1: % For ApJ submission -----------------------------------------
2: %\documentclass[12pt,preprint]{/home/cheongho/LatexStyle/AAS5.02/aastex}
3: \documentclass[12pt,preprint]{aastex}
4:
5: %\documentclass{emulateapj}
6: %\usepackage{apjfonts}
7:
8: \lefthead{HAN \& GAUDI}
9: \righthead{DOUBLE-PEAK LENSING EVENTS}
10:
11: %==== CUSTOMIZED LATEX MACROS ========================================
12:
13: \newcommand{\vect}[1]{\ensuremath{\mbox{\boldmath $#1$}}}
14: \newcommand{\zetavec}{\mbox{\boldmath $\zeta$}}
15:
16:
17: \newcommand{\te}{t_{\rm E}}
18: \newcommand{\re}{r_{\rm E}}
19: \newcommand{\rh}{r_{\rm H}}
20: \newcommand{\retilde}{\tilde{r}_{\rm E}}
21: \newcommand{\thetae}{\theta_{\rm E}}
22:
23:
24:
25: % Equation align
26: \def\eqalign#1{\null\,\vcenter{\openup\jot
27: \ialign{\strut\hfil$\displaystyle{##}$&$
28: \displaystyle{{}##}$\hfil \crcr#1\crcr}}\,}
29:
30: %=======================================================================
31:
32:
33: \begin{document}
34: \title{A Characteristic Planetary Feature in Double-Peaked, High-Magnification
35: Microlensing Events}
36:
37:
38: \author{
39: Cheongho Han\altaffilmark{1}
40: and
41: B. Scott Gaudi\altaffilmark{2}
42: }
43:
44:
45: \altaffiltext{1}{Program of Brain Korea 21, Physics Department,
46: Chungbuk National University, Cheongju 361-763, Korea;
47: cheongho@astroph.chungbuk.ac.kr}
48:
49:
50: \altaffiltext{2}{Department of Astronomy, Ohio State University,
51: 140 West 18th Avenue, Columbus, OH 43210-1173;
52: gaudi@astronomy.ohio-state.edu}
53:
54:
55:
56:
57:
58:
59: % ==================================================================
60:
61: %\submitted{Submitted to The Astrophysical Journal}
62:
63: \begin{abstract}
64: A significant fraction of microlensing planets have been discovered in
65: high-magnification events, and a significant fraction of these events
66: exhibit a double-peak structure at their peak. However, very wide or
67: very close binaries can also produce double-peaked high-magnification
68: events, with the same gross properties as those produced by
69: planets. Traditionally, distinguishing between these two
70: interpretations has relied upon detailed modeling, which is both
71: time-consuming and generally does not provide insight into the
72: observable properties that allow discrimination between these two
73: classes of models. We study the morphologies of these
74: two classes of double-peaked high-magnification events, and identify
75: a simple diagnostic that can be used to immediately distinguish
76: between perturbations caused by planetary and binary companions,
77: without detailed modeling. This diagnostic is based on the difference
78: in the shape of the intra-peak region of the light curves. The shape
79: is smooth and concave for binary lensing, while it tends to be either
80: boxy or convex for planetary lensing. In planetary lensing this
81: intra-peak morphology is due to the small, weak cusp of the planetary
82: central caustic located between the two stronger cusps. We apply
83: this diagnostic to five observed double-peaked high-magnification
84: events to infer their underlying nature. A corollary of our study is
85: that good coverage of the intra-peak region of double-peaked
86: high-magnification events is likely to be important for their unique
87: interpretation.
88:
89: \end{abstract}
90:
91: \keywords{gravitational lensing}
92:
93: %{\small {\tt Preprint Version}}
94:
95:
96: %\doublespace
97:
98: \section{Introduction}
99:
100: Microlensing has emerged as an important method of discovering
101: extrasolar planets. Since the first discovery in 2004, six
102: microlensing planets have been reported \citep{bond04, udalski05,
103: beaulieu06, gould06, gaudi07}. The detection rate is rapidly
104: increasing and six additional planet candidates were detected during
105: the 2007 season alone \citep{gould08}. In contrast to the radial velocity
106: and transit methods, which are most sensitive to planets that orbit
107: close to their parent star, the sensitivity of the microlensing method
108: peaks in the cool, outer regions of planetary systems beyond the `snow
109: line' (\citealt{gould92}, see also \citealt{gaudi08}). Furthermore,
110: the sensitivity of the microlensing method extends to very low-mass
111: planets \citep{bennett96}. Thus microlensing is sensitive to planets
112: with physical properties that are very different from those
113: discovered by other methods, and the sample of microlensing planets
114: includes notable planets such as the most distant, the coldest, and
115: the lowest-mass planets detected to date. In addition, the recently
116: reported multiple-planet system OGLE-2006-BLG-109Lb,c \citep{gaudi07}
117: is also noteworthy in that the planet masses and locations relative to
118: the snow line are similar to those of Jupiter and Saturn.
119:
120: Microlensing planet searches are currently conducted using a
121: combination of survey and follow-up observations. The primary
122: microlensing events, caused by stars in the Galactic bulge or
123: foreground disk, are found by survey observations \citep{soszynski01,
124: bond01}, which maximize the event rate by monitoring a large area of
125: the Galactic bulge on a roughly nightly basis. These data are
126: analyzed real time, thereby making it possible to issue alerts of
127: ongoing events in the early stage of lensing magnification. Follow-up
128: observations \citep{yoo04, cassan04} are focused on these alerted
129: events in order to detect the short-lived perturbations to the light
130: curves of the host stars that are the signals of planetary companions
131: to these stars. However, the limited number of telescopes available
132: for follow-up restricts the number of events that can followed at any
133: given time. Thus priority is given to those events which will
134: maximize the planetary detection probability. Currently, the highest
135: priority is given to high-magnification events. There are several
136: reasons for this. First, these events have high intrinsic planet
137: detection efficiency because the source trajectories of these events
138: always pass close to the perturbation region around the central
139: caustic induced by the planet \citep{griest98}. Second, follow-up
140: observations can be prepared in advance because the time of
141: perturbation typically occurs near the peak of the event, which can be
142: predicted reasonably well from data taken on the rise to the peak.
143: Third, the enhanced brightness of the
144: highly-magnified source near the event peak allows for precision
145: photometry, which is essential for the proper characterization of the planetary
146: perturbation. In addition, these bright event peaks can be observed using
147: small-aperture telescopes, which are much more numerous, thus enabling
148: continuous and frequent monitoring, which is also essential for proper
149: characterization. As a result, four
150: (OGLE-2005-BLG-071Lb, OGLE-2005-BLG-169Lb, and OGLE-2006-BLG-109Lb,c) of the six reported
151: microlensing planets were detected through the channel of
152: high-magnification events.
153:
154: A common morphology of perturbed high-magnification events is a
155: double-horned, or double-peaked, structure at the peak of the light
156: curve. A double-peaked morphology at the peak of a high-magnification
157: event can be produced in two very different ways. The first is when
158: the source approaches the blunt, back end of the asymmetric, wedge-shaped, central caustic of a
159: planetary companion at an angle of $\sim 90^\circ$ from the
160: planet/star axis. The second arises when the source approaches the
161: symmetric astroid-shaped caustic of a very wide or very close binary
162: at an angle of $\sim 45^\circ$ from binary axis. The light curves of
163: these two types of events are approximately degenerate in the sense
164: that one can find a value of the shear (for a wide binary lens) or
165: quadrupole moment (for a close binary lens) such that the peak heights
166: and time between the peaks are roughly the same as for the planetary
167: case. Fortunately, as has been demonstrated empirically
168: \citep{albrow02}, they are not perfectly degenerate and thus with
169: sufficient data quality and quantity it is possible to determine
170: whether a light curve is due to a planet or a binary. Distinguishing
171: between the planetary or close/wide binary interpretations of an
172: observed double-peaked high-magnification event has heretofore
173: required detailed modeling (e.g., \citealt{albrow02}).
174:
175: In this paper, we study the morphology of double-peaked high
176: magnification events, and identify a diagnostic feature of the
177: intra-peak region of these light curves that can be used to immediately
178: distinguish between the planetary and binary interpretations.
179: Specifically, the intra-peak region of planetary events is typically
180: boxy or convex, whereas it is smooth and concave for close or wide
181: binary lenses. We provide the physical basis for this difference in
182: the morphology, which is related to the existence of a third, weak
183: cusp in the central caustic of planetary lens, which is absent in the
184: close/wide binary case. While detailed modeling of observed events is
185: ultimately required to derive precise values of the underlying physical
186: parameters, this approach is time-consuming. This diagnostic can be
187: used to quickly identify those events that are most likely caused by
188: planetary companions, and so permit efficient allocation of limited
189: modeling resources. Furthermore, our study provides some insight
190: into the kinds of observations that are needed to discriminate between
191: these two classes of models. This can aid in the planning of
192: observations, and if this diagnostic is applied to events
193: real-time, can inform decisions about which events to follow
194: given limited
195: observational resources.
196:
197: \section{Lensing Properties}
198:
199: For a binary lens, the mapping between the lens plane and source
200: plane can be expressed as
201: \begin{equation}
202: \zeta = z - {m_1/M\over \bar{z}-\bar{z}_{L,1}} -
203: {m_2/M\over \bar{z}-\bar{z}_{L,2}},
204: \label{eq1}
205: \end{equation}
206: where $\zeta=\xi+ i\eta$, $z_{L,j}=z_{L,j}+iy_{L,j}$, and $z=x+iy$
207: are the complex angular positions of the source, lens, and image,
208: respectively, $\bar{z}$ denotes the complex conjugate of $z$, $m_j$
209: are the masses of the individual lens components, and $M=m_1+m_2$ is
210: the total mass \citep{witt90}. Here all angles are normalized to
211: the Einstein radius corresponding to the total mass of the lens system,
212: \begin{equation}
213: \theta_{\rm E}=\left[ {4GM\over c^2} \left({1\over D_L}-{1\over D_S}
214: \right)\right]^{1/2},
215: \label{eq2}
216: \end{equation}
217: where $D_L$ and $D_S$ are the distances to the lens and source,
218: respectively. For a binary lens, there exist three or five
219: images depending on the source position with respect to the positions
220: of the lens components. The magnification of each image is the ratio
221: between the areas of the image and source. This corresponds to the
222: reciprocal of the determinant of the Jacobian of the lens mapping
223: evaluated at the image position, i.e.\
224: \begin{equation}
225: A_i=\left\vert
226: \left(
227: 1-{\partial\zeta\over \partial\bar{z}}
228: {\overline{\partial\zeta}\over \partial\bar{z}}\right)_{z=z_i}^{-1}
229: \right\vert.
230: \label{eq3}
231: \end{equation}
232: Then, the total magnification is the sum of the magnifications of the
233: individual images, $A=\sum_i A_i$.
234:
235: An important characteristic of binary lensing is the existence of
236: caustics. They represent the set of source positions at which the
237: magnification of a point source becomes infinite (i.e.\ where the
238: determinant of the Jacobian is zero), and mark the boundaries of the
239: region in the source position where the number images differs by two.
240: For a binary lens, the caustics form one, two, or three closed curves,
241: interior to which there are five images. Each set of caustics is
242: composed of smooth, concave curves which are fold singularities, which
243: meet at points which are higher-order, cusp singularities.
244:
245:
246: % Figure 1 ----------------------------------------------------
247: \begin{figure}[t]
248: %\epsscale{1.18}
249: \epsscale{0.8}
250: \plotone{f1.eps}
251: \caption{\label{fig:one}
252: Magnification map as a function of the angular position
253: of the source for a planet/star (upper panels) and a wide
254: binary-lens (lower panels) system. The map for the planetary system
255: shows a small region around the planet-hosting star. Angles are
256: in units of the Einstein radius of the planet-hosting star.
257: The planet is located to the left
258: with a separation corresponding to 1.3 times of the star's Einstein
259: radius and the planet/star mass ratio is $10^{-3}$. The map of the
260: binary system shows the region in the vicinity of the primary
261: lens of the two, equal-mass, binary-lens components.
262: Angles are in units of the angular Einstein radius of the primary lens.
263: The other, secondary lens component is located to the left with a
264: separation corresponding to 26.9 times the Einstein radius of one component.
265: In all panels, the coordinates are aligned such that $\xi$ and $\eta$ axes are parallel
266: with and normal to the line connecting the two lens components,
267: respectively. The closed figures drawn in thick black curves
268: represent the caustics. The grey-scale is drawn such that brighter
269: tone represents the region of higher magnification. The straight
270: lines with arrows show three different source trajectories leading
271: to the double-peaked light curves presented in Fig.~\ref{fig:two}.
272: The maps on the right panels show blowups of the regions enclosed
273: by the boxes in the left panels.
274: }\end{figure}
275: % -------------------------------------------------------------
276:
277:
278:
279: % Figure 2 ----------------------------------------------------
280: \begin{figure}[t]
281: %\epsscale{1.18}
282: \epsscale{0.8}
283: \plotone{f2.eps}
284: \caption{\label{fig:two}
285: Examples of double-peaked high-magnification light curves. The geometry
286: of the lens systems and the source trajectories responsible for the
287: light curves are presented in Fig.~\ref{fig:one}. In each panel, blue and
288: red curves are the light curves for the planetary and binary events,
289: respectively, whereas the black curve is the light curve for a single lens with
290: the same mass as the primary lens (i.e., no planetary or binary companion).
291: The time is relative to $t_0$, the time of closest approach to the origin
292: of the lens system, and normalized to $t_{\rm E}$, the Einstein time scale
293: of the primary lens.
294: }\end{figure}
295: % -------------------------------------------------------------
296:
297:
298: \subsection{Planetary Lensing}
299:
300: Planetary lenses correspond to an extreme case of the binary lens
301: where the mass of one of the lenses is much smaller than the other.
302: In this case, the lens equation can be rewritten in a somewhat
303: more intuitive form,
304: \begin{equation}
305: \zeta = z -{1\over \bar{z}} - {q\over \bar{z}-\bar{z}_p}.
306: \label{eq4}
307: \end{equation}
308: Here the angular coordinates are centered at the position of the planet-hosting
309: star, $z_p$ represents the location of the planet, $q$ is the
310: planet/star mass ratio, and the angular positions are now normalized to
311: the Einstein radius of the primary mass.
312: In the case of $q\ll 1$, the analysis of the lensing
313: behavior is amenable to a perturbative approach, which yields considerable
314: insight into the behavior of the caustics and light curves as a function of
315: the planetary parameters \citep{dominik99,bozza99,
316: asada02, an05}.
317:
318: For the planetary case, unless the $|z_p| \sim 1$, there exist two sets of disconnected caustics.
319: One set, which can consist of one or two closed caustic curves, is located away from the host star.
320: This set is typically referred to as the planetary
321: caustic (or caustics). The location of the planetary caustic relative to the source trajectory
322: depends on the separation between the planet and star, as well as
323: the angle between the planet/star axis and the direction of motion of the source. Thus perturbations
324: due to planetary caustics are not predictable.
325:
326: In contrast to the planetary caustic or caustics, the other caustic is always located
327: close to the host star, and so is known as the central caustic. Thus central
328: caustic perturbations always occur at the peak of high-magnification events.
329: The central caustic has a
330: wedge-like shape with {\it four} cusps (see Figure~\ref{fig:one}). One cusp is located
331: on the star-planet axis and corresponds to the point of the wedge. This cusp is
332: strong, in the sense that light curves from source trajectories that pass reasonably close
333: to such a cusp will exhibit strong deviations from the single-lens
334: expectation. Two of the cusps are located off the axis on the opposite
335: side of the caustic, and define the `blunt' end of wedge-shaped caustic.
336: These two cusps are also strong. Between these cusps
337: is region of significant demagnification relative to the single-lens expectation.
338: The fourth cusp, which is located between these cusps on the planet-star axis, is weak,
339: in the sense that it creates relatively weak positive deviations. See Figure~\ref{fig:one}.
340: Because of this wedge-shaped geometry, central caustic perturbations typically
341: occur when the source passes close to either the point or the blunt end of the wedge on a
342: trajectory that is approximately perpendicular to the
343: planet/star axis. In the case that it passes the blunt end, the resulting light curve is
344: double peaked. In fact, both of the planetary microlensing
345: events arising from central caustic perturbations have been double-peaked, suggesting
346: that this class of events might be quite common.
347:
348: The size of the central caustic depends on both the star-planet
349: separation and planet/star mass ratio. When the size is measured as
350: the separation between the two on-axis cusps, it is related to the
351: separation and mass ratio by
352: \begin{equation}
353: \Delta\xi_c \simeq {4q\over (s-s^{-1})^2},
354: \label{eq5}
355: \end{equation}
356: where the separation $s$ is expressed in units of the Einstein radius.
357: Unlike the size, which depends on both $s$ and $q$, for $q\ll 1$ the shape of the
358: caustic is solely dependent on $s$ and it becomes more elongated as
359: $s\rightarrow 1$. For a given mass ratio, a pair of central caustics
360: with separations $s$ and $s^{-1}$ are identical to first order in $q$.
361: For more details about the properties of central caustics, see \citet{chung05}.
362:
363:
364: In the upper panel of Figure~\ref{fig:one}, we present the central
365: caustic of an example planetary lens system and the magnification
366: pattern around the caustic. The planet has a mass ratio $q=10^{-3}$
367: and it is located on the left side of the host star with a separation
368: $s=1.3$. The coordinates are centered at the location of the host star
369: and the axes are aligned such that $\xi$ and $\eta$ axes are parallel
370: with and perpendicular to the star-planet axis, respectively. All angular
371: positions are normalized in units of the Einstein radius corresponding to the
372: mass of planet-hosting star. The grey-scale is drawn such
373: that brighter tone represents the region of higher magnification. The
374: straight lines with arrows are example source trajectories producing
375: double peaked events where the light curves of the resulting events
376: are presented in Figure~\ref{fig:two} (blue curves). The source
377: trajectories have a common impact parameter from the primary star but the
378: angles with respect to the star-planet axis (source trajectory angle
379: $\alpha$) are different. For the planetary case, the two peaks have
380: a similar height when the source trajectory angle is $\alpha\sim
381: 90^\circ$ and the difference in heights increases as the angle
382: deviates from this angle. The map in the upper right panel shows
383: the blowup of the region enclosed by a box in the left-side map.
384:
385:
386: \subsection{Wide/Close Binary Lensing}
387:
388: In the limiting case of a binary lens where the projected separation
389: between the lens components is much larger than the Einstein radius
390: ($s\gg 1.0$), the lensing behavior in the vicinity of one of the
391: lens components (the primary) can be approximated by a Chang-Refsdal lens
392: \citep{chang79, chang84, dominik99}, i.e.
393: \begin{equation}
394: \hat{\zeta}=\hat{z}-{1\over \hat{z}} + \gamma \hat{z}.
395: \label{eq6}
396: \end{equation}
397: Here the notations with `hat' represent angular scales normalized
398: by the Einstein radius of the primary.
399: The quantity $\gamma$ represents the shear induced by the other
400: binary component (companion) and it is related to the lens parameters
401: by
402: \begin{equation}
403: \gamma = {q\over \hat{s}^2},
404: \label{eq7}
405: \end{equation}
406: where $q=m_2/m_1$ is the companion/primary mass ratio, $m_1$ and
407: $m_2$ are the masses of the primary and companion, respectively,
408: and $\hat{s}$ is the separation between the primary and
409: the companion in units of the
410: Einstein radius of the primary, which is related to the
411: separation in units the Einstein radius of the total mass of the binary by $\hat{s}=(1+q)^{1/2}s$.
412:
413: The shear exerted by the companion
414: results in the formation of a small caustic near
415: the location of the primary
416: lens. In the Chang-Refsdal limit, the caustic has a shape of hypocycloid with four cusps (an astroid)
417: regardless of the binary separation and mass ratio. Two of the cusps
418: are located on the binary-lens axis, and the other two are along a line perpendicular
419: to the axis. All of these cusps are of equal strength. Thus a source trajectory
420: that passes close to the caustic on a trajectory with an orientation of $\sim 45^\circ$
421: with respect to the binary axis will produce a double-peak event with roughly
422: equal peak heights.
423: The size of
424: the caustic as measured by the separation between the two on-axis cusps is,
425: \begin{equation}
426: \Delta\xi_c \simeq 4\gamma,
427: \label{eq8}
428: \end{equation}
429: and thus $\Delta\xi_c \propto q$ and $\Delta\xi_c \propto \hat{s}^{-2}$.
430:
431: For a close binary with $s\ll 1.0$, the caustic and the magnification
432: pattern around it are approximately identical to those of the wide
433: binary with a separation of $s^{-1}$ except that the caustic is
434: located at the center of mass of the binary. In this case, the
435: size of the caustic is set by the quadrupole moment of the binary.
436:
437: In the lower left panels of Figure~\ref{fig:one}, we present the
438: magnification pattern in the vicinity of the primary of a wide binary
439: lens. The companion is located on the left side with a separation of
440: $\hat{s}=26.9$ and the companion/primary mass ratio is 1.0. In the
441: map, we also mark several example source trajectories resulting in
442: double peaked high-magnification events, where the light curves are
443: presented in Figure~\ref{fig:two} (red curves). These
444: trajectories have $\alpha\sim 45^\circ$, and
445: thus the two peaks have a similar height. The notations are same as those of the
446: maps of the planetary lens.
447:
448: \section{Difference in Magnification Pattern}
449:
450: The perturbations from the single-lens form exhibited in double-peaked
451: high-magnification events can be characterized by three gross
452: observables: the height of each peak and the time between the two peaks.
453: Given an observed double-peaked high-magnification event, it is
454: always possible to find a planetary or binary-lens model that can
455: reproduce these observables. In particular, these three observables
456: can be matched by varying the following three parameters: (1) the angle of the source
457: trajectory relative to the binary lens axis, which sets the relative
458: peak heights, (2) the impact parameter from the primary lens\footnote{It
459: is not strictly true that it is possible to vary the impact parameter
460: arbitrarily, as this parameter is constrained by the light curve
461: data away from the peak. However, for high-magnification events in the usual
462: highly-blended case, the impact parameter is poorly constrained, and thus
463: our discussion is approximately correct.}, which sets
464: the average height of the two peaks, and (3) either the shear (in the
465: case of a wide binary), the quadrupole moment (in the case of a
466: close binary), or the parameter combination $q (s-s^{-1})^{-2}$ (in
467: the case of the planetary lens), which set the size of the caustic and so
468: the time between the peaks.
469:
470: Although it is possible to match these three gross observables with either a planetary or
471: wide/close binary lens, the morphology of these two classes of
472: double-peaked high-magnification events are not identical, as
473: illustrated in Figure~\ref{fig:two}. The most noticeable difference
474: arises in the shape of the intra-peak, trough region. In the case of
475: a close/wide binary lens, this region has a smooth, rounded, concave
476: shape. On the other hand, in the case of the planetary lens curve,
477: this region has a boxy, slightly convex morphology.
478:
479: The morphology of the intra-peak trough region in the planetary case
480: is caused by the existence of the fourth cusp in the central caustic
481: located in between the two stronger cusps (see the blowup of the
482: planetary central caustic in the upper right panel of
483: Fig.~\ref{fig:one}). The general pattern of magnification around a
484: caustic is a lobe of positive perturbation in the region
485: immediately surrounding the cusp, flanked on the side of the
486: cusp by a more extended region of relative demagnification.
487: For the planetary case, the weak fourth cusp
488: results in either a boxy intra-trough region, or even a slight
489: convexity, caused by the lobe of positive perturbation associated with
490: the weak cusp `filling in' the trough created by the two neighboring,
491: stronger cusps. Since the cusp is weak, this bump is generally weak,
492: but its effect on the morphology of the light curve is generally not
493: negligible. For close/wide binary lenses, on the other hand, there is
494: no weak middle cusp. Thus the intra-peak morphology for a binary lens
495: caustic is characterized by the double peaks that occur when the
496: source approaches the strong cusps of the astroid Chang-Refsdal
497: caustic, and a smooth, concave intra-peak trough that occurs as the
498: source passes through the negative perturbation region between the two
499: cusps.
500:
501: The impact of the weak middle cusp on the planetary lensing light
502: curve morphology varies depending on the underlying parameters. Two
503: factors affect the shape of the feature. The first is the
504: overall shape of the central caustic. This can be seen in
505: Figure~\ref{fig:three}, where we present magnification patterns around
506: central caustics of various shapes, as well as example light curves. In order
507: to isolate the effect of the caustic shape on the morphology, we have
508: adjusted the parameters so that the caustics have a similar size
509: (i.e., so that the parameter combination $q (s-s^{-1})^{-2}$ is
510: constant). The shape of the caustic depends on the star-planet
511: separation $s$, and thus the shape of the intra-peak feature depends
512: on the planetary separation. The variation is such that the cusp
513: responsible for the intra-peak feature is stronger and thus the
514: intra-peak feature is more prominent for planet located further from
515: the Einstein radius of the central star. The other factor that
516: affects the shape of the intra-peak feature is the impact parameter of
517: the source trajectory. As the source trajectory passes closer to the
518: weak cusp, the resulting feature becomes more obvious. The variation
519: of the intra-peak feature with the impact parameter is shown in
520: the second panels of Figure~\ref{fig:three}.
521:
522: We note that although the intra-peak region shows the most obvious
523: morphological differences between the close/wide binary and planetary
524: light curves, there are other, somewhat more subtle differences which
525: may also be used to distinguish between these two interpretations. In
526: particular, the detailed shapes of the peaks appear to differ. In the
527: planetary case, the lobes of high magnification due to the strong
528: cusps are asymmetric about the symmetry axes of the cusps, whereas for
529: the binary-lens case the magnification patterns near the cusps are
530: nearly perfectly symmetric about the symmetry axes.
531: Thus, even in the case of a poorly sampled intra-peak region, or when
532: $s \rightarrow 1$ in the planetary lens case such that the middle
533: cusp is very weak, detailed information about the shape of the peaks
534: will allow one to distinguish between the two models.
535:
536: An important corollary to our study is that one expects that
537: high-magnification, double-peaked events in which the intra-peak region
538: is poorly sampled to be subject to more severe degeneracies, such
539: that the span of allowed models is larger. Thus good coverage of the intra-peak
540: region is likely
541: to be important for the unique interpretation of these events.
542:
543:
544: % Figure 3 ----------------------------------------------------
545: \begin{figure}[t]
546: %\epsscale{1.18}
547: \epsscale{0.8}
548: \plotone{f3.eps}
549: \caption{\label{fig:three}
550: Variation of the intra-peak morphology of double-peaked high-magnification
551: planetary microlensing events. The left panels show the magnification
552: patterns around central caustics of various shapes. In each panel,
553: the white lines indicate example trajectories, with the
554: corresponding light curves shown in the right panels. In the second row,
555: several trajectories are shown to illustrate the effect
556: the varying the impact parameter of the source
557: trajectory on the intra-peak morphology.
558: Notations for the magnification pattern maps are same as
559: in Fig.~\ref{fig:one}. The values marked in each panel represent the
560: star-planet separation in units of the Einstein radius ($s$) and the
561: planet/star mass ratio ($q$).
562: }\end{figure}
563: % -------------------------------------------------------------
564:
565:
566:
567: \section{Application to Observed Events}
568:
569: The diagnostic feature we have identified is useful as it
570: can be used to distinguish between planetary and close/wide binary
571: interpretations of observed events, without the need for detailed
572: fitting. Thus it can be used to quickly identify those events
573: which are most likely due to planetary companions. Motivated by this,
574: we apply our diagnostic to five high-magnification, double-peaked events
575: with reasonably well-covered peaks
576: \footnote{We do not consider events that are technically
577: double-peaked, high magnification events, but are clearly not produced
578: by the two classes of models we have considered in this paper. An
579: example is MOA-2002-BLG-33, which has a maximum magnification of
580: $\sim 450$ and exhibits a double-peaked morphology, but is clearly
581: caused by a geometry in which the source trajectory crosses a caustic with a size of order
582: the source size \citep{abe03}.}. For
583: two of these events, MACHO 99-BLG-47 \citep{albrow02} and
584: OGLE-2005-BLG-071 \citep{udalski05,dong08}, detailed modeling has
585: already been done. Three additional events were observed during 2007
586: lensing season, for which detailed modeling has not been reported.
587: These events are OGLE-2007-BLG-349/MOA-2007-BLG-379, OGLE-2007-BLG-514,
588: and OGLE-2007-BLG-137/MOA-2007-BLG-091.
589: \begin{enumerate}
590: \item {MACHO 99-BLG-47}: This is the first published high-magnification
591: event with two well-resolved peaks. The trough between the two peaks
592: of this event exhibits a smooth concave shape suggesting that the lens is a
593: wide/close binary, and not a planetary system. The large difference between
594: the heights of the two peaks implies that the source trajectory angle is
595: considerably different from $45^\circ$. This diagnostic matches the
596: results from the detailed analysis of the event conducted by \citet{albrow02}.
597: In this model, the source trajectory angle was estimated to be $\alpha\sim
598: 25^\circ$.
599: \item {OGLE-2005-BLG-071}: The intra-peak trough of this event exhibits
600: a prominent convex feature, implying that the perturbation is caused by
601: a planet. In addition, the two peaks are of almost equal height, implying
602: that the source passes nearly perpendicular to the binary axis. This
603: diagnostic matches the results obtained from the detailed modeling
604: conducted by \citet{udalski05} and \citet{dong08}.
605: \item {OGLE-2007-BLG-349/MOA-2007-BLG-379}: This event was detected
606: during 2007 season and its peak was densely covered by follow-up
607: observations. The two peaks have moderately different heights and the
608: trough between the peaks has a linear structure. These characteristics
609: are similar to the planetary lensing light curve presented in the middle
610: panel of Figure~\ref{fig:two} except with the direction of time
611: reversed. Our diagnostic would therefore indicate that this event
612: is caused by a planet with a source trajectory angle somewhat different
613: from $\alpha=90^\circ$.
614: \item {OGLE-2007-BLG-514}: This is another double peaked high-magnification
615: event observed during 2007 season. The intra-peak trough shows a smooth
616: concave shape and thus we diagnose that the lens is a either wide or close
617: binary and not a planetary system. The heights of the two peaks are similar
618: and thus the source trajectory angle is close to $\alpha=45^\circ$.
619: \item {OGLE-2007-BLG-137/MOA-2007-BLG-091}: The apparent magnification
620: of this double-peaked event is only a modest $\sim 10$, although the
621: true magnification could be substantially higher if it is highly
622: blended. The intra-peak trough is not well-covered, but the data
623: in this region appear to show a relatively sharp change in the slope
624: of the magnification, perhaps followed by a linear rise. This
625: morphology is indicative of a planetary (or at least low mass ratio)
626: companion. The heights of the two peaks are quite different, implying
627: a source trajectory angle significantly different from
628: $\alpha=90^\circ$, assuming the event is due to a low mass-ratio companion.
629: \end{enumerate}
630:
631: \section{Conclusion}
632:
633: We have investigated the morphology of double-peaked high magnification
634: events, which can be produced by two very different classes of models:
635: planetary lenses in which the source trajectory passes close to the
636: back end of the wedge-shaped central caustic, and very wide or very
637: close binary lenses in which the source passes close to two of the
638: cusps of a Chang-Refsdal caustic. From a comparison
639: of the morphology of the light curves produced by these two classes of
640: models we have identified a diagnostic that can be used to
641: immediately distinguish between perturbations produced by a planet
642: from those produced by a binary companion. This diagnostic is based
643: on the difference in the shape of the intra-peak region of the light
644: curve. For binary lensing, the shape is smooth and concave, whereas
645: for planetary lensing, the shape is boxy or convex. The morphology of
646: the intra-peak region of planetary lensing events is due to the
647: existence of a weak cusp located between the two stronger
648: cusps. Finally, we applied this diagnostic to five observed
649: double-peaked high-magnification events.
650:
651: \acknowledgments
652:
653: This work was supported by the Science Research Center (SRC) program.
654: We would like to thank A. Gould for providing light curves of the events observed
655: by the MicroFUN collaboration during the 2007 season.
656:
657: %This work was supported by the grant (KRF-2006-311-C00072) of the
658: %Korea Research Foundation.
659:
660:
661:
662: \begin{thebibliography}{99}
663: \frenchspacing
664:
665: \bibitem[Abe et al.(2003)]{abe03} Abe, F., et al.\ 2003, \aap, 411, L493
666:
667: \bibitem[Albrow et al.(2002)]{albrow02}
668: Albrow, M.\ D., et al.\ 2002, \apj, 572, 1031
669:
670: \bibitem[An(2005)]{an05}
671: An, J.\ H.\ 2005, \mnras, 356, 1409
672:
673: \bibitem[Asada(2002)]{asada02}
674: Asada, H.\ 2002, \apj, 573, 825
675:
676: \bibitem[Beaulieu et al.(2006)]{beaulieu06}
677: Beaulieu, J.\ P., et al.\ 2006, Nature, 439, 437
678:
679: \bibitem[Bennett \& Rhie(1996)]{bennett96}
680: Bennett, D.~P., \& Rhie, S.~H.\ 1996, \apj, 472, 660
681:
682: \bibitem[Bond et al.(2004)]{bond04}
683: Bond, I.\ A., et al.\ 2004, \apj, 606, L155
684:
685: \bibitem[Bozza(1999)]{bozza99}
686: Bozza, V.\ 1999, \aap, 348, 311
687:
688: \bibitem[Chang \& Refsdal(1979)]{chang79}
689: Chang, K., \& Refsdal, S.\ 1979, Nature, 282, 561
690:
691: \bibitem[Chang \& Refsdal(1984)]{chang84}
692: Chang, K., \& Refsdal, S.\ 1984, \aap, 132, 168
693:
694: \bibitem[Chung et al.(2005)]{chung05}
695: Chung, S.-J., et al.\ 2005, \apj, 630, 535
696:
697: \bibitem[Dominik(1999)]{dominik99}
698: Dominik, M.\ 1999, \aap, 349, 108
699:
700: \bibitem[Dong et al.(2008)]{dong08}
701: Dong, S., et al.\ 2008, ApJ, submitted (arXiv:0804.1354)
702:
703: \bibitem[Gaudi et al.(2007)]{gaudi07}
704: Gaudi, B.\ S., et al.\ 2007, Science, 319, 927
705:
706: \bibitem[Gaudi(2008)]{gaudi08}
707: Gaudi, B.\ S.\ 2008, to appear in ASP Conference Series on Extreme Solar Systems (arXiv:0711.1614)
708:
709: \bibitem[Gould \& Loeb(1992)]{gould92} Gould, A., \& Loeb, A.\ 1992, \apj, 396, 104
710:
711: \bibitem[Gould et al.(2006)]{gould06}
712: Gould, A., et al.\ 2006, \apj, 644, L37
713:
714: \bibitem[Gould(2008)]{gould08}
715: Gould, A., et al.\ 2008, to appear in The Variable Universe: A Celebration of Bohdan Paczynski (arXiv:0803.4324)
716:
717: \bibitem[Griest \& Safizadeh(1998)]{griest98}
718: Griest, K., \& Safizadeh, N.\ 1998, \apj, 500, 37
719:
720: \bibitem[Udalski et al.(2005)]{udalski05}
721: Udalski, A., et al.\ 2005, \apj, 628, L109
722:
723: \bibitem[Soszy\'nski et al.(2001)]{soszynski01}
724: Soszy\'nski, I., et al.\ 2001, \apj, 552, 731
725:
726: \bibitem[Bond et al.(2001)]{bond01}
727: Bond, I.\ A., et al.\ 2001, \mnras, 327, 868
728:
729: \bibitem[Yoo et al.(2004)]{yoo04}
730: Yoo, J., et al.\ 2004, \apj, 616, 1204
731:
732: \bibitem[Cassan et al.(2004)]{cassan04}
733: Cassan, A., et al.\ 2004, \aap, 419, L1
734:
735: \bibitem[Witt(1990)]{witt90}
736: Witt, H.\ J.\ 1990, \aap, 236, 311
737:
738:
739: \end{thebibliography}
740:
741:
742:
743:
744:
745: \end{document}
746:
747:
748: