1: %%%%%%%% Last modification 04/02/05
2: %% This is a sample manuscript marked up using the
3: %% AASTeX v5.x LaTeX 2e macros.
4: \documentclass[12pt,preprint]{aastex}
5: % \documentclass[manuscript]{aastex}
6: %% preprint2 produces a double-column, single-spaced document:
7: % \documentclass[preprint2]{aastex}
8: %% You can insert a short comment on the title page using the command below.
9: \slugcomment{IPM/P-2003/067} \shorttitle{Possibility of Magnetic
10: Mass detection by the next generation microlensing experiments. }
11: \shortauthors{Sohrab Rahvar \& Farhang Habibi }
12: %% This is the end of the preamble. Indicate the beginning of the
13: %% paper itself with \begin{document}.
14: \begin{document}
15: %% LaTeX will automatically break titles if they run longer than
16: %% one line. However, you may use \\ to force a line break if
17: %% you desire.
18:
19: \title{ Possibility of Magnetic Mass Detection by the Next Generation of Microlensing Experiments}
20:
21: %% Use \author, \affil, and the \and command to format
22: %% author and affiliation information.
23: %% Note that \email has replaced the old \authoremail command
24: %% from AASTeX v4.0. You can use \email to mark an email address
25: %% anywhere in the paper, not just in the front matter.
26: %% As in the title, you can use \\ to force line breaks.
27:
28: \author{Sohrab Rahvar \altaffilmark{1,2} and Farhang Habibi \altaffilmark{3}}
29: \email{rahvar@sharif.edu}
30:
31: %% Notice that each of these authors has alternate affiliations, which
32: %% are identified by the \altaffilmark after each name. Specify alternate
33: %% affiliation information with \altaffiltext, with one command per each
34: %% affiliation.
35:
36: \altaffiltext{1}{Department of Physics, Sharif University of Technology,
37: P.O.Box 11365--9161, Tehran, Iran}
38: \altaffiltext{2}{Institute for Studies in Theoretical Physics and
39: Mathematics, P.O.Box 19395--5531, Tehran, Iran}
40: \altaffiltext{3}{Department of Physics, Iran University of Science
41: and Technology, Narmak, Tehran 16844, Iran}
42: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
43: %% style, abstract will output a Received/Accepted line after the
44: %% title and affiliation information. No date will appear since the author
45: %% does not have this information. The dates will be filled in by the
46: %% editorial office after submission.
47: \begin{abstract}
48: We study the possibility of magnetic mass detection by the
49: gravitational microlensing technique. Recently the theoretical
50: effect of magnetic mass in the NUT space on the microlensing light
51: curve has been studied. It was shown that in the low photometric
52: signal to noise and sampling rate of MACHO experiment light
53: curves, no signature of NUT factor has been found. In order to
54: increase the sensitivity of magnetic mass detection, we propose a
55: systematic search for microlensing events, using the currently
56: running alert systems and complementary telescopes for monitoring
57: the Large Magellanic Clouds stars. In this strategy of
58: observation, we obtain the magnetic mass detection efficiency and
59: also the lowest observable limit of the NUT factor. This method of
60: survey for gravitational microlensing detection can also be used
61: as a tool for searching other exotic space-times.
62: \end{abstract}
63: \keywords{gravitational lensing -- relativity -- cosmology:
64: Observations -- Cosmology: theory --dark matter.}
65: \section{Introduction}
66: The gravitational microlensing method for detecting MAssive
67: Compact Halo Objects (MACHOs) in the Milky Way halo has been
68: proposed by Paczy\'nski (1986). Many groups have contributed to
69: this experiment and have detected hundreds of microlensing
70: candidates in the direction of the galactic bulge, spiral arms and
71: Large and Small Magellanic Clouds (LMC \& SMC). Due to the low
72: probability of the microlensing detection, less than 20 events
73: have been observed by the EROS and MACHO groups in the direction
74: of the Magellanic clouds (Lasserre et al. 2000; Alcock et al.
75: 2000). The low statistics not only causes ambiguities in
76: identifying the galactic model of the Milky Way, but also in some
77: cases the microlensing results
78: are at variance with the results of other observations (Gates and Gyuk 2001).\\
79: Comparing LMC microlensing events with the theoretical galactic
80: models can give us the mean mass of MACHOs and the fraction of
81: halo mass in the form of MACHOs. In the case that we use a Dirac
82: Delta mass function for the MACHOs, the mass of MACHOs in standard
83: halo model is obtained about $\sim 0.5 M_{\sun}$. This means that
84: the initial mass function of MACHO progenitors in the galactic
85: halo should be different from that of the disk, because we neither
86: see the low mass stars which should still exist nor heavier stars
87: that would have exploded in the form of supernova (Adams \&
88: Laughlin 1996; Chabrier, Segretain \& Mera 1996). Another
89: contradiction is that if there were as many white dwarfs in the
90: halo, as suggested by the microlensing experiments, they would
91: increase the abundance of heavy metals via Type I Supernova
92: explosions (Canal., Isern $\&$ Ruiz-Lapuente 1997). Also, recently
93: Green \& Jedamzik (2002) and Rahvar (2004) showed that the
94: observed distribution of duration of microlensing events is not
95: compatible with what is expected from the standard and
96: non-standard halo models. They showed that the observed
97: distribution is significantly narrower
98: compared to what is expected from the galactic models.\\
99: The mentioned problems can be a motivation for establishing the
100: next generation of the microlensing experiments. The new surveys
101: will have the potential to increase the number of microlensing
102: candidates and reduce the ambiguities due to Poisson statistics.
103: The other improvements of the new surveys can be the higher
104: sampling rates and the higher precision photometry of the light
105: curves. More precise light curves will enable us to distinguish
106: the deviations between the standard and non-standard light curves
107: due to parallax or source finite-size effects (Rahvar et al.
108: 2003). In the so-called non-standard microlensing candidates the
109: degeneracy partially can be broken between the lens parameters,
110: such as the distance and the mass of a lens. A better
111: determination of the distance and the mass distributions of the
112: lenses can help us to better identify the Milky
113: Way halo model (Evans 1994).\\
114: Although the mentioned effects are in the background of a
115: Schwarzschild space, it is also be possible that a MACHO which
116: plays the role of the lens, resides in an exotic space-time such
117: as the Kerr or the NUT space. Deviation of the space-time from the
118: Schwarzschild metric causes deviation of the microlensing light
119: curve from the standard one. Thus, studying the microlensing light
120: curves not only can be used to determine the dark matter in the
121: form of MACHOs but also as a unique tool to explore the other
122: exotic space-times as well.\\
123: In the paper by Nouri-Zonoz and Lynden-Bell (1997) the
124: gravitational lensing effect on the light rays passing by a NUT
125: hole has been considered, using the fact that all the geodesics in
126: the NUT space, including the null ones, lie on cones. The
127: extension of this work to the microlensing light curve in the NUT
128: space has been studied by Rahvar and Nouri-Zonoz (2003) and the
129: possible existence of magnetic mass on the light curves of the
130: MACHO group microlensing candidates has been tested. According to
131: the analysis of the light curves, no magnetic mass effect has been
132: found. Although the result showed that the effect of the NUT
133: factor is almost negligible, one can not rule out the existence of
134: NUT charge on that basis. The next generation of microlensing
135: experiments may prove the (non-) existence of magnetic mass
136: through a more careful study of the microlensing light curves.\\
137: Here in this work we simulate the microlensing light curves in the
138: NUT metric according to a strategy for the next generation of
139: microlensing surveys. The aim of this work is to obtain the
140: observational efficiency for the magnetic mass detection and to
141: find the lowest limit for the NUT charge that can be observed.
142: Large Magellanic Cloud (LMC) stars are chosen as the target stars
143: for monitoring. The advantage of using LMC stars as compared to
144: the spiral arms and the galactic bulge stars is the lower
145: contamination by blending and source finite-size effects, which
146: can affect the NUT light curves. The other advantage of LMC
147: monitoring is that it enables us to increase the microlensing
148: statistics to put a better limit on the mass of
149: the lenses and the mass fraction of the galactic halo in form of MACHOs. \\
150: The outline of this paper is as follows. In Section 2, we give a
151: brief account on the microlensing light curve in the NUT metric
152: and compare it with the Schwarzschild case. In Section 3, we
153: introduce the observational strategy and perform a Monte-Carlo
154: simulation to generate the microlensing light curves. Section 4
155: contains the fitting process to the simulated light curves to
156: obtain the observational efficiency of the magnetic mass
157: detection. The results are discussed in Section 5.
158: \section{Gravitational microlensing in Schwarzschild and NUT metrics}
159: The gravitational lensing effect occurs when the impact parameter
160: of a lens with respect to the un-deflected observer-source line of
161: sight is small enough that the deviation of source shape becomes
162: detectable. In the case that of a point like source, the
163: deflection angle is too small to be resolved by the present
164: telescopes. This type of gravitational lensing which amplifies the
165: brightness of the background star is called the gravitational
166: microlensing. In the Schwarzschild metric the magnification is
167: given by (Paczy\'nski 1986):
168: \begin{equation}
169: \label{pac}
170: A(t) = \frac{u(t)^2 +2}{u(t)\sqrt{u(t)^2 + 4}},
171: \end{equation}
172: where $u(t) = \sqrt{u_0^2 + (\frac{t - t_0}{t_E})^2}$ is the
173: impact parameter (position of the source in deflector plane
174: normalized by the Einstein radius, $R_E$) and in which $t_E$ is
175: the Einstein crossing time (duration of event) defined by $t_E =
176: R_E/v_t$, where $v_t$ is the transverse velocity of deflector with
177: respect to the line of sight. The Einstein radius is given by $
178: R_E^2 = \frac{4GMD}{c^2}$, where M is the mass of the deflector
179: and $D = \frac{D_{l}D_{ls}}{D_{s}}$. $D_l$, $D_ls$ and $D_s$ are
180: the observer-lens, lens-source and observer-source distances,
181: respectively. The only physical parameter that can be obtained
182: from a light curve is the duration of the event which is a
183: function of the lens parameters such as mass, the distance of lens
184: from the observer and the relative transverse velocity of
185: the lens with respect to our line of sight.\\
186: In the case of gravitational microlensing, the configuration of
187: the lens changes within the time scales of dozen of days while in
188: the cosmological scales the lensing configuration is almost
189: static. Since the magnification factor depends on the space-time
190: metric, the gravitational microlensing technique may also be a
191: useful tool to explore the other exotic metrics like the NUT
192: space. In the NUT space the magnification due to the microlensing
193: depends in addition, to an extra factor (magnetic mass) compared
194: to the Schwarzschild space. It should be mentioned that the NUT
195: space reduces to the Schwarzschild one when the magnetic mass
196: '${\it l}$' is zero\footnote{NUT space is give by the following
197: metric: $ds^2 = f(r)(dt-2l\cos\theta d\phi)^2 -\frac{1}{f(r)}dr^2
198: -(r^2 +l^2)(d\theta^2 +sin^2\theta d\phi^2)$, where $f(r) = 1 -
199: 2(Mr +l^2)/(r^2 + l^2)$.}. So we expect that the microlensing
200: amplification reduces to Equation (\ref{pac}) for zero magnetic
201: mass. Rahvar and Nouri-Zonoz (2003) obtained the magnification in
202: this space-time as follows:
203: \begin{equation}
204: \label{nut_amplification}
205: A(u) = \frac{2 + u^2}{u\sqrt{4 + u^2}} +
206: \frac{8R^4(2+u^2)}{u^3(4+u^2)^{3/2}} +{\mathcal{O}}(R^8)+ ...,
207: \end{equation}
208: where $R_{NUT}=\sqrt{2lD}$ is defined as the NUT radius (analogous
209: to the Einstein radius) and $l$ is the magnetic mass of the lens.
210: Parameter $R$ in Equation (\ref{nut_amplification}) is defined by
211: dividing NUT radius to the Einstein radius, $R =R_{NUT}/R_E$. It
212: is seen that in the NUT space the magnification factor, like the
213: in the Schwarzschild case, is symmetric with respect to time. The
214: extra second term implies a bigger relative maximum of the
215: magnification factor for a given minimum impact parameter. we have
216: also a shape deviation
217: of the light curve with respect to the case of Schwarzschild metric.\\
218: The detectability of the NUT factor through studying microlensing,
219: depends on the light curves quality (i.e. sampling rate and
220: photometric error bars). In the next section we introduce a new
221: strategy for microlensing observations in order to improve the
222: microlensing light curves, both from the point view of the
223: sampling rate and the photometric precision.
224: \section{Light curves simulation in the NUT space}
225: %In this section we introduce a new strategy for the microlensing
226: %observation to have better light curves from the events.
227: % make better light curves of the ongoing
228: %microlensing events compare to present surveys.
229: %in order
230: %to increase the sensivity to the magnetic mass detection that
231: %accurate photometry measurement and high sampling rate are
232: %necessary.
233: The strategy of the observation is based on using a survey as an
234: alert system for microlensing detection with a follow-up setup.
235: EROS is one of the groups that used an alert system to trigger
236: ongoing microlensing events. We simulate EROS like telescope with
237: the same sampling rate, considering $70\%$ clear sky at {\it La
238: Silla} during the observable seasons of the LMC. A follow-up
239: telescope is considered to observe with one percent photometry
240: precision and sampling rate of at least once per night those
241: events, that have been triggered by the first telescope. Here our
242: aim is to simulate microlensing light curves in the NUT space by
243: using the observational strategy, mentioned above.\\
244: It should be noted that there are at least two other important
245: effects which are called blending and source finite-size effects
246: that can change the light curves in symmetric manner like the NUT
247: factor. Those effects are important because they may dominant over
248: the effect of the NUT factor in the light curves. So, before
249: starting the simulation procedure we give a brief account on those
250: effects and include them in generating microlensing light curves
251: in the NUT
252: space.\\
253: %Those effects are considerable because they may wash the effect of NUT factor in the light
254: %curves.\\
255: %Amongst the possible
256: %targets in the Milky Way such as spiral arms, galactic bulge and
257: %Large and Small Magellanic Clouds, LMC is chosen for this study.
258: The blending effect is due to the mixing of a lensed star and its
259: neighbors lights, which is given as:
260: \begin{equation}
261: F(t) = F_b + A(t)F_s,
262: \end{equation}
263: where $F(t)$ is the measured flux, $F_s$ is the lensed source,
264: $F_b$ is from the vicinity of lensed source and $A(t)$ is the
265: amplification (Wozniak and Paczynski 1997). This effect is
266: described by the blending parameter which is defined as $ b =
267: \frac{F_s}{F_s + F_b}$ and the observed magnification factor can
268: be written as
269: \begin{equation}
270: A_{obs}(t) = 1 + b(A(t) -1).
271: \end{equation}
272: The second altering effect on a light curve in a NUT space is the
273: source finite-size effect which is caused by the non-zero size of
274: projected source star on the lens plane. In this case, different
275: parts of the source star are amplified by different factors. The
276: relevant parameter of this effect is the projected size of the
277: source radius on the lens plane, normalized to the corresponding
278: Einstein radius ($U =\frac{xR}{R_E}$), where $x=\frac{D_l}{D_s}$
279: is the ratio of lens and source distances from the observer and
280: $R$ is the size of the source radius. In the case of close
281: source-lens distance compared to the observer-source
282: distance, this effect becomes important.\\
283: To find the best field of source stars, we compare possible fields
284: of observation such as the galactic bulge, the spiral arms and the
285: Magellanic clouds to find the least blending and source
286: finite-size effects. In the direction of galactic bulge the
287: blending effect is high, since the field of target stars is
288: crowded, except for the clump giants (Popowski et al 2000). For
289: the spiral arms stars the blending effect is less than towards the
290: galactic bulge while the source finite-size effect due to the
291: self-lensing by the spiral arms stars is considerable. For the
292: Small Magellanic Cloud (SMC) according to the blending and
293: parallax studies of long duration event (Palanque-Delabrouille et
294: al 1998), it seems that SMC is quite elongated along our line of
295: sight, with a depth varying from a few kpc (the tidal radius of
296: the SMC is of the order of $4$ kpc) to as much as $20$ kpc. So it
297: is seen that due to high blending and source finite-size effects,
298: SMC is not suitable for searching gravito-magnetic parameters. It
299: seems that LMC is the best choice for this study. The other
300: advantage of using this field is increasing the microlensing
301: statistics which
302: can be used in dark matter studies of the galactic halo.\\
303: Here in our simulation, we use the distribution of the blending
304: factor according to the reconstructed blending parameter that has
305: been obtained by the best fit to the LMC microlensing events. For
306: the source finite-size effects of LMC stars, which become
307: important in the case of itself-lensing, first we compare relative
308: self-lensing abundance as compared to the galactic halo lensing
309: and then evaluate the finite-size effect of
310: those events on the light curves. \\
311: Comparing the optical depth for the standard galactic halo model
312: $\tau_{halo}= 1.2^{+0.4}_{-0.3} \times 10^{-7}$ (Alcock et al.
313: 2000) with the optical depth obtained by the LMC itself
314: $\tau_{self-lensing}=[0.47-7.84]\times 10^{-8}$ (Gyuk, Dalal and
315: Griest 2000), with the mean value of $2.4 \times 10^{-8}$ shows
316: that the expected microlensing events lensed by the halo MACHOs
317: are about one order of magnitude more than the LMC's. The optical
318: depth value of LMC self-lensing can be confirmed by studying the
319: parallax effect on the light curves. Rahvar et al. (2003) showed
320: that for using the same observational strategy that is proposed
321: here, if the self-lensing is dominant, very few lenses (only those
322: which belong to the disk) will produce
323: a detectable parallax effect. \\
324: % In the case of standard halo model
325: %the parallax
326: %events will be dominated.\\
327: In order to evaluate the source finite-size effect on the
328: microlensing light curves of LMC we perform a Monte-Carlo
329: simulation to produce the distribution of the relevant parameter
330: $U$. We use the LMC model introduced by Gyuk, Dalal and Griest
331: (2000) to see the matter distribution in our line of sight. The
332: probability of a microlensing event by a lens at LMC at a given
333: distance from us is
334: $$\frac{d\Gamma(x)}{dx}\propto \sqrt{x(1-x)}\rho(x), $$
335: where $\rho(x)$ is the matter density distribution of LMC. \\
336: The source stars at LMC are chosen according to their
337: color-magnitude distribution. We use the mass-radius relation
338: (Demircan and Kahraman 1990) to evaluate the radius of stars in
339: our simulation. The radius of source stars are projected on the
340: lens plane and normalized to the corresponding Einstein radius to
341: obtain the distribution of $U$s for the LMC self-lensing events.
342: The mean value of $U$ according to our simulation is about
343: $10^{-3}$, which we applied to obtain the gravitational
344: microlensing light curves. For an impact parameter as small as
345: $u_0=0.01$, where the NUT factor becomes important, the maximum
346: magnification difference of a standard light curve and that of
347: obtained by considering source finite-size effect is about one
348: percent. On one hand this difference is less than our photometric
349: accuracy and on the other hand the optical depth due to
350: self-lensing is one order of magnitude smaller as compared to the
351: galactic halo. The conclusion is that the source finite-size
352: effect is not important in our analysis.
353: %In order to avoid finite-size effect in our
354: %simulation we ignore those high amplitude self-lensing events. In
355: %this case the finite-size effect makes the peak of light curves
356: %shallower in opposite to the NUT parameter. Two more parameter
357: %makes the fitting process highly degenerated. Hereafter we only
358: %encounter the blending effect as the contaminator.
359: %the observation of microlensing alerts triggered by EROS with a
360: %follow-up telescope, and estimate the detection efficiency of the
361: %magnetic mass effect.
362: \subsection{Simulation of light curves}
363: The aim of this section is to simulate the microlensing light
364: curves according to the observational strategy that was described
365: before. We use the theoretical light curves to fit the simulated
366: ones and evaluate the magnetic mass parameter of the NUT metric.
367: The final result in this procedure is the observational magnetic
368: mass detection efficiency, which can be applied in different
369: galactic models. To start simulating the light curves,
370: we use a uniform random function to generate the lens parameters.\\
371: The standard microlensing light curve in the Schwarzschild metric
372: depends on 4 parameters, namely the base flux, $u_0$ (minimum
373: impact parameter), $t_e$ (duration of the event) and $t_0$ (the
374: moment of minimum impact parameter or maximum magnification).
375: Taking into account the magnetic mass needs an extra parameter,
376: $R$. The relevant parameters in simulating the light curves are
377: chosen in the following intervals: $u_0\in[0,1]$, $t_0\in[0,2yr],$$t_E\in[5,365]$
378: days and $R\in[0,0.5]$.\\
379: The base fluxes $F_b$ of the background stars in the direction of
380: LMC are chosen according to the magnitude distribution in the EROS
381: catalogs (Lasserre 2000). Since it was shown that the contribution
382: of the blending effect is important in this study, we use the
383: blending distribution that has been obtained from the observed LMC
384: microlensing events in order to use them in generating the light
385: curves (Alcock et al. 2000). The light curves are simulated by
386: using the sampling rate of EROS which is about one observation per
387: six nights in average and is variable during the seasons. The
388: average relative photometric precision $\Delta F/F$ for a given
389: flux $F$ (in ADU unit) is taken from the EROS phenomenological
390: parametrization which has been found for a standard quality image
391: \cite{fre99}.
392: %\begin {equation}
393: % g = n\exp(-b \log(F)),
394: %\frac{\Delta U}{F} = 3.5\times F^{-0.85}.
395: %\end{equation}
396: In simulating the light curves, every photometric measurement is
397: randomly shifted according to a Gaussian distribution that
398: reflects the photometric uncertainties. Since the photometric
399: uncertainty depends on the apparent magnitude of the background
400: stars, the error bars of light curves decrease by increasing the
401: brightness of background source during the lensing, (see Fig.
402: \ref{lc}).
403: \subsection{Simulation of a simple alert system}
404: The next step is to simulate an alert system to trigger the
405: ongoing events and the follow-up observation by the secondary
406: telescope.
407: %which is necessary to trigger follow-up observations.
408: According to one of the EROS alert algorithms, the events will be
409: announced as soon as their light curves exhibit 4 consecutive flux
410: measurements above 4 standard deviations from the base line
411: \cite{man}. It is clear that only the most significant
412: microlensing events are selected by this algorithm. We have in
413: fact considered several trigger thresholds, from a loose criterion
414: (3 consecutive measurements above $3\sigma$ from the base line) to
415: the strict criterion that was finally used. Even using this strict
416: criterion, in average one false alarm due to variable stars or
417: instrumental artifacts is expected per true microlensing alert
418: \cite{JFGprivate}. This false alarm rate will induce some lost
419: follow-up time, but for very limited durations, as it is usually
420: very fast to discard a non-microlensing event.
421: %for the weak lensing we will miss the event.
422: Fig. \ref{lc} shows an example of microlensing light curve that
423: has been simulated, using the specifications of the primary and
424: the secondary follow-up telescopes.\\
425: The efficiency of the alert system depends on the parameters of
426: the lenses. In order to obtain the trigger efficiency in terms of
427: the physical parameters such as the duration of events and $R$, we
428: integrate over the irrelevant parameters such as the minimum
429: impact parameter and the time of maximum magnification. Equation
430: (\ref{nut_amplification}) shows that NUT parameter increases the
431: maximum magnification or in another word decreases the effective
432: minimum impact factor. The result is more trigger rate of
433: microlensing events for those that have larger $R$. This effect is
434: shown in Fig. \ref{trig_proj}. It shows that the trigger
435: efficiency is increased by the long duration of microlensing
436: events, which reflects a bigger probability for the observation of
437: long duration events as compared to short events.
438: \section{ Follow-up telescope and fitting process to the light curves}
439: We use a Monte-Carlo simulation to generate a large number of
440: microlensing events. At the first step the lens parameters are
441: chosen and the light curve is generated according to the primary
442: telescope specification. Using the trigger system, in the case
443: that an event is alerted, the secondary telescope starts its
444: measurements with high sampling rate and photometry precision of the
445: ongoing microlensing event.\\
446: The second telescope is supposed to be a partially dedicated
447: telescope which follows the measurements of alerted events. The
448: telescope is assumed to have about one percent precision in
449: photometry and perform the sampling of events through all the
450: clear nights. According to the Meteorological statistics of the
451: {\it La Silla} observatory about $70$ percent of nights per year
452: are clear. A one-meter telescope could achieve this
453: precision with a long exposure of about $30$ {\it min}. \\
454: After simulating a large number of events by this strategy, we use
455: the NUT and Schwarzschild theoretical microlensing light curves to
456: fit the simulated ones. The least square method is used to fit the
457: theoretical light curves on the data. An example of the fitting
458: routine is shown in Fig. {\ref{lc}. In the case of fitting data
459: with the NUT curve, with $R<0.1$ we encounter the degeneracy
460: problem of fitting. It means that for $R$ close to zero we may
461: obtain from the fitting a non-zero reconstructed value for $R$. To
462: distinguish between the microlensing light curves affected by NUT
463: charge and the standard ones we use the following criterion
464: denoted by $\Delta\chi^2$, to be more than two
465: \begin{equation}
466: \Delta\chi^2 = \frac{\chi_{Sch}^2 -
467: \chi_{NUT}^2}{\chi_{NUT}^2/N_{d.o.f}}\frac{1}{\sqrt{2N_{d.o.f}}},
468: \end{equation}
469: where indices of the $\chi^2$ correspond to the type of the metric
470: and $N_{d.o.f}$ is the number of degree of freedom in the NUT
471: fitting. As complementary criterion in addition to the mentioned
472: one we use the signal to noise ratio of $R$ to be more than two.
473: We obtain the magnetic mass detection efficiency of MACHOs by
474: dividing the reconstructed parameters of those events that pass
475: two mentioned criteria, to the generated events. Fig. \ref{2d_eff}
476: shows two dimensional efficiency of magnetic mass detection in
477: terms of $R$ and the duration of events.
478: %The efficiency is obtained by dividing the population of
479: %reconstructed parameters to the generated ones.
480: %Figure (\ref{2d_eff}) shows two dimensional
481: %detection efficiency of magnetic mass in terms of duration of
482: %events and $R$.
483: The detection efficiency of magnetic mass has direct correlation
484: with $R$ as well as to the duration of the events. It is more
485: practical to have efficiencies in terms of duration of events and
486: $R$ which are shown in Fig.\ref{nut_eff}. It should be mentioned
487: that the blending effect decreases the detection efficiency of
488: magnetic mass, as shown in Fig.\ref{nut_eff}.
489: %The magnetic mass detection
490: %efficiency also in terms of duration or $R$ are shown in Figure
491: %(\ref{nut_eff}).
492: \section{Conclusion}
493: In this work we proposed a new strategy for microlensing
494: observation that not only can be used for searching MACHOs of the
495: galactic halo through observing the LMC stars, but also it can be
496: a useful tool to explore exotic space-times around compact objects
497: such as the NUT metric. As a result of our Monte Carlo simulation,
498: we obtained the detection efficiency for magnetic mass. The
499: minimum value for $R$ that can be observed by this method is about
500: $0.1$. In order to evaluate the amount of detectable magnetic mass
501: {\it l}, we use the relation between the magnetic mass and $R$
502: (Rahvar and Nouri-Zonoz 2003):
503: \begin{equation}
504: R = c\sqrt{\frac{{\it l}}{2GM}}.
505: \label{rl}
506: \end{equation}
507: EROS and MACHO experiments results propose that the mean value of
508: the mass of MACHOs is about $0.5M_{\odot}$ (Alcock et al. 2000;
509: Lasserre et al. 2000). It is worth to mention that this result is
510: obtained in the standard halo model where the mean mass of MACHOs
511: depends on the model that is used for the Milky Way. Assuming
512: standard model for the Milky Way halo, according to equation
513: (\ref{rl}) the minimum observable magnetic mass ${\it l}$ is
514: evaluated to be about $14$ m. Non-existence of magnetic mass
515: signal in the microlensing light curves can also put an upper
516: limit for the value {\it l} $<$ 14 m in the MACHOs of the Milky
517: Way.
518:
519: \acknowledgments The authors thank M. Nouri-Zonoz, H. Hakimi
520: Pajouh and S. Arbabi Bidgoli for their useful comments.
521:
522: %\section{Appendicial material}
523:
524: \begin{thebibliography}{}
525: \bibitem[Adams and Laughlin(1996)]{ada96} Adams, F., Laughlin, G., 1996, APJ, 468, 586.
526: \bibitem[Alcock et al.(2000)]{alc00} Alcock C. et al. (MACHO), 2000, APJ, 542, 281.
527: \bibitem[Canal., Isern and Ruiz-Lapuente(1997)]{can97}
528: Canal, R., Isern, J., Ruiz-Lapuente, P., 1997, APJ, 488, L35.
529: \bibitem[Chabrier., Segretain and Mera(1996)]{cha96}
530: Chabrier, G., Segretain, L., Mera D., 1996, APJ, 468, L21.
531: \bibitem[Derue 1999]{fre99} Derue F., 1999a, Ph.D. thesis, {\sc
532: CNRS/IN2P3}, {\sc LAL99-14 report}.
533: \bibitem[Demircan., Kahraman (1990)]{dem90}
534: Demircan O., Kahraman G., 1990, Ap\&SS, 181, 313
535: \bibitem[Evans 1994]{eva94}
536: Evans N. W., 1994, MNRAS, 267, 333.
537: \bibitem[Gates and Gyuk(2001)]{gat01}
538: Gates I. E., Gyuk G., 2001, APJ, 547, 786.
539: \bibitem[Gyuk., Dalal and Griest 2000]{gyuk}
540: Gyuk G., Dalal N. and Griest K., 2000, APJ, 535, 90.
541: \bibitem[Glicenstein 2002]
542: {JFGprivate}Glicenstein J-F., 2002, private communication.
543: \bibitem[Green and Jedamzik(2002)]{gre02}
544: Green A. M., Jedamzik K., 2002, A\&A 395, 31.
545: \bibitem[Paczy\'nski 1986]{pac86} Paczy\'nski B., 1986, APJ 304, 1.
546: \bibitem[pop00]{pop00}
547: Popowski P. et al. (MACHO), 2000, AAS,
548: 197th AAS Meeting, \#04.17; Bulletin of the American Astronomical Society, Vol. 32, p.1391
549: \bibitem[Lasserre (2000)]{lasth00} Lasserre, T.,2000. PhD. thesis, CNRS/IN2P3, LAL-report 97-19.
550: \bibitem[Lasserre et al.(2000)]{las00} Lasserre, T. et al. (EROS), 2000, A\&A 355, L39.
551: \bibitem[Mansoux 1997]{man}
552: Mansoux B., 1997, Ph.D. thesis, {/sc CNRS/IN2P3, LAL report
553: 97--19}.
554: \bibitem[Nouri-Zonoz and Lynden-Bell 1997]{nou97}
555: Nouri-Zonoz M., Lynden-Bell D., 1997, MNRAS, 292, 714
556: \bibitem[palanque]{pal98}
557: Palanque-Delabrouille et al. (EROS), 1998, A\&A, 332, 1
558: \bibitem[rahvar(2003a)]{rah03a}
559: Rahvar S., Nouri-Zonoz M., 2003, MNRAS, 338, 926
560: \bibitem[Rahvar(2003b)]{rah03b}
561: Rahvar S., Moniez M., Ansari R., Perdereau O., 2003, A\&A, 412, 81
562: \bibitem[Rahvar(2004)]{rah04}
563: Rahvar S., 2004, MNRAS, 347, 213
564: \bibitem[Woz97]{woz97}
565: Wozniak P., Paczynski B., 1997, ApJ, 487, 55
566: \end{thebibliography}
567:
568:
569: %\clearpage
570:
571: %% Use the figure environment and \plotone or \plottwo to include
572: %% figures and captions in your electronic submission.
573:
574:
575: \begin{figure}
576: %\epsscale{0.9}
577: \includegraphics[angle=0,scale=0.72]{f1.eps}
578: %\plotone{light_curve.eps}
579: \caption{ An example of the simulated light curves according to
580: our proposed observational strategy for the next generation
581: microlensing survey. The parameters of the light curve are chosen
582: to be $t_e = 100$ days, $t_0 = 365$ days, $u_0 = 0.3 $, $b=0.88$
583: and $R = 0.5$. The background star is chosen to have an apparent
584: magnitude of 22. The dashed and solid lines show the result of
585: least square fit of the the Schwarzschild and NUT theoretical
586: light curves to the simulated data, respectively. The
587: reconstructed NUT parameter derived from the fitting is $R_{rec}
588: =0.501668 $ with one sigma uncertainty of $0.001946$.
589: $\chi^2/N_{dof}$ for this light curve from the NUT and the
590: Schwarzschild fittings are 0.26 and 30.88.} \label{lc}
591: \end{figure}
592:
593: %\begin{figure}
594: %\plotone{trig_eff.eps} \caption{ This figure shows two dimensional
595: %trigger efficiency in terms of events duration and $R$. It is seen
596: %that for long duration events the chance of alert if hight than
597: %the short term events. \label{trig_eff}}
598: %\end{figure}
599:
600: \begin{figure}
601: \plotone{f2.eps} \caption{The panels from up to down show the
602: trigger efficiency in terms of the blending parameter, duration of
603: events and $R$. The efficiency of the alert system depends on the
604: blending parameter. This means that the bigger blending factor
605: produce a lower maximum magnification. Also for the long duration
606: events, there is a bigger chance to be alerted by the primary
607: telescope. For the case of events with bigger $R$s, the peak of
608: maximum magnification is elevated and the result is a bigger
609: probability for those events to be alerted.} \label{trig_proj}
610: \end{figure}
611:
612: \begin{figure}
613: \plotone{f3.eps} \caption{ These contours show the two dimensional
614: magnetic mass detection efficiency in terms of duration of events
615: and $R$. The numbers between the contours show the amount of
616: detection efficiency.} \label{2d_eff}
617: \end{figure}
618:
619:
620: \begin{figure}
621: %\plottwo{nut_ef_te.eps}{nut_ef_r.eps}
622: \plotone{f4.eps} \caption{The panels from up to down show the
623: magnetic mass detection efficiency in terms of blending parameter,
624: duration of events and $R$. According to the first panel, the
625: magnetic mass effect can be dominated by the blending. The
626: detection efficiency has also direct dependence on the duration of
627: events and $R$. A rough value for the minimum $R$ that can be
628: detected is about $R=0.1$.}
629: \end{figure}
630: \label{nut_eff}
631: \end{document}
632: