1: \documentclass{elsart}
2:
3: \usepackage{graphicx,epsfig,epsf,amssymb}
4: \usepackage{amssymb}
5:
6: \newcommand{\gr}{$\gamma$-ray}
7: \newcommand{\grs}{$\gamma$-rays}
8: \newcommand{\vhe}{V\textsc{HE}}
9: \newcommand{\hess}{H.E.S.S.}
10:
11: \begin{document}
12:
13: \begin{frontmatter}
14:
15: \title{Precision Measurement of Optical Pulsation using a Cherenkov Telescope}
16:
17: \author[HD,LSW]{J. A. Hinton\corauthref{add1}}
18: \author[HD]{, G. Hermann}
19: \author[HD,UK]{, P. Kr\"otz}
20: \author[HD]{, S. Funk}
21:
22: \address[HD]{Max-Planck-Institut f\"ur Kernphysik, P.O.~Box~103980,
23: D-69029~Heidelberg, Germany}
24: \address[LSW]{Landessternwarte, K\"onigstuhl, D-69117 Heidelberg, Germany}
25: \address[UK]{now at: Physikalisches Institut, Universit\"at zu K\"oln,
26: Germany}
27:
28: \corauth[add1]{Corresponding author, Jim.Hinton@mpi-hd.mpg.de}
29:
30:
31: \begin{abstract}
32:
33: During 2003, a camera designed to measure the optical pulsations of
34: pulsars was installed on a telescope of the H.E.S.S. array. The
35: array is designed for $\gamma$-ray astronomy in the $\sim$ 100 GeV -
36: 100~TeV energy regime. The aims of this exercise were two-fold: to
37: prove the pulsar timing capabilities of H.E.S.S. on all relevant
38: time-scales, and to explore the possibility of performing
39: sensitive optical pulsar measurements using the $\sim\,100$ m$^{2}$ mirror
40: of a Cherenkov telescope. Measurements of the Crab pulsar with this
41: instrument demonstrate an order of magnitude sensitivity improvement
42: over previous attempts using Cherenkov telescopes.
43: Here we describe the design and performance
44: of the system and discuss design considerations for future
45: instruments of this type.
46:
47: \end{abstract}
48:
49: \begin{keyword}
50: \PACS 97.60.Gb, 95.85.Kr\\
51: Crab pulsar, optical pulsations, Cherenkov telescopes
52: \end{keyword}
53:
54: \end{frontmatter}
55:
56: %_________________________________________________________________
57: \section{Introduction}
58: \label{intro}
59:
60: The six nearby high spin-down luminosity pulsars known at
61: $\gamma$-ray energies provide a laboratory for the study
62: of high energy processes under extreme conditions~\cite{EGRET_PULSARS}.
63: The extremely strong gravitational and magnetic fields close
64: to pulsars make the modelling of such objects rather complex.
65: Indeed, the acceleration processes at work in pulsar magnetospheres, and the
66: associated broad-band non-thermal \emph{pulsed} emission, remain
67: poorly understood despite several decades of effort. Observations in
68: a wide range of wavelength bands are required to provide information
69: not just on different particle energies, but also on different regions within the
70: magnetosphere.
71: Observations above 10~GeV are widely seen as key
72: to our understanding of the origin of the high energy emission (see
73: for example \cite{POLARCAP,OUTERGAP}). A small number of
74: (apparently pulsed) photons above this energy were detected by
75: the EGRET satellite but statistics were insufficient to extend
76: pulsed spectra much beyond 10~GeV~\cite{EGRET_VHE}. The much larger
77: collection area of ground based instruments is required to explore
78: the very high energy (\vhe) regime.
79: The major new air-Cherenkov experiments for ground based $\gamma$-ray
80: astronomy, such as H.E.S.S.~\cite{HESSproject}, MAGIC~\cite{MAGIC} and
81: VERITAS~\cite{VERITAS} exhibit greatly improved sensitivity around 100
82: GeV compared to previous instruments~\cite{STACEE,CELESTE}.
83: Detections or upper limits on pulsed emission from these new
84: instruments will likely be based on data taken over months or even
85: years. Especially in the case of non-detection of a pulsed signal, it
86: is necessary to prove the long term stability and accuracy of the
87: timing systems of the experiment. The key components of
88: the timing system are the timing hardware (usually based on GPS clock
89: technology), with associated interface electronics,
90: and the software used to derive phase information for an
91: individual object. Some of these $\gamma$-ray pulsars exhibit
92: pulsed emission detectable at optical wavelengths.
93: The detection of pulsed optical emission using the
94: timing system of a $\gamma$-ray instrument can be used
95: as a means to demonstrate its absolute timing capability.
96:
97: Optical pulsar measurements are also interesting in their own right
98: (see for example \cite{WHYOP}). The six known optical pulsars exhibit
99: dominantly non-thermal (optical) emission and are all relatively
100: young, and energetic or nearby. The class of pulsars with measured
101: optical pulsations therefore falls close to the class of potential
102: very-high-energy $\gamma$-ray emitting pulsars. The optical emission
103: also provides a bridge between the well studied radio and X-ray
104: bands. The recent claim of enhanced optical emission associated with
105: giant radio pulses (GRPs) of the Crab pulsar~\cite{SHEARER}, acts as
106: an additional motivation for further studies.
107:
108: Optical observations of pulsars using conventional optical
109: telescopes are performed either with fast readout CCD
110: cameras~\cite{ULTRACAM} or photon counting instruments~\cite{OPTIMA}.
111: The requirements for such a device are time-resolution
112: of $\sim$100 $\mu$s and sufficient signal/noise
113: to resolve the pulsed emission against the night sky background
114: (NSB).
115: Sensitivity improves linearly with both optical PSF width and
116: mirror diameter. Traditional devices have excellent angular
117: resolution and modest mirror area. An alternate concept is to use
118: the large mirror area of Cherenkov telescopes to compensate
119: for their modest angular resolution. The potential advantages of this
120: approach include smaller fluctuations
121: due to \emph{seeing} (intensity fluctuations on small angular scales
122: caused be atmospheric turbulence), potentially deadtime free
123: operation and in general a different set of systematic effects.
124: The idea of measuring the optical pulsations of pulsars with Cherenkov
125: telescopes is at least a decade old and pulses from the Crab pulsar
126: have been measured successfully by several instruments:
127: Whipple~\cite{WHIPPLECRAB}, STACEE~\cite{STACEE_OP},
128: CELESTE~\cite{CELESTE}, and HEGRA~\cite{CT1CRAB},
129: with a typical sensitivity of $0.1\,\sigma/\sqrt{t/\mathrm{s}}$
130: (i.e. a one second observation leads on average to an 0.1 standard
131: deviation detection significance, increasing with the square-root of
132: time, $t$).
133: The most recent measurement, using the modestly sized (1.7~m diameter)
134: mirror of the HEGRA CT-1 telescope, is described in
135: \cite{CT1CRAB}. This measurement achieved a sensitivity of 0.14
136: $\sigma/\sqrt{t/\mathrm{s}}$ and the authors estimate a value of 1
137: $\sigma/\sqrt{t/\mathrm{s}}$ can be achieved for the MAGIC telescope
138: using the same device (indeed a preliminary study using MAGIC
139: \cite{MAGIC_OP} confirms this predicted sensitivity.
140: However, as all other known optical pulsars are dimmer than the
141: Crab by a factor $>200$, even with this level of sensitivity
142: over 600~hours of observations would be required to detect pulses from
143: any pulsar apart from the Crab.
144:
145: The advantage of the H.E.S.S. telescopes (15~m
146: focal length, 107~m$^{2}$ of reflector) is that the optical point
147: spread function is considerably smaller than the pixel size for
148: on-axis observations, with a FWHM of only
149: $0.07^{\circ}$\cite{HESSOPT,HESSOPT2}. To exploit this advantage a
150: single-channel prototype system was installed on the first H.E.S.S.
151: telescope for 8 nights of Crab pulsar observations in January
152: 2003~\cite{FRANZEN}. A 7-pixel optical pulsar camera with improved
153: electronic and optical properties was installed in October 2003 on the
154: last of the H.E.S.S. telescopes to receive a Cherenkov camera. The
155: observations described here were performed with this new camera during
156: October and November 2003.
157:
158: \section{The Camera System}
159: \label{camera}
160:
161: The main improvement over the (single-channel) prototype device described in
162: \cite{FRANZEN} is the addition of a 6 channel ``veto'' camera surrounding the
163: central pixel. The additional channels have the dual purpose of
164: identifying optical atmospheric transients such as air-showers and
165: meteorites, and of monitoring the NSB level close to the target
166: pulsar. The optical pulsar camera is shown in figure~\ref{fig:Photo}.
167: The camera box contains seven photomultiplier tubes (PMTs) with
168: associated light cones, a high voltage distribution system and two
169: positioning LEDs. The Photonis XP2960 PMTs and the light collecting
170: cones are the same as used for the H.E.S.S. Cherenkov
171: camera~\cite{HESSCAMERA}. The wavelength dependence of the instrument
172: response is given in figure~\ref{fig:Wavelength}. The central pixel is equipped with an
173: exchangeable aperture (of 1 to 22~mm diameter). 20 or 22~mm apertures
174: were used for most of the data discussed here, providing optimal
175: signal to noise given the on-axis optical point-spread-function of a
176: H.E.S.S. telescope at $46^{\circ}$ zenith angle (the culmination of
177: the Crab). The 1~mm (pin-hole) aperture is used for pointing tests
178: using stars. The camera box is equipped with a pneumatic lid. The
179: positioning LEDs are used to monitor pointing and tracking accuracy
180: via images acquired by a CCD camera mounted at the centre of the
181: telescope dish, with the pulsar camera in its field of view.
182:
183: \begin{figure}[ht]
184: \begin{center}
185: \mbox{\epsfig{file=figure1.eps, width=1.0\textwidth }}
186: \end{center}
187: \caption{Photograph showing the principle mechanical components of the optical
188: pulsar camera: the central pixel aperture, the ``veto'' pixel light
189: guides, the positioning LEDs and the HV inputs/signal outputs for
190: the photomultipliers mounted within the casing.}
191: \label{fig:Photo}
192: \end{figure}
193:
194: \begin{figure}[ht]
195: \begin{center}
196: \mbox{\epsfig{file=figure2.eps, width=0.97\textwidth }}
197: \end{center}
198: \caption{Wavelength dependent efficiency of the instrument compared to
199: the optical spectrum of the Crab nebula~\cite{HUBBLE}, before and
200: after atmospheric absorption. The lower solid curve gives the combined
201: quantum and collection efficiency of the PMT. The dashed curve
202: includes also the reflectivity of the primary mirror and the
203: overall collection efficiency of the light cone.}
204: \label{fig:Wavelength}
205: \end{figure}
206:
207: The analogue signals from the seven camera pixels were fed via 50~m
208: coaxial cables to an electronics container at the base of the
209: telescope. The camera electronics were housed in this temperature
210: controlled environment. Signals from all channels were amplified with
211: DC-coupled spectroscopic shaper modules, with a characteristic
212: shaping time of 100~$\mu$s (FWHM). The signal was then digitised with
213: an over-sampling factor of $\sim\,4$ using a HYTEC VTR 2536 14-bit Flash
214: ADC. This scheme provides a current measurement proportional to the
215: average photon flux on 100~$\mu$s time-scales. Any optical pulsed
216: signal is measured on top of a DC background generated by NSB photons,
217: which have a typical rate of $\sim$100~MHz.
218:
219: Timing information was provided by two Meinberg 167BGT GPS clocks, one
220: associated with the H.E.S.S. central trigger system~\cite{HESSTRIGGER}
221: and a second identical clock housed on the telescope tracking room.
222: These clocks supply 10~MHz and 1~Hz TTL outputs synchronized to
223: UTC. For the H.E.S.S. clock these signals were distributed via optical
224: fibre from the central control building. Two VME counter models are
225: used to count these TTL pulses and generate event timestamps. The 1~Hz
226: channel is used to reset the 10~MHz counter. Every two seconds an
227: ASCII time-string was read via a serial port on the GPS clock, as an
228: absolute time reference. The clock counters are read out together
229: with each FADC sample via a VME based CPU. A custom data acquisition
230: system (DAQ) developed for this measurement achieved the required
231: over-sampling factor of $\sim$4 with an average sampling rate 28~kHz
232: on all channels. Additional electronics in the electronics container
233: provided temperature monitoring and remote control of photomultiplier
234: HV and the camera lid. Data transmission to a remote machine
235: responsible for data storage occurred in parallel to data taking ---
236: resulting in deadtime free operation.
237: Due to the large sampling rate the pulsar camera produced data
238: at a rate comparable with that of the H.E.S.S. 960 pixel Cherenkov
239: cameras ($\sim$1~MB/s).
240:
241:
242: \section{Observations}
243:
244: The observations described here were performed in late October and early
245: November 2003, partly contemporaneous with observations using
246: the Effelsberg 100~m radio telescope (discussed in \cite{JESSNER}).
247: The optical measurements described here were accompanied by
248: $\gamma$-ray observations of the Crab nebula with the 3 remaining
249: H.E.S.S. telescopes. Overall 60~hours of optical Crab observations
250: were performed of which 41 hours of on-source observations pass all
251: data quality criteria. Over 300~GB of raw data were obtained.
252:
253: The observations consisted of data taken tracking the Crab pulsar
254: itself, interspersed with occasional off-source runs and periodic
255: pedestal monitoring runs taken with the camera lid closed.
256: Observations occurred semi-automatically, with intervention from the
257: H.E.S.S. shift-crew only at the beginning and end of each night. To
258: keep the camera centred on the pulsar it was necessary to make online
259: corrections for atmospheric refraction and bending in the arms of the
260: telescope. The verification of the absolute pointing of the
261: instrument was done by scanning several stars across the aperture of
262: the central pixel, and also using CCD images of the positioning LEDs
263: and star images on the camera lid. The telescope tracking precision
264: was monitored during the measurement and exhibited a maximum rms
265: deviation of 3~$''$. The total systematic pointing error is estimated
266: as $\sim 30''$, introducing a negligible error on the recorded optical
267: signal.
268:
269: For roughly half of the observing time, two independent GPS clocks
270: were read-out in parallel: the H.E.S.S. central trigger timing system
271: and the independent system of the optical pulsar instrument. The mean
272: difference between the two event time-stamps during this time was
273: 2.1~$\mu$s, consistent with the length of the optical fibre over which
274: the central clock time was transmitted. The rms of this time
275: difference was 150~ns, consistent with the precision quoted by the
276: clock manufacturers.
277:
278: The night sky background level in the camera FOV is rather
279: inhomogeneous. The Crab nebula has a V-band magnitude of 8.4 and a
280: diameter of $\approx$ 6$'$, covering essentially the entire central
281: pixel. The brightest star present in the surrounding pixels has
282: $M_{\mathrm{v}}$ = 9.9, causing significant variation in pixel
283: currents as it rotates around the outer pixels. The mean NSB level in
284: the outer pixels was found to be of $\sim 3\times10^{12}$ photons
285: m$^{-2}$ s$^{-1}$ sr$^{-1}$. Previous measurements of the NSB level
286: at the H.E.S.S. site found $\approx2.4\times10^{12}$ photons m$^{-2}$
287: s$^{-1}$ sr$^{-1}$ (300-650~nm) close to the zenith and levels higher
288: by a factor $\sim\,2$ on the galactic plane in the outer
289: galaxy~\cite{HESS_NSB}. NSB measurements of this field using a HEGRA
290: telescope on La Palma~\cite{CT1CRAB} yielded a value of
291: $4.3\times10^{12}$ photons m$^{-2}$ s$^{-1}$ sr$^{-1}$, in reasonable
292: agreement, given the different site and the different altitude range
293: of our observations.
294:
295:
296: \section{Analysis \& Results}
297:
298: The first step of the analysis is the summation of every four
299: consecutive samples to produce statistically independent measurements
300: of $\approx$ 140\,$\mu$s duration. A timestamp is generated based on
301: the average time of the four samples. For each five minute observation
302: run, pedestal values are subtracted based on the ADC values with the
303: camera lid closed. Periods with unstable weather conditions are
304: removed by cuts on the rms and gradient in the signals measured in the
305: veto pixels. Transient background events are removed by excluding 2~s
306: blocks in which the signal in any \emph{surrounding} pixel exceeded 5
307: times the signal rms for any 5 consecutive measurements. No cuts are
308: made based on the central pixel signal.
309:
310: \subsection{Timing and Light-Curve}
311:
312: A crucial aspect of the analysis is the barycentering and
313: phase-folding of the event times. For this purpose software developed
314: specifically for H.E.S.S. was employed~\cite{SGPhD}. The software has
315: been compared against the standard TEMPO package~\cite{Till,TEMPO} and no
316: difference greater than 1~$\mu$s found for non-binary pulsars.
317: Ephemerides from Jodrell Bank were used for the phase
318: calculation~\cite{Jodrell}. Comparing the ephemerides provided for
319: October and November we derive a second derivative of the pulsar
320: frequency: $\ddot{f}\,=\,
321: 9.3\,\times\,10^{-21}\,\mathrm{s}^{-3}$. This value of $\ddot{f}$,
322: together with $\dot{f}$, $f$ and $t_{0}$ taken from the published
323: November 15$^{\mathrm{th}}$ ephemerides, was used to calculate the
324: absolute phase information given here.
325:
326: Following phase-folding a clear signal from the Crab pulsar was
327: visible in all datasets. The average signal/noise of the full dataset
328: is such that an average significance of $4\,\sigma/\sqrt{t/\mathrm{s}}$ can be
329: assigned to the pulsed signal, more than an order of magnitude better
330: than previous Cherenkov telescope measurements. The level of
331: precision achievable on different time-scales is illustrated in
332: figure~\ref{fig:Phaso10sec}. Phase-folded light-curves extracted from
333: 10\,s, 100\,s and 1\,hour datasets are shown. Even in a 10\,s
334: exposure, the peak position is resolved (inset in
335: figure~\ref{fig:Phaso10sec}) to better than 1\,ms.
336: The position of the peak in subsets of the data is determined
337: by a fit of a smoothed version of the overall measured phasogram
338: with two free parameters: the normalisation and the relative
339: phase shift. This empirical fit function is shown in the inset
340: of figure~\ref{fig:Phaso10sec}.
341:
342: \begin{figure}[t]
343: \begin{center}
344: \mbox{\epsfig{file=figure3.eps, width=1.01\textwidth }}
345: \end{center}
346: \caption{Average optical signal versus phase for the Crab pulsar for 10
347: second, 100 second and 1 hour integration times. The inset shows
348: a fit to the main peak for the 10 second dataset. For clarity,
349: offsets of $\pm$10 counts have been added to the 10 s and 1 h
350: datasets. The DC signal produced by the Crab nebula and the NSB
351: has been subtracted ($\sim 2 \times 10^{4}$ ADC counts).
352: }
353: %\vspace{10mm}
354: \label{fig:Phaso10sec}
355: \end{figure}
356:
357:
358: As can be seen from figure~\ref{fig:PeakDrift} the position of the
359: main peak can be measured with an accuracy of $<20\,\mu$s in one
360: 5-minute run. Unfortunately the available radio ephemerides have an
361: absolute precision of $\sim 70\mu$s, so any comparison is limited by
362: the radio accuracy.
363: The apparent drift during the period of our measurements ($\sim$ one
364: month) was $\approx$60 $\mu$s, consistent within the accuracy of the
365: Jodrell ephemerides for October and November. The apparent phase shift
366: of the mean peak between January and November 2003 was
367: $\approx$100 $\mu$s, again consistent within the accuracy of the
368: radio ephemerides used.
369:
370: \begin{figure}[t]
371: \begin{center}
372: \vspace{6mm}
373: \mbox{\epsfig{file=figure4.eps, width=1.07\textwidth }}
374: \end{center}
375: \caption{Reconstructed relative main peak position for the Crab pulsar over one
376: week in October 2003. Each point represents one 5 minute run.}
377: \label{fig:PeakDrift}
378: \vspace{3mm}
379: \end{figure}
380:
381: The phasogram extracted from the full 39~hour dataset is shown in
382: figure~\ref{fig:Phasogram}. The best published optical phasogram for
383: the Crab nebula is that extracted from 2~hours of observations with
384: the Hubble Space Telescope (HST)~\cite{HUBBLE}.
385: We find generally good agreement between the H.E.S.S. result and that
386: of HST. For example, the ratio of the height of the main pulse to that
387: of the inter-pulse: HST 3.78$\pm$0.11, H.E.S.S. 3.721$\pm$0.003; and
388: the FWHM of the main peak: HST 0.0431$\pm$0.0003,
389: H.E.S.S. 0.0445$\pm$0.0001. The H.E.S.S. measurement is 40 $\mu$s
390: wider than that from HST, consistent with the expected smearing
391: introduced by imperfect ephemerides over the long exposure
392: time. Indeed, fits to all 489 individual 5 minute phasograms show a
393: mean value of 0.0432$\pm$0.0001, identical to the HST result. In our
394: data the phase position of the main optical pulse precedes that of the
395: radio peak by 134$\pm$2 $\mu$s, again consistent with previous
396: measurements (100~$\mu$s \cite{SHEARER}) within the error introduced
397: by the imperfect radio ephemeris.
398:
399: \begin{figure}[h]
400: \begin{center}
401: \mbox{\epsfig{file=figure5.eps, width=1.0\textwidth }}
402: \end{center}
403: \caption{ Phasogram extracted from the complete 39 hour dataset.
404: The main phase regions are marked: the bridge emission (0.1-0.25),
405: the inter-pulse (0.25-0.55), the main pulse (0.9-1.1) and the off
406: pulse region (0.55-0.9). The inset shows the 3\% of
407: the phase around the main peak position on an expanded scale.
408: The 100~$\mu$s scale bar in the inset corresponds to the shaping
409: time of the measurement, any structures in the light-curve on
410: shorter timescales are unresolvable.
411: }
412: \label{fig:Phasogram}
413: \end{figure}
414:
415: The signal measured with our instrument corresponds approximately to the
416: wavelength range 300-650~nm, with a rather non-uniform response
417: illustrated in figure~\ref{fig:Wavelength} (more details on the individual optical
418: components can be found in~\cite{HESSOPT}).
419: The most meaningful way to compare the pulsed flux measured here with
420: previous measurements is to convolve the measured Crab pulsed spectrum
421: with our wavelength dependent response function. Using the spectrum
422: reconstructed by~\cite{HUBBLE} we predict a pulsed signal of 1000
423: photoelectrons per pulse (at 46$^{\circ}$ zenith).
424: The mean measured value is 1200 p.e./pulse, in good agreement within
425: the $\approx$30\% systematic error introduced by uncertainties in
426: instrumental reflectivities, efficiencies and atmospheric transmission.
427:
428: \subsection{Search for Giant Pulses}
429:
430: Extreme pulse height variability is a common feature of radio pulsars.
431: Giant radio pulses (GRPs) are normally coincident with main pulse and inter-pulse ~\cite{SHEARER}
432: but can apparently also occur in other phase
433: regions~\cite{JESSNER}. While an analogous phenomenon has not so far been
434: seen in the optical, an average 3\% enhancement in flux of the optical
435: pulse, in coincidence with GRPs, has been reported~\cite{SHEARER}.
436: Given the signal/noise of our measurement a few $\times 10^{4}$
437: simultaneous GRPs would be required to confirm this value.
438: However, given the relatively long duration of
439: our measurement it is useful to derive a limit on the rate of
440: \emph{large} optical pulses, independent of the radio pulse height.
441:
442: The major background for such a search in our dataset is meteorites. These are normally
443: detected in several pixels and take $\approx$50~ms to cross the
444: camera. Such events, as for example that shown in
445: figure~\ref{fig:Meteorite}, are readily rejected from the analysis
446: using the outer pixel information.
447: However, an irreducible background of events travelling almost on-axis
448: and hence illuminating only the centre pixel, is present.
449: Figure~\ref{fig:AmplitudeDists} shows the distribution of integrated
450: signal in each pulse (or on- and off- phase regions) normalised by the
451: expected poisson fluctuations.
452: No fluctuations $>$ 20$\times$ the mean pulse amplitude were observed
453: during our measurement, neither on- nor off-phase. We can therefore
454: place a limit on the fraction of such large pulses of $1.4 \times\,10^{-6}$
455: (95\% confidence).
456: This result should be compared to the situation at radio frequencies
457: where pulses exceeding the mean flux by several orders of magnitude
458: are frequent.
459: We note that our instrument is sensitive to
460: giant pulses of very short ($<$ 1 $\mu$s) duration, as seen in
461: the radio band, which may not be the case for photon counting
462: instruments.
463:
464: \begin{figure}[ht]
465: \begin{center}
466: \mbox{\epsfig{file=figure6.eps, width=1.0\textwidth }}
467: \end{center}
468: \caption{Light-curve in three pixels indicating the passage of a
469: meteorite candidate. The central pixel (upper curve) shows a narrower
470: profile than that of the outer pixels (lower curves) due to its reduced aperture.}
471: \label{fig:Meteorite}
472: \vspace{3mm}
473: \end{figure}
474: \begin{figure}[h]
475: \begin{center}
476: \mbox{\epsfig{file=figure7.eps, width=1.01\textwidth }}
477: \end{center}
478: \caption{Relative fluctuations in the integrated signal in on- and
479: off-phase regions for each pulse (linear scale - left and log. scale
480: - right). The on-phase region corresponds to the main pulse and
481: inter-pulse regions shown in figure~\ref{fig:Phasogram}. Off-pulse
482: here corresponds to a combination of the true off-pulse
483: region and the bridge region as shown in figure~\ref{fig:Phasogram}.
484: After tight quality selection to
485: remove optical transients and unstable weather, 2.2 million pulses
486: remain. The off-phase distribution follows closely a normal
487: gaussian distribution (dashed-line) as expected, but with a
488: significant tail to large values. This tail can be attributed to
489: transient events occurring only in the central pixel. A similar
490: tail is seen on the on-phase distribution.
491: }
492: \vspace{3mm}
493: \label{fig:AmplitudeDists}
494: \end{figure}
495:
496: \section{Summary and Outlook}
497:
498: We have constructed and tested an optical pulsar monitoring system for
499: installation on Cherenkov telescopes. We have shown that such a
500: system can approach the sensitivity of conventional optical telescopes
501: (with custom built cameras) in measuring short-time-scale
502: (millisecond-second) optical pulsations. In addition, the excellent
503: agreement on the shape of the Crab pulsar light-curve validates the
504: timing hardware and software used for H.E.S.S., thus demonstrating the
505: validity of pulsed emission limits from
506: H.E.S.S~\cite{pulsedlimits}. We also derive an upper limit on the
507: frequency of \emph{giant pulses} which is complementary to existing
508: limits and measurements in that it includes also very short ($\sim$
509: nanosecond) pulses.
510:
511: As the energy threshold of the IACT technique are pushed down further
512: (e.g. with the second phases of both the H.E.S.S. and MAGIC
513: experiments), the detection of pulsed VHE $\gamma$-ray emission will
514: become more likely, thereby increasing the importance of optical monitoring
515: devices. We consider two possibilities for future instruments of this
516: type. Firstly a small device for monitoring purposes, gathering data
517: in parallel to $\gamma$-ray with the Cherenkov camera. Ideally such an
518: instrument would replace one pixel of the Cherenkov camera. The
519: problem with this approach is the observing strategy of modern
520: instruments, for example 'wobble' mode and convergent pointing, both
521: of which move the target source away from the centre of the field of
522: view of the telescopes. Field rotation then prevents the observation
523: of a source in a signal pixel through an entire run. The second
524: possibility is to adapt the photo-sensor current monitoring of a
525: Cherenkov camera to provide high rate ($>$1 kHz) and resolution
526: sampling of the sky brightness. With integration and sampling rates on
527: comparable time scales such a camera could be used to measure optical
528: pulsations (and transients) anywhere in the $\sim\,4^{\circ}$ field of view of the
529: instrument, in parallel to $\gamma$-ray observations. A camera with these properties is currently under
530: test~\cite{SPC}. For the 600~m$^{2}$ class telescopes of the
531: next generation,
532: deep observations of young pulsars with such a system could lead to
533: the discovery of new optical counterparts to known radio pulsars. Such
534: a system is also desirable in the search for MeV/GeV emission from
535: short time-scale GRBs~\cite{SGARFACE}.
536:
537: \section*{Acknowledgements}
538:
539: The authors would like to acknowledge the support of their host
540: institutions, and additionally support from the German Ministry for
541: Education and Research (BMBF). We appreciate the excellent work of the engineering and
542: technical support staff in Heidelberg and Namibia in the construction and operation of the
543: equipment. We would like to thank the entire H.E.S.S. collaboration
544: for their cooperation and assistance, in particular: F.\ Breitling,
545: K.\ Bernl\"ohr, O.\ Bolz, S.\ Gillessen, M.\ Holleran.
546:
547: \begin{thebibliography}{}
548:
549: \bibitem{EGRET_PULSARS} Nolan, P.L., {\it et al.}, 1996,
550: Astron. Astrophys. Suppl. Ser. \textbf{120}, 61.
551: \bibitem{POLARCAP} Daugherty, J.K. \& Harding, A. K., 1982, ApJ {\bf 252}, 337.
552: \bibitem{OUTERGAP} Hirotani, K., 2001, ApJ {\bf 549}, 495.
553: \bibitem{EGRET_VHE} Thompson, D. J. {\it et al.}, 2005, ApJ Suppl., \textbf{157}, 324.
554: \bibitem{HESSproject} Hinton, J.A., 2004, New Astron. Rev. \textbf{48}, 331.
555: \bibitem{MAGIC} Lorenz, E., 2004, New Astron. Rev. \textbf{48}, 339.
556: \bibitem{VERITAS} Weekes, T.C. {\it et al.}, 2002, Astropart. Phys. \textbf{17}, 221.
557: \bibitem{STACEE} Hanna, D. S., {\it et al.}, 2002, Nucl. Inst. \&
558: Methods \textbf{A491}, 126.
559: \bibitem{CELESTE} de Naurois, M., {\it et al.}, 2002, ApJ
560: \textbf{566}, 343. %ALSO describes optical Crab
561: \bibitem{WHYOP} Shearer, A. \& Golden. A, 2002, in
562: \emph{Neutron stars, Pulsars and Supernova remnants},
563: eds W. Becker, H. Lesch \& W. Tr\"umper, MPE report 278, p. 44-53 (astro-ph/0208579)
564: \bibitem{SHEARER} Shearer, A., {\it et al.}, 2003, Science
565: \textbf{301}, 493.
566: \bibitem{ULTRACAM} Dhillon, V. \& Marsh, T., 2001, New Astron. Rev. \textbf{45}, 91.
567: \bibitem{OPTIMA} Straubmeier, C., Kanbach, G. \& Schrey, F., 2001,
568: Exp. Astron. \textbf{11}, 157.
569: % WHIPPLE CRAB PULSAR (includes optical)
570: \bibitem{WHIPPLECRAB} Lessard, R.W., {\it et al.}, 2000, ApJ \textbf{531}, 942.
571: % STACEE (pascals thesis)
572: \bibitem{STACEE_OP} Fortin, P., 2005, Ph.D. Thesis, McGill University.
573: %\emph{Observations of the Crab Nebula and pulsar in the optical and gamma-ray bands with STACEE},.
574: \bibitem{CT1CRAB} O\~na-Wilhelmi, E. {\it et al.}, 2004, Astropart. Phys. \textbf{22}, 95.
575: \bibitem{MAGIC_OP} Lucarelli, F. {\it et al.}, 2005, Proc. 29th ICRC (Pune), \textbf{5}, 367.
576: \bibitem{HESSOPT} Bernl\"ohr, K. {\it et al.}, 2003, Astropart. Phys. \textbf{20}, 111.
577: \bibitem{HESSOPT2} Cornils, R. {\it et al.}, 2003, Astropart. Phys. \textbf{20}, 129.
578: \bibitem{FRANZEN} Franzen, A. {\it et al.}, 2003, Proc. 28th ICRC (Tsukuba), Univ. Academy Press, Tokyo. p. 2987.
579: \bibitem{HESSCAMERA} Vincent. P. {\it et al.}, 2003, Proc. 28th ICRC (Tsukuba), Univ. Academy Press, Tokyo. p. 2887.
580: \bibitem{HESSTRIGGER} Funk, S. {\it et al.}, 2004, Astropart. Phys. \textbf{22}, 285.
581: \bibitem{JESSNER} Jessner, A. {\it et al.}, 2005,
582: Advances in Space Research, \textbf{35}, 1166.
583: \bibitem{SGPhD} Gillessen, S. 2004, PhD Thesis, University of Heidelberg.
584: \bibitem{Till} { Eifert, T. 2006, Diploma Thesis, Humbolt University, Berlin.}
585: \bibitem{TEMPO} { \it http://pulsar.princeton.edu/tempo/}
586: \bibitem{Jodrell} Jodrell Bank pulsar group page, {\it
587: http://www.jb.man.ac.uk/$\sim$pulsar/}
588: \bibitem{HUBBLE} Percival, J.W., {\it et al.}, 1993, ApJ \textbf{407}, 276.
589: \bibitem{HESS_NSB} Preuss, S. {\it et al.}, 2002, Nucl. Instrum. Meth. A \textbf{481}, 229.
590: %%\bibitem{HUBBLE2} Sollerman, J, {\it et al.}, 2000, ApJ \textbf{537}, 861.
591: \bibitem{pulsedlimits}
592: Schmidt, F. {\it et al.}, 2004,
593: 2nd Int. Symp. on High Energy Gamma Ray Astronomy, Heidelberg, APS Conf. Proc. \textbf{745}, 377
594:
595: \bibitem{SPC} Hermann, G. {\it et al.}, 2005, Proc. Cherenkov 2005,
596: Paris ({\it astro-ph/0511519}).
597: \bibitem{SGARFACE} LeBohec, S., Krennrich, F \& Sleege, G. Astropart. Phys. \textbf{23}, 235.
598:
599: \end{thebibliography}
600:
601: \end{document}
602:
603: