Status of CELESTE Cherenkov Low Energy Sampling

1
Status of CELESTECherenkov Low Energy Sampling
Timing ExperimentE. Paré et al., 2002, Nuclear
Instruments and Methods, A490, 71-89
12th May 2003Centre de Physique de Particules de
Marseille
Javier Bussons Gordo, Groupe d'Astroparticules de
Montpellier (France)
2
Detector configuration
Heliostats 40, now 53 x 54 m2 40 cable- and 13
radio-controlled Secondary optics 3 sets of
mirrors (near, middle, far field) one PMT per
heliostat Intermediate electronics pre-amps,
current read-out Counting house
electronics Trigger, Flash ADCs Master Run
Controller (config, database) currents,
heliostats, weather, FADC, Lynx, Event Display
3
Detection principle sampling and timing
Pulses (1 Ghz FADCs)
Intensities
Arrival times
4
Trigger
Trigger rate
Cherenkov pulse (a few ns) S 3 p.e. /
heliostat at 30 GeV Night sky background B 1
p.e. / heliostat / ns Trigger multiplicity
3/5, now 4/6 groups above a given threshold
4 p.e. / heliostat Helps reject
hadrons Rate 20 Hz (mostly from cosmic
background) Must take care of electronics dead
time
Threshold setting
Jan 2002 40 heliostats (5 groups) gt 53
heliostats (6 groups)
5
Energy Threshold and Collection Area

Fns Omega T
Eth sqrt(B) / S sqrt
A EFFpm,opt
2
-5
Omega pi RFOV 8 x 10 sr
RFOV 5 mrad CELESTE lowers threshold via
large A
(PHYSICAL) COLLECTION AREA CELESTE 53 x 54
2800 m2 STACEE 64 x 39 2500 m2 CAT 16 m2
HESS 4 x 113 450 m2
Inconvenients timing, optical throughput,
acceptance evolve during source tracking
6
Veto
Pointing configurations SINGLE all heliostats
at 11 km DOUBLE half at 11 km, half at 25 km
now VETO 41 at 11km and 12 as veto Normal
heliostats point at shower max Veto heliostats
point 150m to the side Better hadron rejection
at lower energies
d150m
7
Crab Nebula detectionM. De Naurois, Holder et
al., ApJ 566, 343-357 (2002)
integral rate on a differential plot
8
Pulsars Emmanuel Durand's thesis approved
January 2003
7 EGRET pulsars, none seen by IACTs 6 seen at
Egt100MeV (3 in N hemisphere) Cutoffs expected at
10 100GeV Polar cap (super-exponential cutoff
model) Outer gap (exponential cutoff model) E0
polar cap lt E0 outer gap
9
RUN SELECTION Duration gt 7 mins (most runs 18
mins) Rate Crab PSR1951 3/5 20 - 30 Hz 12 - 17
Hz 4/5 6 - 16 Hz 5 - 10 Hz Rate stability Hour
angle -2 h lt H lt 2 h Remaining data
(Crab) Set 1 SP, trigger 3/5 48 runs (16.02
h) Set 2 SP, trigger 4/5 21 runs (7.78 h) Set
3 DP, trigger 3/5 29 runs (8.91 h) Remaining
data (PSR1951) Set 1 SP, trigger 3/5 22 runs
(6.66 h) Set 2 DP, trigger 3/5 19 runs (5.93
h) Set 3 DP, trigger 4/5 24 runs (7.71 h) SP
DP single double pointing
Crab PSR1951 Distance kpc 2 2.5 Perio
d ms 33.4 39.5 Age yrs 949 10000 L
1034 erg/s 44 15 B 1012 G 3.8 0.5 Obj
ect SNR Crab Nebula CTB80
OPTICAL CRAB Check barycentering, phase
calculation and GPS timing Optical peaks aligned
with radio peaks gt OK
10
Cuts reject high E events with mean(S) reject
hadrons with timing, light sampling no
good eff(MC gammas) eff(real hadrons) Q
factor (Elt50GeV) mean(S) lt 15 pe /
heliostat 100 55 1.3 Elt50GeVmax(residual) lt
5ns 100 35 1.7 both cuts 100 15 2.7
11
Crab results
Further cut in hour angle -3/4h lt h lt 3/4h Set
1 signal but lt 5? Compatible with p1ltp2 w/bridge
2-100keV, 100keV-10MeV Incompatible with
p2gtp1 no bridge 0.5-2keV, gt100MeV EGRET
favourable P2, bridge, TW1, LW2 Profile
compatible with EGRET extrapolation Assume flux
from this lot lt 2 sigma expected in other
lots, result not incompatible with result
sets 2,3 Need to master acceptance at low
energies atmosphere effects Set 2 no
signal Set 3 no signal Upper limit (SP, 3/5, H
cut) AE gt A (1? 0.30)E 0.7 most
conservative f phase interval, e dead
time B1 (Outer Gap) or B2 (Polar Cap) Integrate
and find E0 (cutoff energy)
Inf
Nul
A(E) k E-g e-(E/Eo)dE
B
T f e
0
12
PSR1951 no signal
13
UL compatible with both models
E0lt57GeV exp
E0lt80GeV exp
E0lt75GeV super-exp
E0lt62GeV super-exp
14
Crab Pulsar (Limit or Flux?) (comparing ApJ to
ED's thesis)
E0lt 26GeV ApJ
LIMIT
Exponential cutoff model 14 lt E0 lt 57GeV for
A(1.3E) gt A gt A(0.7E) E0 25GeV at nominal A(E)
2.7 ? /min 5.4 ? /min instantaneous Super-expo
nential cutoff model 19 lt E0 lt 60GeV for
A(1.3E) gt A gt A(0.7E) E0 31GeV at nominal A(E)
FLUX
15
University of Tokyo Workshop 2002 - September
25-28, 2002, Kashiwa (Japan)
Energy reconstruction method
Mean charge (p.e.)
Smean collected charge over 40
heliostatsdimpact parameter (/-15m) at
11km Energy estimate f (mean collected charge,
impact parameter) Simulate EAS 30GeV -
1TeVlog(S)a1a2 log(E) for a given d gt
for d50md'd - 50 E50 / E 1 c
d' for 20m lt d lt 80m (85 of gammas) The
slope happens to be c c1 (1/ E) c2 and
therefore
Ea Sb
_ a Sb-c1d'
Impact parameter (m)
Em
1c2d'
16
Energy resolution
E resolution varies smoothly with E Unbiased
above threshold 12 lt (Em - E) / E lt 24
Spectrum derivationEnergy bins 50 - 100 - 200
- 400 GeVSeparate acceptance from energy resol
Max likelihood fit (over all events) yield
s mean true energy in each bin. Assume
power-law spectrum Iterate to find g Method
validated on various test indices
f(EmE0)dE dP/dE (1/sqrt(2pi s2))
x exp(-(Em-E)2 / 2 s2)
dP/dEP0 exp(-E/E0)
A(E) Eg dE
ON-OFF
A
F
A T dE
Eg dE
17
Crab Nebula Egt50GeV
Cuts 4/5 majority peaks gt 10
18
Mrk421 Egt70GeV
2001
2000
19
Seasonal effects
Rate vs DewPoint
o Mrk421 Mrk501
We understand trigger rate variations large for
Mrk501 (warm, humid season) small for Mrk421
(cold, dry season)
Rate vs Temperature
Rate vs Pressure
Rate vs Elevation
20
ATMOSPHERELIDAR CCD photometry radiometry
pseudo-photometry (star track runs)Monte Carlo
parameters atmospheric profile, absorption table
height
Radiosounding data gt customised profiles by month
August mid-latitude summer
STREAMER transmission vs CCD measurements
Cherenkov light density
January mid-latitude winter
Rayleigh
Rayleigh Mie
Overburden
Uncertainty in estimated energy and
acceptance due to choice of atmosphere Aerosol
distrib at Themis (boundarylayer)
Core distance
21
Laser Nd-YAG rep rate 10 Hz ?1064, 532, 355
nm pulse width ?4ns pulse E180mJ
(532nm) 70mJ (355nm) beam divergence
0.5mrad Optics Cassegrain dichroic
splitter primary mirror Ø 60cm
,f102cm secondary Ø 8cm, f10cm
reflectivity(300-600nm) 80 PMTs Philips
XP2020 Acquisition 12-bit, 5000-word card, 2
inputs 25 Mhz (40ns/sample) Pointing azimuth
0-359 elevation 36-89 encoder step 0.14
LIDAR
_ http //doc.in2p3.fr/themis/LIDAR
22
PREDICTIONS (Cherenkov light density on ground)
Attenuation (50GeV photon) 25 40 70
std atmosphere aerosols clouds
Seasonal variations 20 at mid-latitude
sites GOALS Select good quality data Model
Thémis atmosphere with CORSIKA (extinction)
Correct for seasonal variations
?????extinction (scattering absorption) ?????bac
kscattering
LIDAR EQUATION (?????unknowns)
P Po Leff O b(R) T2
R
gas particles Rayleigh Mie
TI/I0exp(-a(r) dr)
Transmittance
0
R range O ?solid angle Leff c t ? ?
Nightly shots, quicklook by observer (sky
quality, run selection)
FFT-cleaned, solid-angle-corrected
Signal in lidarists form
Slope Method
range(km)
23
G. Debiais, B. Fabre (Univ. Perpignan) Kramers-K
ronig analysis (nik) Two-colour difference
(aerosol response)
Refrective index
green
violet
Optical index (absorption)
Aerosols 0 - 10 km
24
Flux predictions for neutralino annihilation A.
Falvard et al., submitted to Astroparticle
Physics, astro-ph/0210184 Supersymmetric dark
matter in M31 can one see neutralino
annihilation with CELESTE?
The DRACO case Dwarf spheroidal M/L
200 November 2001 evidence for extended DM
halo April 2002 observation proposal with
CELESTE Possible detection if DM spike near
black hole
25
MY MESSAGENew results on Crab pulsar marginal
detection cut-offOther pulsars PSR1951
cut-offTwo spectra Crab (Egt50GeV)
Mrk421(Egt70GeV)Operating at 50 GeV (hit by bad
weather)Working on understanding our
systematicsAttention to best shots next
seasonReady to shut down in July 2004
MY MESSAGENew results on Crab pulsar marginal
detection cut-offOther pulsars PSR1951
cut-offTwo spectra Crab (Egt50GeV)
Mrk421(Egt70GeV)Operating at 50 GeV (hit by bad
weather)Working on understanding our
systematicsAttention to best shots next
seasonReady to shut down in July 2004
26
THE END