Download CHERENKOV TELESCOPE ARRAY Dainis Dravins OPTIMIZING THE FOR INTENSITY INTERFEROMETRY

Intensity Interferometry workshop – Salt Lake City, January 2009
OPTIMIZING THE
CHERENKOV TELESCOPE ARRAY
FOR INTENSITY INTERFEROMETRY
Dainis Dravins
Lund Observatory, Sweden
www.astro.lu.se/~dainis
CTA, Cherenkov Telescope Array
PRIORITIES IN EUROPEAN ASTRONOMY 2010-2020
ASTRONET Infrastructure Roadmap
http://www.astronet-eu.org/
For the section on High-Energy Astrophysics, Astroparticle Physics and Gravitational Waves,
highest-priority near-term (−2015) project is CTA; in overall list is 2nd highest priority among
medium-scale ground-based projects (following the European Solar Telescope).
ESFPRI , European Strategy Forum on Research Infrastructures
ftp://ftp.cordis.europa.eu/pub/esfri/docs/esfri_roadmap_update_2008.pdf
Eight prioritized projects within Physical Sciences and Engineering, include CTA
ASPERA network on astroparticle physics
http://www.aspera-eu.org/
The priority project for VHE gamma astrophysics is the Cherenkov Telescope Array, CTA.
Cherenkov Telescope Array
www.cta-observatory.org
The CTA Design Study is to optimize the planned observatory
Primary targets of are to constrain design and technology options;
Optimize the cost/performance ratio;
Define how CTA is to be constructed and operated;
Build and test prototype telescope(s)
CTA CONSORTIUM DESIGN STUDY
Approximately four years, 2008-2011
11 work packages
PHYS
Astrophysics and astroparticle physics
MC
Optimization of array layout, performance studies and analysis algorithms [MonteCarlo]
SITE
Site evaluation and site infrastructure
MIR
Design of telescope optics and mirror
TEL
Design of telescope structure, drive and control systems
FPI
Focal Plane Instrumentation
ELEC
Readout electronics and trigger
ATAC
Atmospheric monitoring, associated science and instrument calibration
OBS
Observatory operation and access
DATA
Data handling, processing, management and data access
QA
Risk assessment and quality assurance
PHYSICS WORK PACKAGE
Work Package Coordinator: Diego F. Torres
ICREA & Institut de Ciencies de l'Espai (IEEC-CSIC), Barcelona, Spain
Physics WP topics & Task Leaders
Dark matter / Fundamental physics — Jan Conrad
Extragalactic background light / Cosmology — Daniel Mazin
AGNs — Helene Sol, Catherine Boisson, Andreas Zech
Cosmic rays / Clusters / Starbursts — Olaf Reimer
Microquasars / Binaries — Josep M. Paredes
Cosmic rays / SNRs / Molecular clouds — Stefano Gabici
Pulsar-wind nebulae — Okkie de Jager
Pulsars / Globular clusters — Bronek Rudak
Galactic center — Stefan Funk
Multi-wavelength / Transients / GRBs — Sera Markoff
Timing — Dimitri Emmanoulopoulos
Surveys / Sub-arrays — Guillaume Dubus
Extended / Diffuse Sources — Sabrina Casanova
Intensity Interferometry — Dainis Dravins
Direct-Cherenkov light / CR composition — Rolf Bühler
Cherenkov Telescope Array
CTA General meetings
May 4-5, 2006, Berlin, Germany
March 1-2, 2007, Paris, France
January 24-25, 2008, Barcelona, Spain
November 3-5, 2008, Padua, Italy
May 11-13, 2009, Cracow, Poland
CTA desiderata
Isochronous telescope design ?
Parabolic or Schmidt better than
Davies-Cotton for Δt < few ns
Cherenkov telescopes are usually Davies–Cotton or parabolic
In a Davies–Cotton layout, all reflector facets have same
focal length f, arranged on a sphere of radius f.
In a parabolic layout, mirrors are arranged on a paraboloid,
and the focal length of the (usually spherical) mirror facets
varies with the distance from the optical axis.
Both have significant aberrations off the optical axis,
the parabolic slightly worse than Davies–Cotton.
Time dispersion introduced by the reflector should not exceed
the intrinsic spread of the Cherenkov wavefront of a few ns.
Parabolic reflectors are isochronal – apart from minute
effects caused by individual mirror facets being spherical
rather than parabolic.
Davies–Cotton layout causes a spread of photon arrival times
at the camera; a plane incident wavefront results in photons
spread over Δt ≈ 5 ns, with an rms width ≈ 1.4 ns.
The optical system of the H.E.S.S. imaging atmospheric Cherenkov
telescopes. Part I: Layout and components of the system
K.Bernlöhr, O.Carrol, R.Cornils, S.Elfahem P.Espigat, S.Gillessen, G.Heinzelmann,
G.Hermann, W.Hofmann, D.Horns. I.Jung, R.Kankanyan, A.Katona, B.Khelifi,
H.Krawczynski, M.Panter, M.Punch, S.Rayner, G.Rowell, M.Tluczykont, R.van Staa
Astropart.Phys. 20, 111 (2003)
DAVIES-COTTON SPHERICAL REFLECTOR DESIGN
The Davies–Cotton configuration
forms a focal surface at the center
of curvature of the optical support,
7.3 m from the mirror surface.
A Davies–Cotton layout gives smaller
aberrations off the optical axis
compared to a parabolic design.
A disadvantage is that the structure
is not isochronous.: Rays striking
mirrors at different distances from
the optic-axis have different
transit times to the focal plane.
For the 10 m Whipple telescope the
spread of transit times is 6.5 ns.
The Whipple Observatory 10 m Gamma-Ray Telescope, 1997–2006
J. Kildea et al. , Astropart.Phys. 28, 182 (2007)
Parabolic reflector of MAGIC, Roque de los Muchachos, La Palma
MAGIC has isochronous parabolic reflectors with an intrinsic time spread of 400 ps,
sufficient to resolve the time structure of the cosmic showers
INTRINSIC TIME SPREAD IN 20 m ∅ CHERENKOV TELESCOPES
Top: Spherical (Davies–Cotton)
A spherical reflector substantially
widens the photon pulse.
At detecting 10 GeV γ-showers,
the pulse width on the spherical
telescope's focal plane may reach 15–
20 ns instead of the inherent 5–8 ns.
Angles of incidence = 2°
Bottom: Parabolic
Performance of a 20 m diameter Cherenkov imaging telescope
A.Akhperjanian & V.Sahakian
Astropart.Phys. 21, 149 (2004)
INTRINSIC TIME SPREAD IN 20 m ∅ CHERENKOV TELESCOPES
Top: Spherical (Davies–Cotton)
Dish arrival time and camera arrival
times of photoelectrons initiated by
the photons from 10 GeV γ-showers.
The observation height is 5 km a.s.l.,
and the showers impact distances
are: 50 m (solid), 100 m (dashed), 150
m (dotted) and 200 m (dash-dotted).
Bottom: Parabolic
Performance of a 20 m diameter Cherenkov imaging telescope
A.Akhperjanian & V.Sahakian
Astropart.Phys. 21, 149 (2004)
IACT Schmidt telescope
Diameter 7.0 m
F-ratio 0.8
Focal length 5.6 m
Field of View 15°
Resolution (RMS) < 1′
Non-isochronicity ≤ 0.03 ns
The mirror and the focal plane
have their centre of curvature
at the centre of the corrector
plate.
The Schmidt corrector is
shown with the aspheric shape
magnified by a factor of 20.
Both the nominal corrector
and a Fresnel version is shown.
R. Mirzoyan, M.I. Andersen: A 15 deg Wide Field of View Imaging Air Cherenkov Telescope
Astropart.Phys. (2009) = astro-ph 0806.0297
DIGITAL PHOTON CORRELATORS @ Lund Observatory 2008/09:
700 MHz clock rate (1.4 ns time resolution)
200 MHz maximum photon count rates per channel (pulse-pair resolution 5 ns)
Photon pulses at TTL voltages
High-speed correlators
may be limited by
telescope non-isochronicity
CTA desiderata
Sharper PSF gives less background
Sky brightness:
(a) Dark sky; mV ≈ 21.5 mag / arcsec2
(b) Full Moon; mV ≈ 18 mag / arcsec2
⇒ mV ≈ 9.4 (a) and 5.9 (b) for 5 arcmin ∅
⇒ mV ≈ 12.9 (a) and 9.4 (b) for 1 arcmin ∅
R.H.Garstang: Night-sky brightness at observatories and sites, Publ.Astron.Soc.Pacific 101, 306 (1989)
SKY BACKGROUND COUNT RATES
Expected count rates in HEGRA CT1 (4.2 m ∅, FoV 15 arcmin) and MAGIC I (17 m ∅, FoV 6 arcmin)
”Background” = Crab nebula background + Light Of the Night Sky during dark-sky conditions
Determination of the night sky background around the Crab pulsar using its optical pulsation
E.Oña-Wilhelmi, J.Cortina, O.C.de Jager, V.Fonseca
Astropart.Phys. 22, 95 (2004)
? ? ?
Intensity interferometry
“should” be possible to carry
out in full moonlight
when Cherenkov observations
are not feasible
CTA desiderata
Detectors for huge photon fluxes ?
Photon counting @ 100 MHz – 10 GHz ?
Silicon detector arrays ?
CTA desiderata
Handling high data rates ?
Can photon time-tagging to 1 ns-100 ps
be preserved until a computing location ?
CTA desiderata
Detectors – only central pixel(s) ?
or should one have separate detectors ?
7-pixel camera on the
lid of the H.E.S.S.
Cherenkov camera
A 7-pixel camera was
custom-built and mounted on
the lid of the Cherenkov
camera of a H.E.S.S.
telescope using a plane
secondary mirror to put it
into focus.
Its central pixel was used to
continuously record the
light curve of the target,
while a ring of six ‘outer’
pixels was used both to
monitor the sky background
level and as a veto system
to reject background events
occurring in the atmosphere
Capability of Cherenkov Telescopes to Observe Ultra-fast Optical Flares
C.Deil, W.Domainko, G.Hermann, A.-C.Clapson, A.Förster, C.van Eldik, W.Hofmann
Astropart.Phys., in press (2009) = astro-ph 0812.3966
Observations with a 7-pixel camera mounted on the lid of the Cherenkov camera of a H.E.S.S. telescope
Examples of background events most likely caused by a meteor (left) and lightning (right).
Airplanes, satellites and meteors passing through the field of view produce time-shifted flares.
Lightning at the horizon is scattered in the atmosphere and illuminates all pixels in the same way.
Cosmic-ray induced air showers last a few ns, appear as an elliptical light distribution on the sky over ≈ 30 arcmin.
Capability of Cherenkov Telescopes to Observe Ultra-fast Optical Flares
C.Deil, W.Domainko, G.Hermann, A.-C.Clapson, A.Förster, C.van Eldik, W.Hofmann
Astropart.Phys., in press (2009) = astro-ph 0812.3966
Observations with a 7-pixel camera mounted on the lid of the Cherenkov camera of a H.E.S.S. telescope
An event likely caused by Sun-illuminated space debris, plotted before and after the event in the central pixel.
Assuming h=1000 km, a deduced angular speed of 0.6°/s corresponds to 10 km/s, typical for space debris orbits.
Capability of Cherenkov Telescopes to Observe Ultra-fast Optical Flares
C.Deil, W.Domainko, G.Hermann, A.-C.Clapson, A.Förster, C.van Eldik, W.Hofmann
Astropart.Phys., in press (2009) = astro-ph 0812.3966
CENTRAL PIXEL IN THE MAGIC I TELESCOPE
Support of the central pixel, and a camera rear-side photograph with the PMT installed
The mechanical support holding the PMT at the central aperture position, consists of two parts:
* One part is fixed to the metal support plate (dubbed “Swiss cheese” because of its many holes)
* The second part, containing the PMT, is screwed into the central aperture of the “Swiss cheese” plate
The Central Pixel of the MAGIC Telescope for Optical Observations
F.Lucarelli, J.A.Barrio, P.Antoranz, M.Asensio, M.Camara, J.L.Contreras, M.V.Fonseca, M.Lopez, J.M.Miranda, I.Oya, R.De los Reyes,
R.Firpo, N.Sidro, F.Goebel, E.Lorenz, N.Otte
Nucl.Instr.Meth.Phys.Res.A, 589, 415 (2008)
CTA desiderata
Provision for inserting optical
elements in front of detector
such as focusing lenses or color filters
CTA desiderata
Provision for focusing at infinity
rather than on Cherenkov light in
the upper atmosphere
CTA desiderata
Telescopes in optimal pattern to
cover interferometric (u,v)-plane ?
Important only for larger telescopes,
or will the plane be filled in anyway ?
”OPTIMAL” TELESCOPE PLACEMENTS FOR INTERFEROMETRY ?
Examples of optimization using different criteria:
Top: Noise at the spatial frequency that is most attenuated by the optical system
Bottom: Average noise at all relevant frequencies
Aperture configuration optimality criterion for phased arrays of optical telescopes
L.M.Mugnier, G.Rousset, F.Cassaing
J.Opt.Soc.Am. A 13, 2367 (1996)
ESO
ESOParanal
Cerro Paranal
”OPTIMAL” TELESCOPE PLACEMENTS FOR INTERFEROMETRY ?
Array configurations optimized for Earth-rotation aperture synthesis:
The pupil configuration is at left, and its autocorrelation function at right graph.
Apertures are scaled to provide a full (u, v)-plane coverage during half a rotation.
Circles show the path of the autocorrelation domains during the rotation.
Their regular spacing demonstrates a very good (u, v) coverage.
Aperture Rotation Synthesis: Optimization of the (u,v )- Plane Coverage for a Rotating Phased Array of Telescopes
O.Guyon, F.Roddier
Publ.Astron.Soc.Pacific 113, 98 (2001)
”OPTIMAL” TELESCOPE PLACEMENTS FOR INTERFEROMETRY ?
Array configurations optimized for Earth-rotation aperture synthesis:
The pupil configuration is at left, and its autocorrelation function at right graph.
Apertures are scaled to provide a full (u, v)-plane coverage during half a rotation.
Circles show the path of the autocorrelation domains during the rotation.
Their regular spacing demonstrates a very good (u, v) coverage.
Aperture Rotation Synthesis: Optimization of the (u,v )- Plane Coverage for a Rotating Phased Array of Telescopes
O.Guyon, F.Roddier
Publ.Astron.Soc.Pacific 113, 98 (2001)
IMAGE RECONSTRUCTION ALGORITHM FOR
EARTH-ROTATION APERTURE SYNTHESIS
For each snapshot exposure (no rotation), the
entrance pupil (a) samples a domain (b) of the
Fourier transform of the image.
The image’s Fourier transform information is
extracted by Fourier transform (d) of the snapshot
image (c).
This process is repeated for each rotation step
until the Fourier transform of the image is
completely known up to a cutoff frequency.
An inverse Fourier transform is then performed to
recover an image (e) which can be Fourier filtered
to reduce ringing effects.
Aperture Rotation Synthesis: Optimization of the (u,v ) - Plane
Coverage for a Rotating Phased Array of Telescopes
O.Guyon, F.Roddier
Publ.Astron.Soc.Pacific 113, 98 (2001)
Left: Distribution of interferometer baselines in one possible large-scale array of 81 telescopes placed in a
1 km2 square grid with 125 m spacing. The upper scale indicates the baseline for the first interferometric
minimum for a uniform stellar disk observed at 350 nm.
Right: The two-dimensional baseline distribution, with scales in meters.
S.LeBohec, M.Daniel, W.J.de Wit, J.Hinton, E.Jose, J.A.Holder, J.Smith, R.J.White: Stellar Intensity Interferometry with Air Cherenkov Telescope Arrays, AIP Conf. 984 (2008); D.Dravins & S.LeBohec: Towards a diffractionlimited square-kilometer optical telescope: Digital revival of intensity interferometry, SPIE Proc. 6986 (2008)
Tentative recommendations
Desire also smaller telescopes to be isochronous ( < 1 ns)
Desire Schmidt-type small telescopes for image quality
Provision to mechanically refocus on infinity
Use standard CTA detectors in central pixel(s)
Provision to mount optics in front of central pixel(s)
Provision to mount equipment in front of camera cover
Limit photon count rates with wavelength filters
Photon pulse train precise to < 1 ns to computing station
Avoid placing the large telescopes on a regular grid
“Homework” outside CTA
Prototype instrumentation & test observations
Real-time or off-line correlation ?
Understanding detector noise sources
Understanding noise sources in the sky
Software tracking stars across the sky (LOFAR, MeerKAT ?)
Astrophysical targets and their spectral features ?
Simultaneous observations in different wavelengths ?
Understanding information content in correlation functions
Demonstrate possible ”full” image reconstruction ?
The competition…
Visions towards kilometric-scale
optical imagers
Spatial Resolution
E-ELT
8m
Seeing
+ AO
HST
(diffraction
limited
limit a few
milliarcsec in
the near-IR)
ESO Paranal: Auxiliary telescopes of the Very Large Telescope Interferometer
SHAPE OF ACHERNAR
Image of the rapidly rotating
( Vsin i ≈ 250 km/s )
star Achernar (α Eri, B3 Vpe),
from VLTI observations.
Axis ratio = 1.56, the most
flattened star seen so far.
Because of the projection effect
this ratio is a minimal value;
the star could be even flatter.
Individual diameter measurements
are shown by points with error bars.
The spinning-top Be star Achernar from VLTI-VINCI
A.Domiciano de Souza, P.Kervella, S.Jankov, L.Abe, F.Vakili, E.di Folco, F.Paresce
Astron.Astrophys. 407, L47 (2003)
VLBI maps at two epochs of SiO maser emission around TX Cam.
The color bar gives flux in Jy/beam
J.Yi, R.S.Booth, J.E.Conway, P.J.Diamond
SiO masers in TX Cam. Simultaneous VLBA observations of two 43 GHz masers at four epochs
Astron.Astrophys. 432, 531
COOL SUPERGIANT SIMULATION
Bernd Freytag (Uppsala)
DONUT-SHAPED STARS ?
Models for nonspherical stellar models with differential rotation. From the surface inward, the surfaces enclose
mass fractions of 1.0, 0.995, 0.95, and 0.5. Radiative portions of the interior are white; convective regions are gray.
K.B.MacGregor, S.Jackson, A.Skumanich,T.S.Metcalfe
On the Structure and Properties of Differentially Rotating, Main-Sequence Stars in the 1–2 Mo Range, Astrophys.J. 663, 560
Luciola* Hypertelescope
* genus of fireflies
The Luciola flotilla of many small collector mirrors operates like one giant diluted mirror.
Focal beam-combiners independently exploit the sky image formed at the focal surface.
A.Labeyrie, H.Le Coroller, J.Dejonghe, O.Lardière, C.Aime, K.Dohlen, D.Mourard. R.Lyon, K.G.Carpenter
Luciola hypertelescope space observatory, Exp.Astron., in press (2008) & ESA Cosmic Vision 2015-2025 proposal
Concordia Station
http://www.concordiastation.org/
F.Vakili et al.: KEOPS: Kiloparsec Explorer for Optical Planet Search
Imaging synthesis optical array proposed at Dome C in Antarctica.
KEOPS individual telescopes are grouped around the optical recombiner.
Concordia station is visible in the distance.
CTA, Cherenkov Telescope Array
Which will come first ?
The science cases for kilometric-scale
optical imaging are overwhelming
“Our
local Universe is teeming with stars, but despite 400 years
of telescopic observations, astronomy is still basically
incapable of observing stars as such!
Although we can observe the light radiated by them, we do not
(with few exceptions) have the capability to observe the stars
themselves, i.e., resolving their disks or viewing structures
across and outside their surfaces (except for the Sun, of
course!).
In 2009, we
celebrate
400 years of
telescopic
astronomy
One can just speculate what new worlds will be revealed once
stars no longer will be seen as mere point sources but as
extended and irregular objects with magnetic or thermal spots,
flattened or distorted by rapid rotation, and with mass ejections
monitored in different spectral features as they flow towards
their binary companions.
It is not long ago that the satellites of the outer planets passed
from being mere point sources to a plethora of different worlds,
and one might speculate what meager state extragalactic
astronomy would be in, were galaxies observed as point
sources only.”
(Dravins & LeBohec, SPIE Proc. 6986, 2008)