0804.4482/ms.tex
1: 	\documentclass{emulateapj}
2: \slugcomment{Accepted for Publication in Astrophysical Journal}
3: \lefthead{Rau et al.}
4: \righthead{A Survey for Fast Transients in the Fornax Cluster of Galaxies}
5: 
6: 
7: %\documentclass[11pt,preprint2]{aastex}
8: %\documentclass[12pt,preprint]{aastex}
9: %% manuscript produces a one-column, double-spaced document:
10: 
11: %\documentclass[manuscript]{aastex}
12: 
13: 
14: %% preprint2 produces a double-column, single-spaced document:
15: 
16: %\documentclass[preprint2]{aastex}
17: 
18: %% Sometimes a paper's abstract is too long to fit on the
19: %% title page in preprint2 mode. When that is the case,
20: %% use the longabstract style option.
21: 
22: %% \documentclass[preprint2,longabstract]{aastex}
23: 
24: %% If you want to create your own macros, you can do so
25: %% using \newcommand. Your macros should appear before
26: %% the \begin{document} command.
27: %%
28: %% If you are submitting to a journal that translates manuscripts
29: %% into SGML, you need to follow certain guidelines when preparing
30: %% your macros. See the AASTeX v5.x Author Guide
31: %% for information.
32: 
33: %\usepackage{graphicx,rotating}
34: %\usepackage{multirow,multicol} 
35: %\usepackage{natbib}
36: 
37: %\usepackage{latexsym,psfig}
38: %\usepackage{epstopdf}
39: %\usepackage{longtable}
40: 
41: \newcommand{\vdag}{(v)^\dagger}
42: \newcommand{\myemail}{arne@astro.caltech.edu}
43: 
44: % MY DEFS
45: \def\msun{M$_{\odot}$}
46: \def\lsun{L$_{\odot}$}
47: \def\rsun{R$_{\odot}$}
48: \def\asec{\ifmmode ^{\prime\prime}\else$^{\prime\prime}$\fi}
49: \def\farcs{\hbox{$.\!\!^{\prime\prime}$}}  	% Fractions of arcseconds
50: \def\farcm{\hbox{$.\!\!^{\prime}$}}  % Fractions of arcminutes
51: \def\degs{\ifmmode ^{\circ}\else$^{\circ}$\fi}
52: \def\farcm{\hbox{$^{\prime}$}}  % arcminutes
53: 
54: 
55: %% You can insert a short comment on the title page using the command below.
56: 
57: %\slugcomment{}
58: 
59: %% If you wish, you may supply running head information, although
60: %% this information may be modified by the editorial offices.
61: %% The left head contains a list of authors,
62: %% usually a maximum of three (otherwise use et al.).  The right
63: %% head is a modified title of up to roughly 44 characters.
64: %% Running heads will not print in the manuscript style.
65: 
66: \shorttitle{Fornax Fast Transient Survey}
67: \shortauthors{Rau et al.}
68: 
69: %% This is the end of the preamble.  Indicate the beginning of the
70: %% paper itself with \begin{document}.
71: 
72: \begin{document}
73: 
74: %% LaTeX will automatically break titles if they run longer than
75: %% one line. However, you may use \\ to force a line break if
76: %% you desire.
77: 
78: \title{A Survey for Fast Transients in the Fornax Cluster of Galaxies}
79: 
80: %% Use \author, \affil, and the \and command to format
81: %% author and affiliation information.
82: %% Note that \email has replaced the old \authoremail command
83: %% from AASTeX v4.0. You can use \email to mark an email address
84: %% anywhere in the paper, not just in the front matter.
85: %% As in the title, use \\ to force line breaks.
86: 
87: 
88: \author{A. Rau\altaffilmark{1}, E.O. Ofek\altaffilmark{1},   S.R. Kulkarni\altaffilmark{1}, B.F. Madore\altaffilmark{2}, O. Pevunova\altaffilmark{3}, M. Ajello\altaffilmark{4}}
89: 
90: \email{arne@astro.caltech.edu}
91: 
92: \altaffiltext{1}{Caltech Optical Observatories, MS 105-24, California Institute of Technology, Pasadena, CA 91125, USA}
93: \altaffiltext{2}{Observatories of the Carnegie Institution of Washington, 813 Santa Barbara Street, Pasadena, CA 91101, USA}
94: \altaffiltext{3}{Infrared Processing and Analysis Center, MS 100-22, California Institute of Technology, Jet Propulsion Laboratory, Pasadena, CA 91125, USA}
95: \altaffiltext{4}{Max-Planck Institut f\"ur Extraterrestrische Physik, Giessenbachstr.1, 80748 Garching, Germany}
96: 
97: \begin{abstract}
98: 
99:   The luminosity gap between novae ($M_{\rm R}\le -10$) and supernovae
100:   ($M_{\rm R}\ge -14$) is well  known since the pioneering research of
101:   Zwicky and Hubble.  Nearby  galaxy clusters and concentrations offer
102:   an  excellent opportunity  to  search for  explosions brighter  than
103:   classical novae  and fainter than  supernovae. Here, we  present the
104:   results of  a $B$-band  survey of 23  member galaxies of  the Fornax
105:   cluster, performed at the Las Campanas 2.5-m Irene duPont telescope.
106:   Observations  with a  cadence of  32\,minutes discovered  no genuine
107:   fast  transient   to  a  limiting  absolute   magnitude  of  $M_{\rm
108:     B}=-9.3$\,mag.  We provide a  detailed assessment of the transient
109:   detection  efficiency and the  resulting upper  limits on  the event
110:   rate  as  function  of  peak  magnitude.  Further,  we  discuss  the
111:   discoveries of five previously unknown foreground variables which we
112:   identified as two flare stars, two W~Uma type eclipsing binaries and
113:   a candidate $\delta$~Scuti/SX~Phe star.
114: 
115: \end{abstract}
116: 
117: \keywords{surveys --- stars: flare --- (stars: variables:) delta Scuti ---
118: stars: variables: other}
119: 
120: 
121: % --------------------------------------------------------------------
122: 
123: \section{Introduction}
124: 
125: The recent  findings of  a number of  enigmatic transients,  e.g., new
126: types of  novae \citep{Kulkarni:2007uq} and  supernovae related events
127: \citep{Ofek:2007fj,Smith:2007qy,Quimby:2007qy,Pastorello:2007jk}, have
128: demonstrated that  the phase space  of eruptive transients  is already
129: richer  than discussed  in  astronomy texts.   Even more  discoveries,
130: especially on  timescales of  days to weeks,  are anticipated  for the
131: upcoming     large-area      facilities,     such     as     SkyMapper
132: \citep{Schmidt:2005qy},  the  Panoramic  Survey  Telescope  and  Rapid
133: Response  System \citep[Pan-STARRS;][]{Kaiser:2002fk},  and  the Large
134: Synoptic Survey  Telescope \citep[LSST;][]{Tyson:2005lr}.  However, on
135: shorter timescales (minutes to hours), their observational designs are
136: less optimal and dedicates, small  experiments are more likely to lead
137: to great findings.  A small number of such surveys have been performed
138: previously
139: \citep[e.g.,][]{Becker:2004fk,Rykoff:2005lr,Morales-Rueda:2006qy,Ramsay:2006lr}
140: and informed us  of the difficulty of discovering  genuine new classes
141: of fast transients and variables\footnote{In the following we refer to
142:   an event as transient when  it has no known quiescent counterpart at
143:   any wavelength and  to a source which shows  a quiescent counterpart
144:   in our images or has  a catalogued counterpart as variable.}.  Here,
145: the  greatest challenge  is  to  pierce the  fog  of known  foreground
146: contaminators.   Asteroids   and  flares  from  M   dwarfs  have  been
147: established   to    dominate   the   short-timescale    variable   sky
148: \citep{Kulkarni:2006lr}.   However, once  this curtain  is penetrated,
149: exciting  and  rare events  can  be  found,  e.g., the  high-amplitude
150: optical  flickering   ($\sim$3.5\,mag  in  6\,min)   accompanying  the
151: outburst   of  the   Galactic   X-ray  binary   Swift~J195509.6+261406
152: \citep{Stefanescu:2007fk,Kasliwal:2007qy}.   The   latter  source  was
153: brought    to   attention   by    its   preceding    gamma-ray   flare
154: \citep{Pagani:2007fk}.  Nonetheless, high-cadence optical surveys will
155: be  capable   of  detecting   similar  events  independently   of  the
156: high-energy emission.
157: 
158: 
159: Another,  yet  unidentified,  bright  ($R=11.7$\,mag),  high-amplitude
160: candidate  transient  ($>$6\,mag in  2\,min)  was recently  discovered
161: along the line  of sight to the nearby  ($\sim$70\,Mpc) galaxy IC~4779
162: \citep{Klotz:2007lr}. If associated with  IC~4779, the event reached a
163: staggering peak  luminosity of $L_{\rm R,peak}\sim3\times10^{44}$\,erg
164: s$^{-1}$,   outshined   only   by   the  brightest   known   supernova
165: \citep[SN2005ap;][]{Quimby:2007qy}.
166: 
167: A  critical  component  of  every  successful transient  search  is  a
168: judicious   target  selection.   This   is  especially   relevant  for
169: short-cadence  experiment, where  only  a small  number  of fields  is
170: repeatedly  observed.  Obvious targets  for such  a search  are nearby
171: massive clusters of galaxies.   Clusters contain a large stellar mass,
172: which  increases the  probability of  finding rare  events  during the
173: transient phase.  Furthermore, their  known distance provides a direct
174: access  to the  absolute brightness  of the  events. In  addition, the
175: variety  of galaxy  types,  and thus  stellar  populations, may  offer
176: constraints on the progenitor ages.
177: 
178: In this paper we present the  results of an optical survey of galaxies
179: in the Fornax Cluster designed  to test the transient and variable sky
180: in a galaxy  cluster environment on timescales shorter  than one hour.
181: At    the    distance    of    $16.2\pm1.5$\,Mpc    \citep[$(m-M)_{\rm
182:   0}=30.91\pm0.19$;][]{Grillmair:1999lr}     and    with    negligible
183: foreground extinction  \citep[E(B-V) = 0.013\,mag;][]{Schlegel:1998ul}
184: observations  of the  Fornax cluster  allow one  to probe  the  gap in
185: absolute peak brightness between classical novae (M$_{\rm R}>-10$) and
186: supernovae (M$_{\rm  R}<-14$) already with  2-m-class telescopes. This
187: part  of  phase  space  has  become  particularly  alluring  with  the
188: recognition  of an  emerging population  of new  types  of explosions,
189: namely           the            Luminous           Red           Novae
190: \citep{Kulkarni:2007uq,Rau:2007kx,Ofek:2007uq}.
191: 
192: 
193: % --------------------------------------------------------------------
194: 
195: \section{Observations and Data reduction}
196: \label{sec:data}
197: 
198: Observations were  obtained with the  Wide Field Reimaging  CCD Camera
199: (WFCCD) at the  2.5-m Irene du Pont telescope  in Las Campanas, Chile.
200: The detector dimensions of  2048$\times$2048 pixels, together with the
201: plate  scale of  0\farcs774, provide  a  circular field  of view  with
202: $\sim$12\farcm5 radius.   Imaging was performed  in the $B$-band  as a
203: compromise between avoiding  flares from Galactic M-dwarfs (especially
204: bright  in  the  ultra-violet)  and increasing  the  contrast  between
205: possible transients and the  predominantly early type cluster galaxies
206: (bright in the red part of the spectrum).
207: 
208: For  the survey,  we selected  ten  fields covering  the 23  brightest
209: Fornax~I        cluster        members        (Table~\ref{tab:fields},
210: Figures available in printed paper).    These   fields   were
211: imaged  with a  sequence of  120\,s exposures  repeated up  to fifteen
212: times in each of five nights in October and December 2006.  Additional
213: single exposures  were obtained in  five nights in November  2006 (see
214: Table~\ref{tab:obslog} for a  complete log). As part of  a sequence, a
215: single exposure was taken at  each position, after which the telescope
216: slewed to the next field.  The resulting distribution of times between
217: consecutive     images    for     each    field     is     given    in
218: Figure~\ref{fig:cadence}.   Throughout  the  survey a  mean  intra-day
219: cadence  of $\Delta$$t=32.25$\,min  was achieved.   The spread  of the
220: observations  between  October  and  December  2006  allowed  for  the
221: possible detection of slowly  evolving events with timescales of about
222: one month (e.g., supernovae).
223: 
224: % ---------------------------------------------------------------------------------------------------------
225: \begin{table}
226: \begin{center}
227: \caption{Fields.
228: \label{tab:fields}}
229: \begin{tabular}{clll}
230: \tableline\tableline
231: \# & RA$_{\rm 2000\tablenotemark{a}}$ & Dec$_{\rm 2000}\tablenotemark{a}$ & Galaxies\tablenotemark{b} \\
232: \tableline
233: A & 03:24:36.7 & -36:26:21&  NGC1326, 1326A, 1326B\\
234: B & 03:22:41.1 & -37:10:54 & NGC1316, 1317\\
235: C & 03:31:04.3 & -33:36:42 & NGC1350\\
236: D & 03:36:29.5 & -34:52:28 & NGC1380, 1380A\\
237: E & 03:35:44.1 & -35:18:50 & NGC1373, 1374, 1375, 1379, 1381\\
238: F & 03:36:49.0 & -35:20:55 & NGC1379, 1381, 1382, 1387, MCG-06-09-008\\
239: G & 03:38:34.0 & -35:29:58 & NGC1399, 1404\\
240: H & 03:36:45.2 & -36:06:17 & NGC1369, 1386\\
241: I & 03:33:37.5 & -36:07:19 & NGC1365\\
242: K & 03:42:21.9 & -35:16:12 &NGC1427, 1428\\
243: \tableline
244: \end{tabular}
245: \tablenotetext{a}{Coordinates of the centers of the fields.}
246: \tablenotetext{b}{Prominent galaxies within the field of view.}
247: \end{center}
248: \end{table}
249: % ---------------------------------------------------------------------------------------------------------
250: 
251: % ---------------------------------------------------------------------------------------------------------
252: 
253: \begin{table}
254: \begin{center}
255: \caption{Observation log.
256: \label{tab:obslog}}
257: \begin{tabular}{ccccccccccc}
258: \tableline\tableline
259: Date (UT)\tablenotemark{a} & \multicolumn{10}{c}{Exposures\tablenotemark{b}}\\ 
260: 		& A	&B	&C	&D	&E	&F	&G	&H	&I	&K\\
261: \tableline
262: 06/10/23	&10	&10	&10	&10	&10	&9	&8	&8	&8	&10\\
263: 06/10/24	&13	&14	&13	&12	&14	&10	&13	&13	&13	&13\\
264: 06/10/25   &\multicolumn{10}{c}{lost due to bad weather}\\
265: 06/11/17	&1	&1	&1	&1	&1	&1	&1	&1	&1	&1\\
266: 06/11/18	&1	&1	&1	&1	&1	&1	&1	&1	&1	&1\\
267: 06/11/19	&1	&1	&1	&1	&1	&1	&1	&1	&1	&1\\
268: 06/11/20	&1	&1	&1	&1	&1	&1	&1	&1	&1	&1\\
269: 06/11/21	&1	&1	&1	&1	&1	&1	&1	&1	&1	&1\\
270: 06/12/19	&14	&15	&14	&14	&14	&14	&14	&14	&14	&14\\       
271: 06/12/20	&13	&13	&13	&14	&13	&14	&14	&14	&14	&14\\              
272: 06/12/21   &12	&12	&12	&12	&12	&12	&12	&10	&12	&12\\
273: \tableline
274: total\tablenotemark{c}: 	&67	&69	&67	&67	&68	&64	&66	&64	&66	&68\\
275: \tableline
276: \end{tabular}
277: \tablenotetext{a}{UTC date of observation in the format YY/MM/DD.}
278: \tablenotetext{b}{Number of 120\,s $B$-band exposure for each field in a given night.}
279: \tablenotetext{c}{Total number of observations for each field.}
280: \end{center}
281: \end{table}
282: % ---------------------------------------------------------------------------------------------------------
283: 
284: Figure~\ref{fig:cadence} indicates that  our primary sensitivity is to
285: variability  on  $\sim32$\,min  timescales.   Here, we  have  a  total
286: exposure time of 13.7 days distributed over 616 images (neglecting the
287: once-per-day  exposures   in  November).   With  a  covered   area  of
288: 0.136\,deg$^2$ per  image, this yields a total  areal exposure, E$_A$,
289: of 1.86\,deg$^2$ days.  We have a  factor of 5 and 44 larger E$_A$ for
290: inter-night and inter-month observation intervals, respectively.
291: 
292: % ---------------------------------------------------------------------------------------------------------
293: \begin{figure}[htbp]
294: \begin{center}
295: \includegraphics[width=0.35\textwidth,angle=-90]{f1.jpg}
296: \caption{Time between two consecutive images of a given pointing. The mean cadence is 32.25\,min.}
297: \label{fig:cadence}
298: \end{center}
299: \end{figure}
300: % ---------------------------------------------------------------------------------------------------------
301: 
302: Image reduction (bias subtraction,  flat fielding) was performed using
303: standard  {\it  IRAF}  routines\footnote{IRAF  was distributed  by  the
304:   National Optical Astronomy Observatories,  which are operated by the
305:   Association of  Universities for Research in  Astronomy, Inc., under
306:   cooperative  agreement with the  National Science  Foundation}.  The
307: astrometric   solutions   were   obtained   in  reference   to   NOMAD
308: \citep{Zacharias:2005lr}       with       {\it      ASCfit       v3.0}
309: \footnote{http://www.astro.caltech.edu/$\sim$pick/ASCFIT/README.ascfit3.html}.
310: The  strategy  to  search  for  transients  and  variables  in  single
311: exposures  implies  that Cosmic-ray  removal  by  median combining  of
312: multiple  images was not  applicable.  Hence,  we applied  a Laplacian
313: algorithm  \citep[{\it lacos\_im};][]{van-Dokkum:2001qy}  which allows
314: the identification of Cosmic-rays in single frames.
315: 
316: For  absolute  photometric calibration  we  used  observations of  the
317: standard   star  field   T~Phe   \citep{Landolt:1992lr}  taken   under
318: photometric  conditions  (2006  November  19).  This  calibration  was
319: applied to  the images  of the ten  Fornax fields obtained  during the
320: same night.  Exposures from  the remaining nights were tied relatively
321: to these absolutely calibrated  frames.  Here, we selected an ensemble
322: of   at   least  40   local,   non-saturated,  non-variable   ($\Delta
323: B<0.01$\,mag) reference stars for  each image.  The photometric offset
324: between  the two  observations was  estimated by  deriving  the median
325: brightness  offset  of these  sets  of  stars  with respect  to  their
326: brightness in the reference frame.  This was successively done for all
327: observations  of a given  field, thus  providing a  common photometric
328: zero-point.  The  resulting photometric accuracy for  point sources is
329: shown in  Figure~\ref{fig:deltaMag}.  In a typical  night a 5-$\sigma$
330: limiting magnitude  of $B~\sim22$\,mag was  reached.  Sources brighter
331: than $B\sim15$\,mag were generally saturated.
332: 
333: % ---------------------------------------------------------------------------------------------------------
334: \begin{figure}[htbp]
335: \begin{center}
336: \includegraphics[width=0.45\textwidth,angle=0]{f4.jpg}
337: \caption{Photometric accuracy. Shown are the combined point source measurements of three exposures obtained on 2006 Dec 19 UT. The seeing was 1\farcs4 and a 5-$\sigma$ limiting magnitude of $B\sim$22\,mag was reached. Sources brighter than $B\sim15$\,mag are saturated.}
338: \label{fig:deltaMag}
339: \end{center}
340: \end{figure}
341: % ---------------------------------------------------------------------------------------------------------
342: 
343: 
344: The search for transients and  variables was performed in two separate
345: steps. First  we applied a  point spread function  (PSF)-matched image
346: subtraction   using   a  software   package   based   on  {\it   ISIS}
347: \citep{Alard:2000fk}.  Reference frames  for all fields were generated
348: by combining the three best  quality exposures taken in November 2006.
349: Candidate  transients and  variables were  detected in  the difference
350: images  using  {\it SExtractor  v2.5.0}  \citep{Bertin:1996uq}. For  a
351: typical  image,  this resulted  in  about  100  sources brighter  than
352: 5-$\sigma$, out of  which all but a few  (see below) were subsequently
353: identified  by   eye  as  false  positives   (e.g.,  image  artifacts,
354: Cosmic-ray residuals, PSF distortion, and  sources near the rim of the
355: circular  field  of view  which  were  in the  input  but  not in  the
356: reference  frame).  The  PSFs of  the resulting  candidates  were then
357: verified against those of stars in the input images.
358: 
359: The efficiency, $\epsilon$, of recovering transients in the difference
360: images was  modeled with Monte  Carlo simulations.  Here, a  subset of
361: the survey data was enriched with PSF-matched artificial point sources
362: and subsequently  passed through the image  subtraction pipeline.  The
363: efficiency was  then calculated from  the ratio of recovered  to input
364: test sources.  In total we obtained 1.9 Million efficiency points with
365: random location in the field and random magnitudes of 14$<$B$<$23.  As
366: the main  driver for our  survey was the  search for events  in nearby
367: galaxies, understanding  $\epsilon$ as  function of the  local surface
368: brightness became  important. Thus, a subset of  the efficiency points
369: was placed inside the extend of the bright Fornax cluster members.  As
370: expected, $\epsilon$ was found  to decrease with increasing background
371: flux  contribution  (see  Fig.~\ref{fig:efficiency}a).  For  ``empty''
372: locations  ($B\sim22.6$\,mag  asec$^{-2}$)  a recovery  efficiency  of
373: $\epsilon=0.85$ (0.7)  was achieved  for sources brighter  than $B=20$
374: (21.3).   In positions  with higher  surface  brightness ($B\sim21.5$,
375: $20.0$\,mag   asec$^{-2}$)   the    maximum   efficiency   was   lower
376: ($\epsilon=0.60$, $0.40$).  Fig~\ref{fig:efficiency}b shows $\epsilon$
377: as function of surface brightness for three test source magnitudes.
378: 
379: % ---------------------------------------------------------------------------------------------------------
380: \begin{figure}[htbp]
381: \begin{center}
382: \includegraphics[width=0.50\textwidth,angle=0]{f5a.jpg}
383: \hspace{-0.9truecm}
384: \includegraphics[width=0.50\textwidth,angle=0]{f5b.jpg}
385: \caption{left: Transient recovery efficiency as function of source magnitude for three levels of background surface brightness. The solid line (B=22.4\,mag asec$^{-2}$) corresponds to the efficiency away  from the Fornax cluster members, while the dotted (B=21.5\,mag asec$^{-2}$) and dashed (B=20.0\,mag asec$^{-2}$) lines are representative for locations within the galaxies. Sources brighter than B=15 are generally saturated. The faint sources cutoff  corresponds to the 5-$\sigma$ detection limits in the difference images. 
386: right: Transient recovery efficiency as function of background surface brightness for test sources with B=21.3 (solid), 19.5 (dotted) and 18.0 (dashed). Fewer efficiency points have been obtained at high surface brightness regions, thus, these bins have the lowest statistic.}
387: \label{fig:efficiency}
388: \end{center}
389: \end{figure}
390: % ---------------------------------------------------------------------------------------------------------
391: 
392: In   a  second   step   we  performed   a  catalog-based   variability
393: search. While this method is inferior to image subtraction for sources
394: overlaying the bright cluster galaxies, it improves the detectivity of
395: low-amplitude,  bright  variables  in  the field.   For  these,  image
396: subtraction is typically very sensitive to PSF stability, convolution,
397: and alignment,  and residuals  unrelated to intrinsic  variability may
398: remain.   The  source catalogs  were  compiled  using {\it  SExtractor
399:   v2.5.0}  on  the  photometrically  calibrated  images  (see  above).
400: Objects  with  two  or   more  detections  brighter  than  $B=20$  and
401: deviations of more than $B=0.1$\,mag from their median brightness were
402: selected as candidate variables.
403: 
404: A serious contamination in optical transient and variable searches can
405: come  from solar  system bodies.   However, due  to the  high ecliptic
406: latitude ($\beta\approx-53$\,\degs) observations of the Fornax cluster
407: are only  marginally affected. Our  choice of using the  B-band filter
408: further  reduced the  impact  of asteroids,  whose  emission peaks  at
409: longer   wavelength.   Moreover  repetitive   visits  of   each  field
410: throughout the nights  were allowing secure astrometric identification
411: of moving objects  with proper motions as low as  3\asec\ per day. For
412: example, the  expected parallax  of a Kuiper  belt object,  at 100\,AU
413: from the Sun  at that location and time of  the year, is $\sim33$\asec
414: day$^{-1}$ and $\sim3\farcs5$\,day$^{-1}$ due  to the Earth and object
415: motion,  respectively.  Thus, it  was  not  surprising  that only  one
416: astrometrically variable was found in the survey dataset.
417:  
418: % --------------------------------------------------------------------
419: 
420: \section{Results}
421: 
422: The  search netted  seven high-confidence  photometric  transients and
423: variables.  Two,  both  earlier  reported  type Ia  SNe  in  NGC~1316,
424: SN~2006dd  \citep{Monard:2006lr} and  SN~2006mr \citep{Monard:2006fk},
425: were only  detected in image subtraction.  The  remaining five objects
426: were variable  point sources found with both  detection methods.  None
427: of        them        was        previously       catalogued        in
428: SIMBAD\footnote{http://simbad.u-strasbg.fr/},
429: NED\footnote{http://nedwww.ipac.caltech.edu/},
430: GCVS\footnote{heasarc.gsfc.nasa.gov/W3Browse/all/gcvsnsvars.html}    or
431: detected in the  {\it ROSAT}/PSPC all-sky survey \citep{Voges:1999fj}.
432: Below  we provide  a  brief  description for  each  of these  variable
433: sources (see also Table~\ref{tab:candidates}).
434: 
435: % ---------------------------------------------------------------------------------------------------------
436: %{\rotate
437: \begin{table*}
438: \begin{center}
439: \caption{Summary of detected sources.
440: \label{tab:candidates}}
441: \begin{tabular}{lllcccl}
442:   \tableline\tableline
443:   \# & RA$_{\rm 2000}$ & Dec$_{\rm 2000}$ & B$_{\rm peak}$\tablenotemark{a} & B$_{\rm quiescence}$\tablenotemark{b} &Classification & D$_{\rm proj}$\tablenotemark{c} \\
444:   & & & (mag) & (mag) & & \\
445:   \tableline
446:   FA-1   &  03:24:24.56 & $-$36:30:26.5 & 18.91$\pm$0.03 & 19.17$\pm$0.03 & M1--2 & 36\,kpc to NGC~1326\\
447:   FB-1    & 03:22:31.38 & $-$37:04:01.5 &  18.48$\pm$0.02 & 18.84$\pm$0.14 & (?)$\delta$~Scuti/SX~Phe & 26\,kpc to NGC~1317\\ 
448: FE-1    & 03:36:18.85 & $-$35:14:58.9 & 16.07$\pm$0.01 & 16.18$\pm$0.09 & W~Uma & 29\,kpc to NGC~1381\\
449: FH-1    & 03:36:53.67 & $-$36:05:29.9 & 18.66$\pm$0.03 & 20.70$\pm$0.09 & M3--4 & 42\,kpc to NGC~1386\\
450: FK-1    & 03:42:33.73 & $-$35:18:29.0 & 15.20$\pm$0.01 & 15.42$\pm$0.20 & W~Uma & 35\,kpc to NGC~1427\\
451: SN~2006dd & 03:22:41.64 & $-$37:12:13.2 & 17.25$\pm$0.03\tablenotemark{d} & -- & SN Ia & 1.2\,kpc to NGC~1316 \\
452: SN~2006mr & 03:22:42.84 & $-$37:12:28.5 & 16.16$\pm$0.03\tablenotemark{e} & -- & SN Ia & 1.1\,kpc to NGC~1316 \\
453: \tableline
454: \end{tabular}
455: \tablenotetext{a}{Observed peak $B$-band magnitude.}
456: \tablenotetext{b}{Observed mean (for periodic variables) or quiescence (for eruptive variables) $B$-band magnitude.}
457: \tablenotetext{c}{Projected distance to center of nearest Fornax cluster galaxy.}
458: \tablenotetext{d}{Observed $B$-band magnitude on 2006 Oct 23.13 UT}
459: \tablenotetext{e}{Observed $B$-band magnitude on 2006 Nov 17.58 UT}
460: \end{center}
461: \end{table*}
462: %}
463: % ---------------------------------------------------------------------------------------------------------
464: 
465: 
466: \subsection{Eruptive variables}
467: 
468: Two of the  variable sources were detected when  they exhibited single
469: flares from their otherwise  constant quiescent brightness. The first,
470: FA-1  (Figure~\ref{fig:fc}a),  is  a  faint,  $B=19.17\pm0.03$,  point
471: source    that    showed   a    distinct    outburst   with    $\Delta
472: B=0.26\pm0.04$\,mag (Figure~\ref{fig:lc}a). The  rise time to peak was
473: shorter  than 1\,hr  (2$\times$cadence)  and the  decline lasted  over
474: $\sim2.5$\,hrs.   The  source  has  a bright  near-IR  counterpart  as
475: detected     by     2MASS\footnote{http://www.ipac.caltech.edu/2mass/}
476: ($J=14.58\pm0.03$,      $K=13.65\pm0.04$).      We      fitted     its
477: $B-J=4.59\pm0.04$\,mag            and           $B-K=5.52\pm0.05$\,mag
478: colors\footnote{Here,  we  assume  that  the  2MASS  observations  are
479:   representative  of  the  quiescence   (or  median  in  case  of  the
480:   non-eruptive  sources  discussed  in  \S~\ref{sec:per})  brightness.
481:   This  is supported  by  generally lower  amplitudes  in the  near-IR
482:   compared    to    the    $B-$band.}     with    stellar    templates
483: \citep{Pickles:1998xy} and  found them to be consistent  with an M1--2
484: star.  The most likely explanation is a low-amplitude long-decay flare
485: of a UV Ceti type Galactic M1--2 dwarf (M$_{\rm B}\sim11.5$\,mag) at a
486: distance of $\sim350$\,pc.
487: 
488: FH-1 (Figure~\ref{fig:fc}d)  was detected  when it underwent  a strong
489: flare with $\Delta$B=2.06$\pm$0.15\,mag.  It rose to peak in less than
490: 1\,hr and decayed to its quiescent brightness of $B=20.70\pm0.09$ over
491: 1.5\,hrs  (Figure~\ref{fig:lc}d).   FH-1  is  bright  in  the  near-IR
492: ($J=15.02\pm0.02$,      $K=14.29\pm0.07$)      and     its      colors
493: ($B-J=5.68\pm0.10$\,mag,  $B-K=6.41\pm0.10$\,mag)  indicate  an  M3--4
494: stellar classification. We suggest that this outburst was a long-decay
495: flare of  an M3--4-dwarf (M$_{\rm  B}\sim12.5$\,mag) at a  distance of
496: $\sim$450\,pc .
497: 
498: % ---------------------------------------------------------------------------------------------------------
499: \begin{figure*}[htbp]
500: \begin{center}
501: \includegraphics[width=1\textwidth,angle=0]{f7.jpg}
502: \caption{Observed $B$-band measurements for the five candidate variables listed in Table~\ref{tab:candidates}. Shown are the lightcurves for the five nights with multiple exposures per field (2006 Oct 23, 24 and Dec 19, 20, 21 UTC). Time is given in fractional days (UTC). A compilation of all data is available in electronic form (Table~\ref{tab:data}).}
503: \label{fig:lc}
504: \end{center}
505: \end{figure*}
506: % ---------------------------------------------------------------------------------------------------------
507: 
508: 
509: \subsection{Periodic variables}
510: \label{sec:per}
511: 
512: Regular  brightness modulations  were  shown by  three sources.   FE-1
513: (Figure~\ref{fig:fc}c)  displayed   a  nearly  sinusoidal  variability
514: (Figure~\ref{fig:lc}c)   with   a   maximum   amplitude   of   $\Delta
515: B=0.31\pm0.05$\,mag.          A        Lomb-Scargle        periodogram
516: \citep{Scargle:1982kx}   of  the  heliocentric   corrected  lightcurve
517: revealed  two possible periods,  one at  $0.19025\pm0.00002$\,days and
518: another    at    twice    this    value    ($0.38050\pm0.00004$\,days;
519: Figure~\ref{fig:phaseE}).   The  latter  is  favored  by  an  apparent
520: difference of $\Delta B=0.07\pm$0.02\,mag  in depth of two consecutive
521: minima.  Note  that FE-1  was included  in the overlap  of two  of our
522: pointings (FE \& FF) and thus has twice the data coverage (132 images)
523: compared to  the remaining candidates.   2MASS detected the  source at
524: $J=15.53\pm0.05$  and $K=15.16\pm0.15$,  which, together  with  a mean
525: brightness of $B=16.18\pm0.09$,  results in $B-J=0.65\pm0.10$\,mag and
526: $B-K=1.02\pm0.17$\,mag.   These  colors  suggest  a spectral  type  of
527: A7--F0.     The    source   is    also    included    in   the    {\it
528:   GALEX}\footnote{http://galex.stsci.edu/GR2/?page=mastform}
529: \citep{Martin:2003qy}  source catalog  and  independent photometry  on
530: archival    images    provides    magnitudes   $FUV=23.4\pm0.1$    and
531: $NUV=19.04\pm0.02$.
532: 
533: A likely explanation for FE-1 is an eclipsing contact binary system of
534: W~Uma type with an  orbital period of $0.38050\pm0.00004$\,days.  This
535: is supported by the apparent  two minima in the lightcurve which could
536: indicate the phase of primary and secondary occultation.
537: 
538: % ---------------------------------------------------------------------------------------------------------
539: \begin{figure}[htbp]
540: \begin{center}
541: \includegraphics[width=0.35\textwidth,angle=90]{f8.jpg}
542: \caption{Lomb-Scargle periodogram (left) for FE-1 showing that the highest power occurs for at a frequency of $(6.083\pm0.001)\times10^{-5}$\,Hz (or period of $0.19025\pm0.00002$\,days). The phase  folded  heliocentric  corrected lightcurve (right) indicates differences in the depth of consecutive minima, suggesting that the real period is likely at twice the above value, $0.38050\pm0.00004$\,days.} 
543: \label{fig:phaseE}
544: \end{center}
545: \end{figure}
546: % ---------------------------------------------------------------------------------------------------------
547: 
548: The  second   periodic  variable,  FK-1   (Figure~\ref{fig:fc}e),  was
549: observed  as a  point source  with  mean brightness  of $B=15.42$  and
550: regular    non-sinusoidal     variations    with    a     period    of
551: $0.31864\pm0.00006$\,days     (Figures~\ref{fig:lc}e,\ref{fig:phaseK}).
552: The  maximum photometric  amplitude  was $\Delta  B=0.65\pm0.07$\,mag.
553: The   source  has  a   bright  2MASS   counterpart  ($J=13.42\pm0.03$,
554: $K=12.93\pm0.03$)    and     its    colors    ($B-J=2.00\pm0.20$\,mag,
555: $B-K=2.49\pm0.20$\,mag) resemble those of a G0-9 giant or dwarf.
556: 
557: While the short modulation  and lightcurve shape in principle resemble
558: also  that of  an RRc~Lyrae  pulsator, the  spectral type  of  G5-9 is
559: inconsistent with this  classification. The most likely interpretation
560: for FK-1 is, similar to FE-1, a W~Uma type eclipsing binary.
561: 
562: % ---------------------------------------------------------------------------------------------------------
563: \begin{figure}[htbp]
564: \begin{center}
565: \includegraphics[width=0.35\textwidth,angle=90]{f9.jpg}
566: \caption{Same as Figure~\ref{fig:phaseE} for FK-1. The highest power occurs at a frequency of $(7.264\pm0.002)\times10^{-5}$\,Hz (or a period of $0.15932\pm0.00003$\,days). The phase  folded  heliocentric  corrected lightcurve again indicates differences in the depth of consecutive minima, suggesting that the real period is likely $0.31864\pm0.00006$\,days.}
567: \label{fig:phaseK}
568: \end{center}
569: \end{figure}
570: % ---------------------------------------------------------------------------------------------------------
571: 
572: The  classification   of  the   remaining  variable,  FB-1,   is  more
573: challenging. Its  lightcurve displayed apparently  erratic variability
574: with   an   rms  of   0.14\,mag   around   the   mean  brightness   of
575: $B=18.84\pm0.14$   and   with   a   maximum   amplitude   of   $\Delta
576: B=0.57\pm0.08$\,mag     (Figure~\ref{fig:lc}b).      A    Lomb-Scargle
577: periodogram (Figure~\ref{fig:phaseB}) revealed  a number of peaks, the
578: most   prominent  being   at  a   period   of  $0.0585\pm0.0001$\,days
579: ($84.24\pm0.15$\,min).  However,  the folded lightcurve  shows several
580: data points that  deviate notably from this phasing.  We also not that
581: the  power  spectrum  shows  evidence  for  an  additional  period  of
582: $\sim107$\,min. Thus, the tentative  timescale of modulation has to be
583: taken  with  caution  and  confirmation by  observations  with  higher
584: sampling rate will be required.
585: 
586: % ---------------------------------------------------------------------------------------------------------
587: \begin{figure}[htbp]
588: \begin{center}
589: \includegraphics[width=0.35\textwidth,angle=90]{f10.jpg}
590: \caption{Same  as Figure~\ref{fig:phaseE}  for  FB-1. The  periodogram
591:   shows  a  large  number  of   peaks,  with  the  most  prominent  at
592:   $(1.975\pm0.005)\times10^{-4}$\,Hz ($84.24\pm0.15$\,min). A second period occurs at ~107\,min.}
593: \label{fig:phaseB}
594: \end{center}
595: \end{figure}
596: % ---------------------------------------------------------------------------------------------------------
597: 
598: FB-1 has no 2MASS counterpart  to a limiting magnitude of $J=17$ which
599: constrains its  color to $B-J<1.84$\,mag.  This, together  with a {\it
600:   GALEX} detection at  NUV=21.39$^{+0.56}_{-0.37}$ and a non-detection
601: in the FUV filter to  $>21$\,mag restricts a stellar classification to
602: A-G.  A spectrum obtained with the ESO Multi-Mode Instrument (EMMI) at
603: the  NTT, La  Silla,  shows  a blue  continuum  with prominent  Balmer
604: absorption lines  (Figure~\ref{fig:spec}). A comparison  with template
605: stellar     spectra    \citep{Pickles:1998xy}    suggests     an    A7
606: classification. This spectral range  is populated by pulsators such as
607: $\delta$~Scuti  stars  and RR~Lyrae.   Of  those, only  $\delta$~Scuti
608: stars,  and the phenomenologically\footnote{$\delta$~Scuti  and SX~Phe
609:   stars   are  generally  distinguished   by  their   hosting  stellar
610:   population and  metallicity. This can  not be accomplished  with the
611:   available  data. }   similar SX~Phe  stars \citep{Rodriguez:1990uq},
612: exhibit modulations on timescales  as short as detected for FB-1.  As
613: indicated by the  number of short period peaks  in the periodogram, we
614: may have  to consider also  events with periods below  84\,mins.  Some
615: white  dwarf  systems  do  show  periodic  pulsations  below  30\,min.
616: ZZ~Ceti    type     sources    have    white     dwarf    temperatures
617: \citep[11--12\,kK;][]{Bergeron:2004lr} in  agreement with the observed
618: colors \citep{Kawka:2006uq}.  However, the lack of pressure broadening
619: of the  Balmer absorption  lines rules out  the white  dwarf scenario.
620: Awaiting  further test  of the  periodic  modulation, we  find that  a
621: Galactic  $\delta$~Scuti  or  SX~Phe  star  appears  to  be  the  most
622: intriguing interpretation for FB-1.
623: 
624: % ---------------------------------------------------------------------------------------------------------
625: \begin{figure}[htbp]
626: \begin{center}
627: \includegraphics[width=0.45\textwidth,angle=0]{f11.jpg}
628: \caption{NTT/EMMI spectrum  (grey) of FB-1 obtained on  2007 Dec 12.19
629:   UT using the low-dispersion  mode (RILD) with grism\#5 (instrumental
630:   FWHM=4.5\AA).  The  data were reduced with  customized IRAF routines
631:   and  flux  calibrated   in  comparison  to  the  spectro-photometric
632:   standard  star  LTT377 \citep{Hamuy:1992zr}.  The  black solid  line
633:   shows   an  A7V   template   spectrum  from  Pickles (1998).
634:   Prominent spectral features are indicated.}
635: \label{fig:spec}
636: \end{center}
637: \end{figure}
638: % ---------------------------------------------------------------------------------------------------------
639: 
640: \section{Discussion}
641: 
642: Here were reported the results  of a dedicated search for optical fast
643: transients  and variables in  the environment  of the  Fornax~I galaxy
644: cluster.   In  total,  the  survey  netted  two  transients  and  five
645: high-amplitude   ($\Delta   B>0.1$\,mag)   variable   sources.    Both
646: transients have been detected as  a result of the inter-month sampling
647: of  our   observations  and  have  independently   been  reported  and
648: classified     as      type     Ia     supernovae      in     NGC~1216
649: \citep{Monard:2006lr,Monard:2006fk}.   The   properties  of  the  five
650: variables  suggests that  they are  located in  the foreground  to the
651: Fornax~I cluster and belong to  the Milky Way stellar population. This
652: is in agreement with the large projected distance (26--42\,kpc) of the
653: events      to     the      nearest      Fornax     galaxies      (see
654: Table~\ref{tab:candidates}).  Two of  the variables were identified as
655: flares from M-dwarfs, while other two are likely eclipsing binaries of
656: the W~Uma type.   The remaining variable shows indications  for a very
657: short periodicity ($\sim84$\,min), and  spectroscopy suggests it to be
658: a $\delta$\,Scuti or SX~Phe  star.  No genuine fast transient brighter
659: than $B=21.3$ has been detected.
660: 
661: Surveys, like the one presented  here, can offer valuable guidance for
662: the  planning  and  execution  of  the next  generation  of  transient
663: experiments.   In order to  provide relevant  event rates,  a complete
664: identification of all sources detected with a given choice of cadence,
665: filter, depth, amplitude, and  field selection, is required. This will
666: pose  inevitable  challenges   for  follow-up  limited  projects  like
667: Pan-STARRS  and  LSST. However,  at  least  for variables,  multi-band
668: photometry and good temporal coverage can already be strong indicators
669: for  the  nature   of  an  event,  as  demonstrated   for  the  Fornax
670: sources. Indeed,  automatic cross-matching  with UV to  near-IR source
671: catalogs with  a similar depth  as the surveys  will be vital  for the
672: success of the future large optical projects.
673: 
674: \subsection{All-Sky Rate of Fast Transients}
675: 
676: In the following,  we will discuss the estimates  and expectations for
677: transients  on  timescales of  $\sim32$\,min  and  with apparent  peak
678: brightness of $15<B<21.3$.  We start by determining the upper limit on
679: the rate  of cosmological fast  transients not associated with  any of
680: the Fornax  cluster galaxies.  As described  in \S~\ref{sec:data}, the
681: areal survey exposure is  $E_A=1.86$\,deg$^2$ and the efficiency for a
682: single  detection  of  a   field  source  brighter  than  $B=21.3$  is
683: $\epsilon=0.7$.  As we require a candidate to be found in at least two
684: images,  $\epsilon$ needs to  be squared.   The non-detection  of fast
685: transients in the survey translates  into a 95\,\% Poisson upper limit
686: of  N=$3$\,  events  \citep{Gehrels:1986qy}.  The  corresponding  rate
687: follows from
688: 
689: \begin{equation}
690: r_A=\frac{N}{\epsilon^2\cdot E_A}
691: \end{equation}
692: 
693: as  ${\rm r_A}  <3.3$\,  events day$^{-1}$  deg$^{-2}$,  or an  annual
694: all-sky rate of ${\rm  r_A}<5\times10^7$.  Note, that these limits are
695: valid only  for an event population that  is homogeneously distributed
696: over the  sky.  In  other words, the  detection probability has  to be
697: independent  of the field  selection.  Under  this assumption,  we can
698: match our  estimates with previously obtained results,  e.g., from the
699: Deep  Lens   Survey  transient  search  \citep[DLS;][]{Becker:2004fk}.
700: Indeed, the Fornax  survey is very similar to the  DLS with respect to
701: exposure  (1.38 vs  1.1  deg$^2$  days) and  probed  timescale (32  vs
702: 22\,min).  As also no genuine fast transient was found in the DLS, the
703: the event rate is comparable.
704: 
705: \subsection{Rates of Fast Transients in Nearby Galaxies}
706: 
707: The main motivation of our survey was the exploration of the transient
708: and variable population in a  specific environment, namely in a galaxy
709: cluster with known distance and  known stellar mass. For this purpose,
710: we will  now calculate the specific  rate of fast  transients per unit
711: stellar mass, or its proxy, the $B$-band luminosity.
712: 
713: As discussed above (see also Fig.~\ref{fig:efficiency}), $\epsilon$ is
714: not  only a  function  of the  magnitude  of a  transient, $M_T$,  but
715: depends  also  on the  underlying  surface  brightness, m$_{\rm  SB}$.
716: Similarly, the product of  the luminosity, L$_{\rm SB,i}$, enclosed in
717: a surface  brightness bin, m$_{\rm SB,i}$, and  the relative exposure,
718: $t$, is different for all targeted Fornax galaxies.  Here, $t$ denotes
719: the time a given galaxy was probed for transients on timescales of the
720: 32\,min  cadence. We  start  by calculating  L$_{\rm  SB,i}$ for  each
721: galaxy  listed in  Table~\ref{tab:fields}, with  $i$ running  from the
722: saturation  limit  of  16.7\,mag  asec$^{-2}$ to  the  sky  background
723: brightness of  22.6\,mag asec$^{-2}$ in  steps of 0.1\,mag.   Next, we
724: estimate the  survey exposure, $E_{SB,i}$, for each  m$_{\rm SB,i}$ in
725: units\footnote{Here, we  use an  absolute solar $B$-band  magnitude of
726:   M$_{B,\odot}=5.48$}  of  M$_{B,\odot}$   years  by  summing  L$_{\rm
727:   SB,i}\cdot t$ for  all galaxies. The rate of  transients as function
728: of the brightness can then be calculated as
729: 
730: \begin{equation}
731: r(M_T)=\sum_i{\frac{N(M_T)}{\epsilon(M_T,m_{\rm SB})_i^2\cdot E_{{\rm SB},i}}}
732: \end{equation}
733: 
734: Formerly, the number of detected  transients, $N$, also depends on the
735: peak brightness. However,  here we use again the  95\,\% Poisson upper
736: limit of $N(M_T)=N=3$  events, as no event in  the entire probed range
737: of   $15<B<21.3$   was   detected.  
738: 
739: The result  is presented in Figure~\ref{fig:rates}, where  we show the
740: upper  limit  on the  rate  as  funtion  of transient  magnitude.   As
741: expected,  the  brightest events  (m$_B<$18.5,  M$_B<-12.4$) have  the
742: lowest limits  ($r<10^{-8}$ events M$_{B,\odot}^{-1}$  yr$^{-1}$). For
743: the   Milky   Way   disk   \citep[$M_B=-19.9$;][]{Quillen:1997}   this
744: translates into  $<140$ events per year,  e.g., a few  times more than
745: the classical novae rate.  One could argue that any Galactic transient
746: population  brighter  than  classical  novae  and  with  a  comparable
747: frequency  may  have  been  discovered already.   However,  the  short
748: timescales tested in our survey require a significantly higher cadence
749: than typically  used in transient searches.  Fast  transients may have
750: previously  been detected in  single images,  but simply  discarded as
751: unconfirmed or spurious  events. However, there have been  a number of
752: planetary transit and variable  star studies, which offer the required
753: high cadence \citep[e.g.,;][]{Kane:2005xx,Kraus:2007xx,Lister:2007xx}.
754: These experiments  typically target dense  Galactic fields and  thus a
755: large  stellar population.  Unfortunately,  results on  transients, as
756: opposed to transits, are rarely recorded or published.
757: 
758: 
759: % ---------------------------------------------------------------------------------------------------------
760: \begin{figure}[htbp]
761: \begin{center}
762: \includegraphics[width=0.45\textwidth,angle=0]{f12.jpg}
763: \caption{95\,\% Poission upper limit on the rate of transients on timescales of 32\,min as function of peak brightness.}
764: \label{fig:rates}
765: \end{center}
766: \end{figure}
767: % ---------------------------------------------------------------------------------------------------------
768: 
769: 
770: \section{Conclusion}
771: 
772: The  existence  of fast  transients  requires  further constraint  with
773: specialized experiments.   The most exciting facility  in this respect
774: will be the  Palomar Transient Factory (PTF, Rau et  al., in prep.), a
775: dedicated transient survey instrument at the Palomar 48-inch telescope
776: with expected  first light  in Winter  2008.  The PTF  will use  a 7.8
777: square degree field of view  to perform a number of transient searches
778: on various time scales. The large field of view will allow us to cover
779: galaxy  clusters  with  fewer  pointings, thus  increasing  the  total
780: stellar mass observed at a given cadence.
781: 
782: However, we have  shown that the probability of  detecting a transient
783: strongly depends on the local  surface brightness. While this is not a
784: surprising  result,  its  relevance  for  future  searches  in  nearby
785: galaxies needs to  be understood. There will always be  the need for a
786: compromise between limiting  magnitude of an event to  be detected and
787: galaxy fraction covered.  Larger apertures or longer exposures are not
788: necessarily   improving   the    detection   probability   for   faint
789: events. Although more source photons will be collected, also more host
790: light is recieved and and a  larger extend of the galaxy will be above
791: saturation. The main factor for  success will be the image quality. At
792: a  given  surface  brightness,   a  narrower  PSF  will  increase  the
793: signal-to-noise, and thus, the probability for a detection.
794: 
795: % --------------------------------------------------------------------
796: 
797: \acknowledgments 
798: 
799: We are grateful to Paul Price for providing his image subtraction code
800: and to  Carrol Wainwright  for his contribution  to the  pipeline.  We
801: thank  Mara  Salvato,  Mansi  M.~Kasliwal  and  S.~Bradley  Cenko  for
802: discussion and  constructive criticism. We  thank the referee  for his
803: stimulating comments and suggestions. This work is based on in part on
804: observations  collected at the  European Southern  Observatory, Chile.
805: This publication  makes use of data  products from the  Two Micron All
806: Sky  Survey,   which  is  a   joint  project  of  the   University  of
807: Massachusetts    and   the    Infrared    Processing   and    Analysis
808: Center/California  Institute  of Technology,  funded  by the  National
809: Aeronautics  and   Space  Administration  and   the  National  Science
810: Foundation. This research has  made use of the NASA/IPCA Extragalactic
811: Database  (NED) which is  operated by  the Jet  Propulsion Laboratory,
812: California Institute  of Technology, under contract  with the National
813: Aeronautics and  Space Administration. This work is  supported in part
814: by grants from the National Science Foundation and NASA.
815: 
816: % --------------------------------------------------------------------
817: 
818: %{\it Facilities:}  \facility{Du Pont:WFCCD}, \facility{NTT:EMMI},
819: 
820: % --------------------------------------------------------------------
821: 
822: 
823: 
824: \bibliographystyle{apj}
825: \begin{thebibliography}{}
826: \bibitem[{Alard} 2000]{Alard:2000fk}
827: {Alard}, C. 2000, \aaps, 144, 363
828: 
829: \bibitem[{Becker} et al. 2004]{Becker:2004fk}
830: {Becker}, A.~C., {Wittman}, D.~M., {Boeshaar}, P.~C., {et al.} 2004, \apj, 611, 
831: 418
832: 
833: \bibitem[{Bergeron} et al. 2004]{Bergeron:2004lr}
834: {Bergeron}, P., {Fontaine}, G., {Bill{\`e}res}, M., {et al.} 2004, \apj, 600, 404
835: 
836: \bibitem[{Bertin} \& {Arnouts} 1996]{Bertin:1996uq}
837: {Bertin}, E. \& {Arnouts}, S. 1996, \aaps, 117, 393
838: 
839: \bibitem[{Gehrels} 1986]{Gehrels:1986qy}
840: {Gehrels}, N. 1986, \apj, 303, 336
841: 
842: \bibitem[{Grillmair} et al. 1999]{Grillmair:1999lr}
843: {Grillmair}, C.~J., {Forbes}, D.~A., {Brodie}, J.~P., {et al.} 1999, \aj, 117, 167
844: 
845: \bibitem[{Hamuy} et al. 1992]{Hamuy:1992zr}
846: {Hamuy}, M., {Walker}, A.~R., {Suntzeff}, N.~B., {et al.} 1992, \pasp, 104, 533
847: 
848: \bibitem[{Kaiser} et al. 2002]{Kaiser:2002fk}
849: {Kaiser}, N., {Aussel}, H., {Burke}, B.~E., {et al.} 2002, in Presented at the Society of  Photo-Optical Instrumentation Engineers (SPIE) Conference, Vol. 4836, Survey  and Other Telescope Technologies and Discoveries. Edited by Tyson, J.  Anthony; Wolff, Sidney. Proceedings of the SPIE, Volume 4836, pp. 154-164  (2002)., ed. J.~A. {Tyson} \& S.~{Wolff}, 154
850: 
851: \bibitem[{Kane} et al. 2005]{Kane:2005xx}
852: Kane, S.~R., Lister, T.~A., Cameron, A.~C., et al. 2005, \mnras, 362, 117
853: 
854: \bibitem[{Kasliwal} et al. 2007]{Kasliwal:2007qy}
855: {Kasliwal}, M.~M., {Cenko}, S.~B., {Kulkarni}, S.~R., {et al.} 2007, ArXiv e-prints, 0708.0226
856: 
857: \bibitem[{Kawka} et al. 2006]{Kawka:2006uq}
858: {Kawka}, A., {Vennes}, S., {Oswalt}, T.~D., {et al.} 2006, \apjl, 643, L123
859: 
860: \bibitem[{Klotz} et al. 2007]{Klotz:2007lr}
861: {Klotz}, A., {Boer}, M., \& {Atteia}, J. 2007, GRB Coordinates Network, 6769, 1
862: 
863: \bibitem[{Kraus} et al. 2007]{Kraus:2007xx}
864: {Kraus}, A.~L., Craine, E.~R., Giampapa, M.~S., et al. 2007, AJ, 134, 1488
865: 
866: \bibitem[{Kulkarni} et al. 2007]{Kulkarni:2007uq}
867: {Kulkarni}, S.~R., {Ofek}, E.~O., {Rau}, A., {et al.} 2007,  \nat, 447, 458
868: 
869: \bibitem[{Kulkarni} \& {Rau} 2006]{Kulkarni:2006lr}
870: {Kulkarni}, S.~R. \& {Rau}, A. 2006, \apjl, 644, L63
871: 
872: \bibitem[{Landolt} 1992]{Landolt:1992lr}
873: {Landolt}, A.~U. 1992, \aj, 104, 340
874: 
875: \bibitem[{Lister} et al. 2007]{Lister:2007xx}
876: Lister, T.~A., West, R.~G., Wilson, D.~M., et al. 2007, \mnras, 379, 647
877: 
878: \bibitem[{Martin} et al. 2003]{Martin:2003qy}
879: {Martin}, C., {Barlow}, T., {Barnhart}, W., {et al.} 2003, in Presented at the Society of Photo-Optical  Instrumentation Engineers (SPIE) Conference, Vol. 4854, Future EUV/UV and  Visible Space Astrophysics Missions and Instrumentation. Edited by J. Chris  Blades, Oswald H. W. Siegmund. Proceedings of the SPIE, Volume 4854, pp.  336-350 (2003)., ed. J.~C. {Blades} \& O.~H.~W. {Siegmund}, 336
880: 
881: \bibitem[{Monard} 2006]{Monard:2006lr}
882: {Monard}, L.~A.~G. 2006, Central Bureau Electronic Telegrams, 553, 1
883: 
884: \bibitem[{Monard} \& {Folatelli} 2006]{Monard:2006fk}
885: {Monard}, L.~A.~G. \& {Folatelli}, G. 2006, Central Bureau Electronic  Telegrams, 723, 1
886: 
887: \bibitem[{Morales-Rueda} et al. 2006]{Morales-Rueda:2006qy}
888: {Morales-Rueda}, L., {Groot}, P.~J., {Augusteijn}, T., {et al.} 2006, \mnras, 371, 1681
889: 
890: \bibitem[{Ofek} et al. 2007a]{Ofek:2007fj}
891: {Ofek}, E.~O., {Cameron}, P.~B., {Kasliwal}, M.~M., {et al.} 2007a, \apjl, 659, L13
892: 
893: \bibitem[{Ofek} et al. 2007b]{Ofek:2007uq}
894: {Ofek}, E.~O., {Kulkarni}, S.~R., {Rau}, A., {et al.} 2007b, ArXiv e-prints, 710
895: 
896: \bibitem[{Pagani} et al. 2007]{Pagani:2007fk}
897: {Pagani}, C., {Barthelmy}, S.~D., {Cummings}, J.~R., {et al.} 2007, GRB Coordinates Network, 6489, 1
898: 
899: \bibitem[{Pastorello} et al. 2007]{Pastorello:2007jk}
900: {Pastorello}, A., {Smartt}, S.~J., {Mattila}, S., {et al.} 2007, Nature, 447, 829
901: 
902: \bibitem[{Pickles} 1998]{Pickles:1998xy}
903: {Pickles}, A.~J. 1998, \pasp, 110, 863
904: 
905: \bibitem[{Quillen} \& {Sarajedini} 1997]{Quillen:1997}
906: Quillen, A.~C., \& Sarajedini, V.~L., 1997, AJ, 115, 1412 
907: 
908: \bibitem[{Quimby} et al. 2007]{Quimby:2007qy}
909: {Quimby}, R.~M., {Aldering}, G., {Wheeler}, J.~C., {et al.} 2007, \apjl, 668, L99
910: 
911: \bibitem[{Ramsay} et al. 2006]{Ramsay:2006lr}
912: {Ramsay}, G., {Napiwotzki}, R., {Hakala}, P., {et al.} 2006, \mnras, 371,  957
913: 
914: \bibitem[{Rau} et al. 2007]{Rau:2007kx}
915: {Rau}, A., {Kulkarni}, S.~R., {Ofek}, E.~O., {et al.} 2007, \apj, 659, 1536
916: 
917: \bibitem[{Rodriguez} et al. 1990]{Rodriguez:1990uq}
918: {Rodriguez}, E., {Rolland}, A., \& {Lopez de Coca}, P. 1990, \apss, 169, 113
919: 
920: \bibitem[{Rykoff} et al. 2005]{Rykoff:2005lr}
921: {Rykoff}, E.~S., {Aharonian}, F., {Akerlof}, C.~W., {et al.} 2005, \apj, 631, 1032
922: 
923: \bibitem[{Scargle} 1982]{Scargle:1982kx}
924: {Scargle}, J.~D. 1982, \apj, 263, 835
925: 
926: \bibitem[{Schlegel} et al. 1998]{Schlegel:1998ul}
927: {Schlegel}, D.~J., {Finkbeiner}, D.~P., {Davis}, M., 1998, ApJ, 500, 525
928: 
929: \bibitem[{Schmidt} et al. 2005]{Schmidt:2005qy}
930: {Schmidt}, B.~P., {Keller}, S.~C., {Francis}, P.~J., {et al.} 2005,  in Bulletin of the American Astronomical Society, Vol.~37, Bulletin of the  American Astronomical Society, 457
931: 
932: \bibitem[{Smith} et al  2007]{Smith:2007qy}
933: {Smith}, N., {Li}, W., {Foley}, R.~J., {et al.} 2007, \apj, 666, 1116
934: 
935: \bibitem[{Stefanescu} et al. 2007]{Stefanescu:2007fk}
936: {Stefanescu}, A., {Slowikowska}, A., {Kanbach}, G., {et al.} 2007, GRB Coordinates Network, 6508, 1
937: 
938: \bibitem[{Tyson} 2005]{Tyson:2005lr}
939: {Tyson}, A. 2005, in Astronomical Society of the Pacific Conference Series,  Vol. 339, Observing Dark Energy, ed. S.~C. {Wolff} \& T.~R. {Lauer}, 95
940: 
941: \bibitem[{van Dokkum} 2001]{van-Dokkum:2001qy}
942: {van Dokkum}, P.~G. 2001, \pasp, 113, 1420
943: 
944: \bibitem[{Voges} et al. 1999]{Voges:1999fj}
945: {Voges}, W., {Aschenbach}, B., {Boller}, T., {et al.} 1999, \aap, 349, 389
946: 
947: \bibitem[{Zacharias} et al. 2005]{Zacharias:2005lr}
948: {Zacharias}, N., {Monet}, D.~G., {Levine}, S.~E., {et al.} 2005, VizieR Online Data Catalog, 1297, 0
949: \end{thebibliography}
950: 
951: \input{tab5}
952: 
953: % ---------------------------------------------------------------------------------------------------------
954: % \begin{figure}[htbp]
955: % \begin{center}
956: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f2a.jpg}}
957: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f2b.jpg}}
958: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f2c.jpg}}
959: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f2d.jpg}}
960: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f2e.jpg}}
961: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f2f.jpg}}
962: % \caption{Survey fields. Prominent galaxies are marked. North is up and East to the left. Images are 25\farcm\ on each side. (for electronic version only)}
963: % \label{fig:fieldsa}
964: % \end{center}
965: % \end{figure}
966: % ---------------------------------------------------------------------------------------------------------
967: 
968: % ---------------------------------------------------------------------------------------------------------
969: % \begin{figure}[htbp]
970: % \begin{center}
971: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f3a.jpg}}
972: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f3b.jpg}}
973: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f3c.jpg}}
974: % \fbox{\includegraphics[width=0.43\textwidth,angle=0]{f3d.jpg}}
975: % \caption{Figure~\ref{fig:fieldsa} continued}
976: % \label{fig:fieldsb}
977: % \end{center}
978: % \end{figure}
979: % ---------------------------------------------------------------------------------------------------------
980: 
981: % ---------------------------------------------------------------------------------------------------------
982: \begin{figure}[htbp]
983: \begin{center}
984: \includegraphics[width=0.45\textwidth,angle=0]{f6a.jpg}
985: \includegraphics[width=0.45\textwidth,angle=0]{f6b.jpg}
986: \includegraphics[width=0.45\textwidth,angle=0]{f6c.jpg}
987: \includegraphics[width=0.45\textwidth,angle=0]{f6d.jpg}
988: \includegraphics[width=0.45\textwidth,angle=0]{f6e.jpg}
989: \caption{Finding charts.}
990: \label{fig:fc}
991: \end{center}
992: \end{figure}
993: % ---------------------------------------------------------------------------------------------------------
994: \clearpage
995: 
996: 
997: \end{document}
998: 
999: