1:
2: %
3: % Manuscript prepared by Hamed Bagherpour & Ronald Kantowski
4: %
5:
6: \documentclass[12pt,preprint]{aastex}
7:
8:
9: \begin{document}
10:
11:
12: \title{Effects of Gravitational Microlensing on P-Cygni Profiles of Type Ia Supernovae}
13:
14:
15: \author{Hamed Bagherpour %
16: \footnote{hamed@nhn.ou.edu %
17: } , David Branch %
18: \footnote{branch@nhn.ou.edu %
19: } , Ronald Kantowski %
20: \footnote{kantowski@nhn.ou.edu %
21: } }
22:
23: \affil{University of Oklahoma, Department of Physics and Astronomy,\\
24: Norman, OK 73019, USA }
25:
26:
27: \begin{abstract}
28: A brief description of the deformed spectra of microlensed SNe Ia
29: is presented. We show that microlensing amplification can have significant
30: effects on line profiles. The resonance-scattering code
31: SYNOW is used to compute the intensity profile in the rest frame
32: of the supernova. The observed (microlensed) spectral lines are predicted
33: assuming a simple stellar-size deflector, and are compared
34: to unlensed cases to show the effects microlensing by solar-size deflectors can have
35: on spectral lines. We limit our work to spherically symmetric deflectors.
36: \end{abstract}
37:
38: \keywords{gravitational lensing --- supernovae: general }
39:
40:
41: \section{Introduction}
42:
43: It is known that amplification of a light source due to microlensing
44: can affect the spectral line profiles if different parts of the
45: profile originate from different emitting regions of the source (e.g., if the emitting
46: region has a ring-like structure) and if the sizes of these regions are comparable to the
47: characteristic lensing radius \citep{KayRefSta86}.
48: Evidence of such effects on line profiles of broad-absorption-line (BAL) quasars was suggested
49: by spectroscopic observations of the multiple images of the strongly lensed quasar
50: H1413+117 \citep{Ang90}. Spectral differences observed between
51: the four images of this quasar were investigated by \citet{Hut93}
52: using the lens model of \citet{ChaRef84}.
53:
54: One expects these `chromatic amplifications' to be observed in a microlensed
55: type Ia supernova as well. SNe Ia are well-studied extended light sources consisting
56: of a central continuum source surrounded by a rapidly
57: expanding atmosphere. The atmosphere accounts for the formation of the observed
58: P-Cygni profiles in the spectral lines of these objects.
59: In $\S$ 2 we show how microlensing by a simple (point-mass) Schwarzschild deflector
60: can deform these P-Cygni line profiles.
61: Results of lensing a source at $z_{s}=1$ by a deflector at $z_{d}=0.05$ are
62: presented in $\S$
63: 3. A flat Friedmann-Lama\^{\i}tre-Robertson-Walker cosmological model
64: with $\Omega _{m}=$ 0.3, $\Omega _{\Lambda }=$ 0.7, and h$_{100}=$ 0.67
65: is assumed to calculate source and deflector distances.
66:
67:
68: \section{Microlensing of a Supernova as an Extended Source}
69:
70: \subsection{Line Profiles of Type Ia Supernovae}
71:
72: P-Cygni profiles are characterized by emission lines together
73: with corresponding blueshifted absorption lines. The latter is produced
74: by material moving away from the source with either relativistic
75: velocities \citep{HutSer90} or nonrelativistic velocities
76: \citep{Bea29}. For an explosive expansion the material in each plane
77: perpendicular to the line of sight has a fixed component of velocity $V_{r}$
78: toward the observer (Fig. 1).
79:
80: Using the normalized,
81: projected radius $P$ (i.e., the photosphere at $P=1$, see Fig. 1), the unlensed flux is
82: defined as
83: \begin{equation}
84: F_{\lambda }=\int _{0}^{2\pi }\int _{0}^{P_{max}}I_{\lambda }(P,\theta )\, P\, dP\, d\theta \: ,
85: \end{equation}
86: where $P_{max}$ is the normalized, projected radius of the supernova and
87: the plane polar angle $\theta $ is measured in the projected source plane.
88:
89: To compute the intensity $I_{\lambda }$, we used a resonance-scattering
90: synthetic spectrum produced with the fast, parameterized supernova
91: synthetic-spectrum SYNOW. This
92: code is often used for making and studying line identifications and initial coarse
93: analysis of the spectra during the photospheric phase of the supernova \citep{Bra03, Bra05}.
94: The code assumes a spherically symmetric and sharp photosphere that
95: emits a blackbody continuum characterized by temperature $T_{bb}$.
96:
97: In SYNOW, the expansion velocity is proportional to radius (homologous
98: expansion: $v=r/t$), as expected for matter coasting at a fixed velocity after
99: an impulsive ejection with a range of velocities. Line formation is
100: treated in the Sobolev approximation \citep{Sob58} and occurs by
101: resonance scattering of photons originating from the photosphere. Line blending
102: is treated in a precise way, within the context of the Sobolev approximation.
103: Each line optical depth $\tau $ is taken to decrease exponentially with radius
104: [See \citet{JefBra90} and \citet{Fis00} for more details].
105: The code does not calculate ionization ratios or rate equations; it
106: takes line identifications to estimate the velocity at the photosphere and
107: the velocity interval within which ions are detected. These quanities give
108: constraints on the composition structure of the ejected matter.
109:
110: SYNOW calculates the intensity emitted from each zone (concentric
111: annuli) of the projected source. These intensity profiles show the
112: absorption features for various $P<1$ as well as emission features for
113: $P>1$. The weighted sum of these intensities over the projected surface of
114: the supernova (eq. [1]) gives the synthetic flux profile.
115: Figure 2 shows the calculated intensity profiles of sodium. These profiles
116: are obtained for optical depths of $\tau = 1$ (upper panel) and $\tau = 1,000$
117: (lower panel). The intensities are calculated for
118: $P>1$, $P=0$, and $P$ just below 1. The asymmetry seen in the emission
119: features is caused by applying the relativistic Sobolev method \citep{Jef93}
120: in the SYNOW code which, in turn, introduces a Doppler boosting in the profiles.
121: Note that to relate the quantities in the comoving and observer frame, one
122: has to use the transformation
123: \begin{equation}
124: I_{\lambda}=I_{\lambda_{o}} \, \frac{\lambda^{5}_{o}}{\lambda^{5}}\: ,
125: \end{equation}
126: where the quantities with subscript `o' denote those
127: measured in the comoving frame \citep{Mih78}.
128:
129:
130: \subsection{Differential Amplification}
131:
132: Microlensing of an extended source such as a type Ia supernova by an isolated compact
133: mass results in two images, both in line with the source and deflector projected on
134: the sky plane. The `primary' image, $r_{p}$, lies on the same side
135: of the deflector as the source while the `secondary' image, $r_{s}$, is on the
136: opposite side. For solar-mass size deflector the angular separation of the two images is
137: of the order of micro arcseconds and as a result, they are seen as
138: a single object as expected
139: for microlensing events \citep{SchEhlFal92}. The
140: apparent brightness of this `single' image differs from that of the
141: unlensed source and is proportional to the apparent area of the image,
142: meaning that the brightness of a source is amplified by a factor
143: \begin{equation}
144: Amp=\frac{A_{o}}{A}\: ,
145: \end{equation}
146: where $A_{o}$ is the area of the image and $A$ is the area of the
147: source both projected on the sky plane. See \citet{Bag05} for details.
148:
149: If the different parts of the spectral profile come from different
150: parts of the source (which is the case for the concentric annuli of
151: an isotropically expanding type Ia supernova), and if the emitting regions are not much bigger
152: than the characteristic lensing radius \citep{KayRefSta86}, one may expect to
153: see not only a rescaling in the observed flux, but also a
154: deformation in the line profiles. For a lensed supernova, the observed
155: line profile becomes
156: \begin{equation}
157: F_{\lambda }=\int _{0}^{2\pi }\int _{0}^{P_{max}}I_{\lambda }(P,\theta )\, Amp(P,\theta )\, P\, dP\, d\theta \: ,
158: \end{equation}
159: where $Amp(P,\theta )$ is the amplification of the surface element
160: centered at point $(P,\theta )$. Assuming that a supernova explodes isotropically,
161: the altered flux becomes
162: \begin{equation}
163: F_{\lambda }=\int _{0}^{P_{max}}I_{\lambda }(P)\, Amp(P)\, P\, dP\: ,
164: \end{equation}
165: where
166: \begin{equation}
167: Amp(P)\equiv \int _{0}^{2\pi }Amp(P,\theta )\, d\theta \: .
168: \end{equation}
169: For an annulus with the width $\delta P$ bounded
170: by inner and outer radii of $P_{-}\equiv P-\frac{\delta P}{2}$ and
171: $P_{+}\equiv P+\frac{\delta P}{2}$, the amplification becomes
172: \begin{equation}
173: Amp(P)=\frac{S_{+}-S_{-}}{\pi \left(P_{+}^{2}-P_{-}^{2}\right)}\: ,
174: \end{equation}
175: where
176: \begin{equation}
177: S_{\pm }=\int _{-\frac{\pi }{2}}^{\frac{\pi }{2}}P_{\pm }\left(P_{\pm }+l\, \sin \varphi \right)\sqrt{1+\frac{4r_{E}^{2}}{l^{2}+P_{\pm }^{2}+2P_{\pm }l\, \sin \varphi }}d\varphi \: .
178: \end{equation}
179: (See \citet{Bag05}). In the above equation, $l$ is the distance between
180: the deflector and supernova, and $r_{E}$ is the radius of Einstein ring. All
181: distances are projected on the deflector plane.
182:
183: The amplification can be calculated using elliptical integrals \citep{WitMao94, MaoWit98}.
184: Figure 3 shows amplification as a function of the expansion velocity
185: of the projected annuli for several values of
186: $l$ in terms of the Einstein ring $r_{E}$. Notice that each curve peaks
187: at the velocity corresponding to the annulus which is partially obscured
188: by the deflector.
189:
190:
191: \section{Lensed Profiles}
192:
193: In this section we calculate the deformed spectral profiles of microlensed
194: SNe Ia. To calculate the differential amplification, we need to specify
195: the distance $l$ between the supernova and the deflector projected
196: on the plane of the deflector (sky plane), normalized by the Einstein
197: ring radius $r_{E}$,
198: \begin{equation}
199: u_{A}\equiv \frac{l}{r_{E}}\, .
200: \end{equation}
201: The radius $r_{E}$ is determined by the mass of the deflector $m_{d}$ and the weighted
202: distance (of luminosity distances) $D$,
203: \begin{equation}
204: r_{E}=\sqrt{\frac{4Gm_{d}D}{c^{2}}}\, ,
205: \end{equation}
206: where $D \equiv D_{ds}D_{d}/D_{s}$ in which, $D_{s}$, $D_{d}$, and $D_{ds}$ are
207: the respective observer-source, observer-deflector, and deflector-source
208: distances. These are angular size distances calculated
209: adopting a $(\Omega _{m},\, \Omega _{\Lambda },\, h)=(0.3,\, 0.7,\, 0.67)$
210: cosmology. We assume that
211: the source is at redshift $z_{s}=1.0$ and is lensed by a deflector
212: located at redshift $z_{d}=0.05$. The point-like deflector has a
213: mass of 1 M$_{\odot }$ and when positioned at different projected distances
214: it magnifies parts of the extended source differently. The Einstein radius $r_{E}$ for
215: the above deflector mass and distances is $\sim1301$ AU.
216:
217: We have considered an SN Ia with fixed radius of 178 AU, corresponding
218: to that of a supernova with a maximum atmospheric speed of 30,000
219: km s$^{-1}$ at eighteen days after the explosion. The radius of the supernova
220: projected on the deflector plane, $r_{SN}$, is $\sim20$ AU, with a photospheric radius of
221: $r_{Ph}\approx8$ AU. The black body continuum temperature $T_{bb}$ is taken to be 14,000 degrees.
222: Optical depths $\tau $ are taken to vary exponentially for each line as
223: \begin{equation}
224: \tau (v)=\tau _{o}\, exp\left(-\frac{v}{v_{e}}\right)\, ,
225: \end{equation}
226: where $v$ is the expansion velocity of each layer and $v_{e}$ is
227: the corresponding e-folding velocity (e.g., 1,000 km s$^{-1}$ for sodium).
228:
229: \setcounter{footnote}{0}
230:
231: We used SYNOW to calculate the unlensed
232: sodium lines as well as their corresponding lensed profile for different
233: optical depths $\tau _{o}$ and normalized distances $u_{A}$ to show
234: the effect microlensing can have on a single, clean line. Figures
235: 4 through 7 show the results of such calculations for $\tau _{o}=1$ and $1,000$,
236: and $u_{A}=0$ and $1/128\ (\approx0.008)$\footnote{We changed $u_{A}$ in terms of
237: negative powers of 2 to let it go to zero assymptotically.}.
238: In these diagrams, flux is plotted in
239: an arbitrary unit as a function of wavelength. As can be seen in Figure 4, the
240: emission feature is reduced with respect to the absorption feature because the $P<1$
241: region is magnified much more than the $P>1$ region. The narrowing of the absorption
242: dip as well as an overall shift of the lensed curve to the left is due to the
243: extreme amplification of the $P=1$ annulus. This is the result of the source-deflector alignment
244: (Fig. 3) which causes the area with
245: the highest blueshift ($P=1$) to get the highest amplification. With $u_{A}=1/128$ (Fig. 5),
246: the emission feature of the apparent line profile is magnified while the absorption
247: feature does not change remarkably because the amplification curve maximizes
248: outside $P<1$ region and flattens inside (Fig. 3). Figure 6 is the same as
249: Figure 4 with $\tau = 1,000$. In this figure, we once again notice the effect of extreme
250: amplification of the central zone of the source in the form of a slight shift
251: of the apparent curve toward lower wavelengths. Contrary to
252: Figure 4, we do not encounter sharp dips here because each dip in the (unlensed) intensity
253: curve is 7,000 km s$^{-1}$ wide. As expected, moving
254: the deflector away from the line of sight to the source results in a stronger emission
255: component while the absorption dip does not vary remarkably (Fig. 7).
256:
257: Figures 8 and 9 show the same calculations as those of Figures 4 to 7 for a
258: SYNOW synthetic spectrum that resembles
259: that of a SN Ia near maximum light with $u_{A}=0$ and $1/128$, respectively.
260: We have normalized the lensed profile at $\lambda = 7,000$ \AA.
261: Because noticeable deformation of the profiles
262: appear only when the deflector is almost aligned with the source (small
263: values of $u_{A}$), we did not include results for $u_{A}>1/128$ (i.e., $l > 10.17$ AU).
264: With $u_{A}=0$ (Fig. 8), central parts of supernova are amplified more than the rest which,
265: as explained for Figures 4 and 6, results in a slight shift toward lower wavelengths.
266: Again, emission features are demagnified
267: with respect to the absorption components. The observed spectral lines show sharp dips
268: because the input value of optical depth $\tau$ for each line is not too high. Figure 9
269: is the same as Figure 8 but with $u_{A}=1/128$. Here, the light coming from the $P>1$
270: area carrying emission features has higher amplification compared to the blueshifted
271: light emitted from $P<1$ and, as expected, the emission component of the P-Cygni
272: features is magnified more than the absorption part.
273:
274: The change in the profiles can, in general, be summarized as a net increase or decrease of
275: the absorption component relative to the emission one. The apparent change in either
276: component may be so strong that an
277: emission feature could look like a typical P-Cygni profile. To see this effect, the
278: projected source (supernova) must be very close to the deflector on the sky plane
279: ($l \ll r_{E}$). Larger impact parameters produce less dramatic deviations from the unlensed
280: profile and could easily be attributed to the intrinsic diversity in spectra of type Ia
281: supernovae rather than gravitational lensing.
282: The deflector redshift used here ($z_{d}=0.05$) is not the most likely redshift for
283: microlensing but results in a remarkable amplification
284: gradient necessary for noticeable deformation of P-Cygni profiles.
285: In general, the probability of microlensing cosmologically distant light source
286: is not negligible, and can exceed 1\% for a source located at $z_{s}=1$ and beyond
287: \citep{Mye95, WyiTur02, ZakPopJon04}.
288: However, for the small values of $u_{A}$ used here
289: the probability of observing such deformations drops below 0.001\% \citep{Bag05}.
290:
291: It should be noted that we have ignored any contribution to lensing from nearby stars and the
292: deflector's parent galaxy. For most microlensing events the amplification due to the host galaxy
293: is expected to introduce a
294: small amplification gradient across the supernova which merely rescales the line profile
295: without introducing any significant deformation effects. Macrolensing by the galaxy should
296: mainly bias the supernovae detection. The rescaling effect is
297: compensated for by normalizing the lensed profile in order to compare the apparent and unlensed
298: profiles, as done in Figures 4 through 9. When large amplification gradients are
299: introduced, either by the galaxy or by microlensing stars nearby, a more complicated lens models will
300: be required.
301:
302:
303: \section{Conclusion}
304:
305: We have shown that microlensing can significantly affect the P-Cygni
306: profile of a cosmologically distant type Ia supernova. We restricted
307: our calculation to $z_{s}=1$ and $z_{d}=0.05$ in the commonly used
308: $\Omega _{m}=0.3$, $\Omega _{\Lambda }=0.7$ flat cosmological model
309: with h$_{100}=0.67$. We found that microlensing can not only increase the
310: flux magnitude but also can cause a change in its line profiles.
311: Microlensing can cause the features in the spectral lines to be blueshifted
312: with respect to the original spectrum and in general, results in a net
313: increase or decrease of the absorption component relative to the emission
314: component.
315:
316: We calculated the deformed line profiles for special cases where the deflector
317: is extremely close to the line of sight to the source.
318: Due to the low probability of microlensing events occuring with such small values of $u_{A}$,
319: a large population of supernovae (around $10^{5}$) would have to be surveyed to observe a
320: single case of deformation in P-Cygni profiles of type Ia SNe. Also, large
321: deformations demand using more complicated lensing models.
322:
323:
324: \acknowledgements{}
325:
326: H. B. wishes to thank Darrin Casebeer for helpful discussions. This work
327: was in part supported by NSF grant AST0204771 and NASA grant NNG04GD36G.
328:
329:
330: \begin{thebibliography}{}
331: \bibitem[Angonin et al.(1990)]{Ang90}Angonin, M., et al. 1990, \aap, 233, L5
332: \bibitem[Bagherpour et al.(2005)]{Bag05}Bagherpour, H., et al. 2005, \apj, submitted
333: \bibitem[Beals(1929)]{Bea29}Beals, C. 1929, \mnras, 90, 202
334: \bibitem[Branch et al.(2003)]{Bra03}Branch, D., et al. 2003, \aj, 126, 1489
335: \bibitem[Branch et al.(2005)]{Bra05}Branch, D., et al. 2005, \pasp, 117. 545
336: \bibitem[Chang \& Refsdal(1984)]{ChaRef84}Chang, K., \& Refsdal, S. 1984, \aap, 132, 168
337: \bibitem[Fisher(2000)]{Fis00}Fisher, A. 2000, PhD thesis, University of Oklahoma
338: \bibitem[Hutsem\'{e}kers \& Surdej(1990)]{HutSer90}Hutsem\'{e}kers, D., \& Surdej, J. 1990, \apj, 361, 367
339: \bibitem[Hutsem\'{e}kers, Surdej, \& Van Drom(1990)]{HutSerVan90}Hutsem\'{e}kers, D., Surdej, J.,
340: \& Van Drom, E. 1992 Lecture Notes in Physics, Gravitational Lenses, ed. R. Kayser et al.
341: (Berlin: Springer-Verlag)
342: \bibitem[Hutsem\'{e}kers(1993)]{Hut93}Hutsem\'{e}kers, D. 1993, \aap, 280, 435
343: \bibitem[Jeffery \& Branch(1990)]{JefBra90}Jeffery, D., \& Branch, D. 1990, Jerusalem Winter School
344: for Theoretical Physics, Vol. 6, ed. J. C. Wheeler, S. Weinberg (Singapore: World Scientific)
345: \bibitem[Jeffery(1993)]{Jef93}Jeffery, D. 1993, \apj, 415, 734
346: \bibitem[Kayser, Refsdal, \& Stabell(1986)]{KayRefSta86}Kayser, R., Refsdal, S., \& Stabell, R. 1986,
347: \aap, 166, 36
348: \bibitem[Mao \& Witt(1998)]{MaoWit98}Mao, S., \& Witt, H. J. 1998, \mnras, 300, 1041
349: \bibitem[Mihalas(1978)]{Mih78}Mihalas, D. 1978, Stellar Atmospheres (San Francisco: Freeman)
350: \bibitem[Myers et al.(1995)]{Mye95}Myers, S. T., et al. 1995, \apj, 447, 15
351: \bibitem[Schneider, Ehlers, \& Falco(1992)]{SchEhlFal92}Schneider, P., Ehlers, J., \& Falco, E. E. 1992,
352: Gravitational Lenses (Berlin: Springer-Verlag)
353: \bibitem[Sobolev(1958)]{Sob58}Sobolev, V. V. 1958, Theoretical Astrophysics, ed. V A. Ambartsumyan
354: (London: Pergamon)
355: \bibitem[Witt \& Mao(1994)]{WitMao94}Witt, H. J., \& Mao, S. 1994, \apj, 430, 505
356: \bibitem[Wyithe \& Turner(2002)]{WyiTur02}Wyithe, J. S. B., Turner, E. L. 2002, \apj, 567, 18
357: \bibitem[Zakharov, Popovi\'{c}, \& Jovanovi\'{c}(2004)]{ZakPopJon04}Zakharov, A. F., Popovi\'{c},
358: L. \v{C}., \& Jovanovi\'{c}, P. 2004, \aap, 420, 881
359: \end{thebibliography}
360:
361:
362: \clearpage
363:
364: \figcaption{}
365: Schematic view of a type Ia supernova as seen
366: by an observer. The normalized radius $P$ and planar speed $V_{r}$ are shown.
367: \label{f1}
368:
369: \figcaption{}
370: Synthetic intensity profile of a type Ia supernova,
371: showing absorption and emission features for $P<1$, $P>1$, and P just below 1
372: for $\tau=1$ (upper panel) and $\tau=1,000$ (lower panel). The emission feature in
373: the upper panel is scaled up.
374: \label{f2}
375:
376: \figcaption{}
377: This figure shows the amplification curves of any point on the projected source
378: as a function of expansion velocity for different values of $l$. The value
379: of photospheric expansion speed determines the zone (absorption or emission)
380: with higher amplification.
381: \label{f3}
382:
383: \figcaption{}
384: Amplified line profile of sodium for $\tau =1$ and $u_{A}=0$. The deflector has a
385: mass of 1M$_{\odot}$.
386: \label{f4}
387:
388: \figcaption{}
389: Same as Fig. 4, with $\tau =1$ and $u_{A}=1/128$.
390: \label{f5}
391:
392: \figcaption{}
393: Same as Fig. 4, with $\tau =1,000$ and $u_{A}=0$.
394: \label{f6}
395:
396: \figcaption{}
397: Same as Fig. 4, with $\tau =1,000$ and $u_{A}=1/128$.
398: \label{f7}
399:
400: \figcaption{}
401: Amplified spectral lines of a SYNOW spectrum that resembles the maximum-light
402: spectrum of a SN Ia, with $u_{A}=0$ and m$_{d}=1$M$_{\odot}$
403: \label{figure8-9}
404:
405: \figcaption{}
406: Same as Fig. 8, with $u_{A}=1/128$ .
407: \label{f9}
408:
409:
410: \clearpage
411:
412:
413: \plotone{f1.eps}
414: \begin{center} Figure 1 \end{center}
415: \eject
416:
417: \epsscale{0.65}
418: \plotone{f2.eps}
419: \begin{center} Figure 2 \end{center}
420: \eject
421:
422: \epsscale{1}
423: \plotone{f3.eps}
424: \begin{center} Figure 3 \end{center}
425: \eject
426:
427: \epsscale{0.65}
428: \plotone{f4.eps}
429: \begin{center} Figure 4 \end{center}
430: \eject
431:
432: \epsscale{0.65}
433: \plotone{f5.eps}
434: \begin{center} Figure 5 \end{center}
435: \eject
436:
437: \epsscale{0.65}
438: \plotone{f6.eps}
439: \begin{center} Figure 6 \end{center}
440: \eject
441:
442: \epsscale{0.65}
443: \plotone{f7.eps}
444: \begin{center} Figure 7 \end{center}
445: \eject
446:
447: \epsscale{1}
448: \plotone{f8.eps}
449: \begin{center} Figure 8 \end{center}
450: \eject
451:
452: \plotone{f9.eps}
453: \begin{center} Figure 9 \end{center}
454: \eject
455:
456:
457: \end{document}
458: