248: {}_2F_1\left( -\frac{\nu}6,\frac{3-\nu}6;\frac{5-2\nu}6;{1-\OM\over \left[1+\OM z(3+3z+z^2)\right]}\right) 
249: {}_2F_1\left( \frac{1+\nu}6,\frac{4+\nu}6;\frac{7+2\nu}6;1-\OM\right)
250: \nonumber\\
251: &&
252: -\ \left[1+\OM z(3+3z+z^2)\right]^{-\frac{1+2\nu}6}
253: {}_2F_1\left( -\frac{\nu}6,\frac{3-\nu}6;\frac{5-2\nu}6;1-\OM\right)\times
254: \nonumber\\
255: &&
256: {}_2F_1\left( \frac{1+\nu}6,\frac{4+\nu}6;\frac{7+2\nu}6;
257: {1-\OM\over\left[1+\OM z(3+3z+z^2)\right]}\right)
258: \Biggr\}.
259: \label{2F1ans}
260: \eea
261: 
262: 
263: 
264: \section{Prospects for constraining $\nu$ from high redshift SNe Ia} 
265: \label{sec-fit}
266: Recent cosmic microwave background observations strongly imply a
267: spatially flat universe (\cite{dB}, \cite{RH}), which naturally motivates
268: application of the 
269: new distance formulae presented in \S2 to a set of standard candles
270: to estimate $\nu$.  We use the 60 SNe Ia from the combined
271: Calan/Tololo + Supernova Cosmology Project (CT+SCP) as presented in
272: \cite{RA}, \cite{PS1}.  Combining these data with those from the
273: High-z SN search (\cite{SB}) would bring the total to about
274: 100 SNe.  However, we shall see that this somewhat complicated task
275: would not increase the numbers of SNe enough to noticeably 
276: improve the estimate.
277: 
278: Rather than subject the data to an in-depth Bayesian re-analysis with
279: the additional beam filling parameter $\nu$ included, we merely use a
280: $\chi^2$ goodness-of-fit estimation.  We assume an intrinsic SN absolute
281: magnitude of $M_B=-19.33$ and $H_0=65$ km s$^{-1}$ Mpc$^{-1}$ and follow the
282: same procedure as \cite{WY2000a} to recover results consistent with
283: \cite{PS1} when $\nu=0$, i.e., for observations in a
284: homogeneous universe.  Then we employ the formulae presented here to
285: obtain confidence contours in the $\OM$--\,$\nu$ plane, constraining
286: $\OL$ so that $\OL=1.0-\OM$. By proceeding this way we are assuming that 
287: inhomogeneous matter, e.g., galaxies, are sufficiently removed from the lines of site 
288: of the 60 SNe and that lensing is negligible.
289: 
290: Figure 1 presents the 68, 90 and 99\% confidence contours for the fit
291: (solid, long-dashed and short-dashed lines respectively).  The results 
292: at $\nu=0$ are clearly consistent with the \cite{PS1}
293: findings.  Unfortunately, no value of $\nu$ can be ruled out with this
294: sample because of its size and depth in redshift space.  The best
295: fit overall is at ($\OM=0.31$, $\nu=0.0$). 
296: 
297: A simple-minded way to estimate the number of additional SNe required
298: to rule out any value of $\nu$ at 99\% confidence is to amplify the
299: contribution of each SN to $\chi^2$ by some factor.  This has the
300: effect of simulating more SNe which are exactly like the real ones,
301: and hence the actual best fit will not move, but the contours will
302: shrink.  Figures 2 and 3 are the same contours as in Figure 1, but
303: with sample size increased by factors of 10 and 100 respectively.  
304: We see that by simply enlarging the CT+SCP
305: sample 10 times will allow a $\nu=2$ value to be ruled out at 99\%
306: confidence.  A factor of 100 will allow a much larger range of $\nu$ to be
307: excluded.  These results are quite promising because a
308: sample size of SNe Ia in the thousands extending to even higher
309: redshifts should be possible with 
310: a ground based SNe pencil beam survey (\cite{WY2000b}) or a satellite mission such as
311: the SuperNova Acceleration Probe
312: (SNAP - http://snap.lbl.gov).
313: 
314: The fact that the current data at $z < 1$ do not rule out any value 
315: of $\nu$ is
316: not surprising.  Assuming ($\OM=0.3$, $\OL=0.7$), the
317: increase in distance modulus incurred by increasing $\nu$ from 0 to 2 at
318: $z=0.5$ is only 0.04 magnitudes (see \cite{KR98}).  However, at redshifts of $z=1.0$, 1.5 and
319: 1.7 to be achieved in the future by SNAP, the increases are 0.14, 0.27
320: and 0.32 magnitudes respectively.  Lensing complications will begin to 
321: occur at these higher redshifts and will have to be corrected for and/or selected against
322: \{see \cite{WJ1}, \cite{PP1}, \cite{HD}, \cite{MS}, 
323: \cite{TPN}, and  \cite{WY1999}\}. 
324: Such data are likely to provide limited leverage in determining ($\OM$,
325: $\OL$), unless $\nu$ is properly constrained. Even though the most 
326: prevalent opinion is that $\nu=2$, i.e., there is no significant intergalactic medium, 
327: there may be a sea of massive neutrinos etc., out there that makes $0<\nu<2$. 
328: 
329: As observations reach $z\sim 3$ lensing by
330: galaxies becomes even more important and details of mass
331: inhomogeneities will significantly distort the Hubble curve. Distances given here 
332: are still useful even without exact knowledge of those inhomoheneities. 
333: A lower bound on the primary image magnification at a given redshift 
334: (relative to the mean) is given by 
335: $\mu_{cutoff} =D_{\ell}^2(\OM,\OL,\nu=0;z)/D_{\ell}^2(\OM,\OL,\nu;z).$ 
336: This bound represents sources which, by chance, are not lensed. 
337: This number can be compared to 
338: histograms given in  \cite{WJ2}, \cite{PP2}, \cite{HD}, and \cite{BGGM}.
339: 
340: \section{Conclusions} \label{sec-conclusions}
341: 
342: We have given useful forms for the luminosity distance in the currently
343: relevant inhomogeneous $\OO=1$ FLRW cosmologies.  These cosmologies are 
344: all dynamically FLRW in the large but differ in how gravitating 
345: matter affects optical observations.  A beam filling parameter 
346: $\nu, \ 0\le\nu\le2$ allows the matter to vary from completely 
347: transparent and homogeneous to completely inhomogeneous and exterior to 
348: any observing beam.
349: In order to determine the values of the cosmic parameters ($\OM$,
350: $\OL$) from Hubble curves at high redshift, the value of $\nu$
351: must also be constrained.  For fixed values of ($\OM,
352: \OL$), increasing $\nu$ from 0 (totally homogeneous universe) to
353: 2 (totally clumped) increases the distance moduli of points on the
354: Hubble curve, {\it especially at higher redshifts} as pointed out in \S
355: \ref{sec-fit}.  When
356: observational error is taken into account, the problem of using
357: standard candles at high redshift while ignoring $\nu$ to obtain
358: $\OL$ will become particularly confounding.
359: 
360: The current sample of SNe Ia at $z < 1$ fail to constrain the value of 
361: the beam-filling parameter $\nu$.  Samples 10 to 100 times larger 
362: than the current sample, and in the same redshift
363: range, will constrain $\nu$.  In order to unambiguously determined 
364: ($\OM$,$\OL$) from even higher redshift observations like 
365: those planned in the future, the distance-enhancing effect of $\nu$ must
366: be accounted for in the luminosity distance formulae.
367: 
368: \acknowledgements
369: The authors wish to thank Y. Wang, G. Kalbfleisch, David Branch, and E. Baron
370: for exceedingly helpful comments.
371: 
372: \begin{thebibliography}{}
373: 
374: \bibitem[Bergstr\"om et al.(2000)]{BGGM} Bergstr\"om, L., Goliath, M., Goobar, A., \& M\"ortsell, E   
375: A\&A in press (astro-ph/9912194)
376: 
377: \bibitem[Damianski et al.(2000)]{DM} Damianski, M., de Ritis, R., Marino, A. A., \&
378: Piedipalumbo, E. 2000,  (astro-ph/0004376)
379: 
380: \bibitem[de Bernardis et al.(2000)]{dB} de Bernardis, P.  et al. 2000, Nature, 404, 955
381: 
382: \bibitem[Dyer \& Roeder(1974)]{DR74}  Dyer, C.  C. \& Roeder, R.  C.  1974, \apj, 189, 167 
383: 
384: \bibitem[Holz \& Wald(1998)]{HD}  Holz, D. E. \& Wald, R. M. 1998, \prd, 58, 063501
385: 
386: \bibitem[Kantowski(1969)]{KR}  Kantowski, R.  1969, \apj, 155, 89 
387: 
388: \bibitem[Kantowski et al.(1995)]{KVB}  Kantowski, R., Vaughan, T., \& Branch, D.  1995,
389: \apj, 447, 35
390: 
391: \bibitem[Kantowski(1998)]{KR98}  Kantowski, R.  1998, \apj, 507, 483 
392: 
393: \bibitem[Kantowski et al.(2000)]{KKT}Kantowski, R.,  Kao, J. K., \& Thomas, R. C. 2000, \apj, 545, 549
394: 
395: \bibitem[Metcalf \& Silk(1999)]{MS} Metcalf, R. B. \& Silk, J. 1999, \apj, 519, L1 
396: 
397: \bibitem[Perlmutter et al.(1999)]{PS1}  Perlmutter, S. et al.  1999, \apj, 517, 565
398: 
399: \bibitem[Premadi et al.(1998)]{PP1}  Premadi, P., Martel, H. \& Matzner, R. 1998, \apj, 493, 10
400: 
401: \bibitem[Premadi et al.(2001)]{PP2}  Premadi, P., Martel, H., Matzner, R. \& Futamase, T. 2001, \apjs, in press astro-ph/0101359
402: 
403: \bibitem[Riess et al.(1998)]{RA}  Riess, A. G. et al.  1998, \aj, 116, 1009
404: 
405: \bibitem[Roos \& Harun-or-Rashid(2000)]{RH} Roos, M. \& Harun-or-Rashid, S. M. 2000, (astro-ph/0005541)
406: 
407: \bibitem[Sachs(1961)]{SR}Sachs, R. K. 1961, Proc. R. Soc. London A, 264, 309
408: 
409: \bibitem[Schmidt et al.(1998)]{SB}  Schmidt, B. P. et al. 1998, \apj, 507, 46 
410: 
411: \bibitem[Seitz \& Schneider(1994)]{SS} Seitz, S. \& Schneider, P. 1994, A\&A 287, 349
412: 
413: \bibitem[Tomita et al.(1999)]{TPN} Tomita, K., Premadi, P., \& Nakamura, T. T. 1999, 
414: Prog. Theor. Phys. Supp. 
415: 
416: \bibitem[Wambsganss et al.(1997)]{WJ1}  Wambsganss, J., Cen, R., Xu, G., \& Ostriker, J.
417: P.  1997, \apjl, 475, L81
418: 
419: \bibitem[Wambsganss et al.(1998)]{WJ2}  Wambsganss, J., Cen, R.,  \& Ostriker, J.
420: P.  1998, \apj, 494, 29
421: 
422: \bibitem[Wang(1999)]{WY1999}  Wang, Y. 1999, \apj, 525, 651
423: 
424: \bibitem[Wang(2000a)]{WY2000a}  Wang, Y. 2000, \apj, 536, 531
425: 
426: \bibitem[Wang(2000b)]{WY2000b}  Wang, Y. 2000, \apj, 531, 676
427: 
428: \bibitem[Zel'dovich(1964)]{Zel}  Zel'dovich, Ya. B. 1964, \sovast--AJ, 8, 13
429: \end{thebibliography}
430: 
431: 
432: \clearpage
433: 
434: \figcaption[fig1_3.eps]{The $\OM$--\,$\nu$ plane with 68, 90 and 99\%
435: confidence contours (solid, long-dashed and short-dashed, respectively) 
436: resulting from an attempt to constrain those parameters using the 60 SNe
437: Ia from the Calan/Tololo Supernova Search + Supernova Cosmology Project 
438: sample \cite{PS1}.  The data were first fit assuming $\nu=0$
439: (completely homogeneous universe) to recover a result consistent with
440: the original findings.  Then assuming $\OM = 1- \OL$ the above $\chi^2$
441: grid was calculated and contours of equal $\Delta \chi^2$ above the
442: minimum $\chi^2$ were plotted.
443: \label{fig1_3}}
444: 
445: \figcaption[fig2_3.eps]{Same as Fig.\,1 except 
446: 10 times more SN measurements exactly like the 60 SNe Ia from the 
447: Calan/Tololo + Supernova Cosmology Project sample
448: were used in computing the confidence levels. 
449: \label{fig2_3}}
450: 
451: \figcaption[fig3_3.eps]{Same as Fig.\,2 except 
452: 100 rather than 10 times more SN were assumed to exist. 
453: \label{fig3_3}}
454: 
455: \plotone{fig1_3.eps}
456: \eject
457: \plotone{fig2_3.eps}
458: \eject
459: \plotone{fig3_3.eps}
460: 
461: \end{document}
462: 
463: 
464: