Extending the Sensitivity Of Air-Cerenkov Telescopes

Steve Biller, Oxford University

(de la Calle & Biller – astro-ph/0602284)

Most effort in ACTs today is generally directed at lower energies (~200 GeV) using small pixels and, typically, relatively narrow fields of view

What about higher energies and wider fields of view ??

IR Background ~5-20 TeV gammas probe IR regime of interestTests of Lorentz Invariance want high energies & moderate redshiftAGN Jet Dynamics spectral shape & variability at high energiesTeV Sky Survey wide field of view perhaps as important as threshold

Gamma-ray Induced Shower

~ 10km a.s.l.

First Interaction

Shower Maximum

Observation Level

~120m

Cerenkov Light Pool

10TeV, 15Tev, 20TeV

500GeV, 1Tev, 2TeV10 Km

5 Km

100 400300 500200

SH

OW

ER

AX

IS

X (m)0.57-1.14 2.29-4.571.14-2.29 1.72-3.43 2.86-5.71(°) :

• Optical Reflector:

Mount: Davis-Cotton design 10 m diameter

Mirror Area: ~87 m2 271 hexagonal facets Focal length: 10 m (f1 optics)

Simulation of Single Facets Ray Tracing Facet Shape : Hexagonal Facet Area : ~0.32 m2

Facet Separation: 0.61 m Facet Diameter : 0.61 m Facet Reflectivity Misalignment : YES

0.61 m

Mirror Facet

• Camera:

935 pixels (PMT) 0.30 /PMT 10 FOV QE according to wavelength

Light Cones

QADCs and TDCs included in the simulations

Light Cones Ray Tracing Hexagonal + Straight PMT Separation : 0.052 m PMT Radius : 0.026 m (0.15 °) Cone Angle : ~17.3° Cone Height : 0.0125 m Cone Reflectivity : 80%PMT

Cone

PMT

Cone

0.052 m

0.046 m

0.01

25 m

0.00

12 m

17.3°

Length

Width TSlope

Alpha

2 TeV Gamma-ray shower

(0° zenith angle, 125m core distance, using only pixels with > 7 phe)

Arr

ival

tim

e of

fir

st p

hot

on (

ns)

Radial distance on camera (m)

[mi – mio(E,R)]2

io(E,R)]2

i=1

3

Fit for E and R as a function ofDisplacement, Summed light and TSlope/D

j

[mi – mio(E,X,Y)]2

io(E,X,Y)]2

i=1

4

j

arrayj=1

n

10 Km

5 Km

100 400300 500200

SH

OW

ER

AX

IS

X (m)0.57-1.14 2.29-4.571.14-2.29 1.72-3.43 2.86-5.71(°) :

r

dhmax

t = (1/c)(hmax2 + r2)

t (r22 – r1

2)2 c hmax

shapeorientation

time-based “depth”

(independent by construction)

50 hours

Integral Sensitivity

1 hour

Integral Sensitivity

1) That the simulations are accurate and that the analysis techniques applied are valid. In this regard we have verified the simulation by comparing

with analytical calculations, alternative simulations and actual experimentaldata. We have been able to reproduce trends seen in other, independentstudies and replicate parameter distributions and published sensitivity curves from existing experiments.

2) That the repeated sampling of showers employed does not lead to significant biases. To this end, we have explicitly tested this with regard to image selection and have found no evidence of any bias to within the limits our statistical uncertainties. Furthermore, throughout this analysis, we have specifically checked to insure that resulting distributions were not unduly influenced by a handful of independent showers with unusually high ``weights.'‘

3) That the background rejection factor for the tandem WFOV designcan be factorised into groups of largely independent contributionswhich can be separately assessed and combined as a product. However,in addition for there being logical arguments for why this ought to bethe case for the parameters used, we have also explicitly verified the independence of these parameters to within the limits of our statistical uncertainties.

4) That the predicted rejection factors due to newly introduced timing parameters are accurate, even though this is yet to be experimentally tested. We find no reason to doubt these factors given that the photon timing is largely governed by

basic air-shower development and geometry, though we certainly encouragefuture experimental efforts to explicitly explore this.

Assumptions: (the small print)

These results predict the 2-telescope design considered here to be more than 3 times more sensitive than existing/planned arrays in the regime above 300 GeV forcontinuously emitting sources; up to 10 times more sensitive for hour-scale emission; significantly more sensitive in the regime above 10 TeV; and possessing a skycoverage which is roughly an order of magnitude larger than existing instruments.

Conclusions: (the big print!)

Next Step: Bigger mirrors, even wider field-of-view, investigate simple array configurations.

Controversial Proposals:

3) Should work towards a prototype via modification of one of the HESS telescopes.

1) Telescope performance has NOT been optimised in current instruments. This should be done before designing large arrays.

2) Work towards any “all-sky” instruments should start by actually building at least one wide-angle instrument with conventional methods before looking at radical solutions.

Also note: Most of the information we’ve obtained from AGN in the TeV regime has come from data taken during strong flares

Not so much background limited!(effective area could buy more than rejection)

Note: Lower intensity gamma images means we’re competing with lower energy background showers

Must improve rejection by at least an order of magnitude to be in the ball park!

EAS Simulation Code*Observation Level: 2300 m a.s.l. (763 g/cm2, same as Whipple Observatory)

XCORE (m)

YC

OR

E (

m)

800m

Distribution of Core Locations

TELESCOPE

GGammasammas:: ~ 5x106 effective Showers 300GeV - 20TeV Crab-like spectrum = 1.5 (Integral) 0 zenith Angle Core Distance < 800m

PProtonsrotons:: 101066 – 10 – 1077 effective Showers effective Showers 200GeV - 20TeV = 1.7 (Integral) 0- 15° zenith Angle Core Distance < 800m

*by S Biller, based on EGS4 & SHOWERSIM

Whipple 10m Crab Data EAS Simulations

Length Width

Distance Alpha

Simulations of the Whipple 10m gamma-ray telescopeSC2000

Cherenkov Signal : ADC i (d.c.) ≡ # detected photoelectrons (pe)

Telescope Trigger Condition : at least 2 pixels 20 pe

Image Cleaning : RMSNoise = 3 pe Picture Pixel : Signal is 4.25 above RMSNoise (from night sky background) Boundary Pixel: Signal is 2.25 above RMSNoise and borders a picture pixel

Image Parameterization : Number of picture pixels in the image 5 Image Moment parameters (Hillas 1985)

Background Rejection : Size and distance dependent cuts on: Length(s,d), Width(s,d), Alpha(s,d), TSlope(s,d)

so as to arbitrarily keep 95% after each individual cut