Gammaray Large Area Space Telescope Update on Beamtest
Gamma-ray Large Area Space Telescope Update on Beamtest 06 CU PSF study • C. Cecchi • S. Germani • M. Pepe • Università di Perugia and INFN Beam Test Worksop II May 15 -17, 2006
Beam Test Worksop 2 Pisa, May 16 th, 2006 Simulated Data Beamtest. Release v 1 r 0801 p 0 – PS simulation from beamtest 06 v 4 r 0 (updated) – 4. 5 M e- generated with E=2. 5 Ge. V /angle(0°, 20°, 40°) – 4. 5 M e- generated with E= 1 Ge. V /angle – 800 k e- from SLAC pipeline /angle(0°, 40°) 40° • Analyzed using Gleam v 6 r 12 20° hit the towers at 3 different Beam Angles (0° and 40° are the same as Mass Production (SLAC)) C. Cecchi S. Germani M. Pepe 0° 2
Beam Test Worksop 2 Pisa, May 16 th, 2006 Basic Event Selection We used the Merit ntuple after adding PS variables - good. Cal = Evt. Energy. Corr>20 && Evt. Energy. Corr<Beam. Energy -good. Num. Tracks = Tkr. Num. Tracks > 0 -good. Tag = e_silicon[i] > 0 i=0, 3 e_silicon[0] e_silicon[1] e_silicon[2] e_silicon[3] C. Cecchi S. Germani M. Pepe 3
Beam Test Worksop 2 Pisa, May 16 th, 2006 Vertex and Track • Good Vertex Merit Variable Vtx. Status = 34 : 2 tracks vertex + two tracks share first hit 162 : 34 + DOCA location lies inside track hits else use only Best. Track After cuts we have roughly the same number of events in each of the two categories C. Cecchi S. Germani M. Pepe 4
Beam Test Worksop 2 Pisa, May 16 th, 2006 Energy and Direction • Reconstructed Energy compare Evt. Energy. Corr Tag. Energy (Beam. Energy – E_rec) • Photon Direction derived using first two tag detectors (Ptag) • PSF defined from angular error: – good Vertex acos(Ptag·PVtx) – best Track acos(Ptag·PTkr 1) C. Cecchi S. Germani M. Pepe 5
Beam Test Worksop 2 Pisa, May 16 th, 2006 Tagger Silicon Detector C. Cecchi S. Germani M. Pepe 6
Beam Test Worksop 2 Pisa, May 16 th, 2006 Angle Error (Vtx Scaled) 2 Ge. V (SLAC) 2. 5 Ge. V 1 Ge. V C. Cecchi S. Germani M. Pepe 7
Beam Test Worksop 2 Pisa, May 16 th, 2006 Mc. Energy (e- 2. 5 Ge. V) Mc. Energy =ΣMc. Eγ Multi. Gamma Any M_gam One Gamma M_gam==1 0 deg 40 deg The Multi. Gamma effect is small we can neglect to distinguish between 1 and multigamma events C. Cecchi S. Germani M. Pepe 8
Beam Test Worksop 2 Pisa, May 16 th, 2006 Energy Rec. Comparison 1 Gamma e- 2. 5 Ge. V 0 deg 40 deg Resolution of the tagger worse at low energies wrt Evt. Energy visible in the scaled PSF C. Cecchi S. Germani M. Pepe 9
Beam Test Worksop 2 Pisa, May 16 th, 2006 Tagger Energy e- 2. 5 Ge. V Tagger Energy = Beam. Energy–E_rec 0 deg C. Cecchi S. Germani M. Pepe 40 deg 10
Beam Test Worksop 2 Pisa, May 16 th, 2006 Evt. Energy. Corr 20 deg e- 2. 5 Ge. V 0 deg 40 deg C. Cecchi S. Germani M. Pepe 11
Beam Test Worksop 2 Pisa, May 16 th, 2006 Mc. Dir. Err – Measured Dir. Err e- 2. 5 Ge. V Vtx Events C. Cecchi S. Germani M. Pepe 12
Beam Test Worksop 2 Pisa, May 16 th, 2006 Tagger Selection (4 Si vs 2 Si) e- 2. 5 Ge. V 0 deg Requiring e_silicon>0 (signal) only from the first two Tagger Planes does not affect the PSF ! C. Cecchi S. Germani M. Pepe 13
Beam Test Worksop 2 Pisa, May 16 th, 2006 Electron Beam 2. 5 Ge. V Vtx Events Tkr Events C. Cecchi S. Germani M. Pepe 14
Beam Test Worksop 2 Pisa, May 16 th, 2006 SLAC Mass Production vs 2. 5 Ge. V C. Cecchi S. Germani M. Pepe 15
Beam Test Worksop 2 Pisa, May 16 th, 2006 e- 2. 5 Ge. V – 2 Ge. V – 1 Ge. V C. Cecchi S. Germani M. Pepe 16
Beam Test Worksop 2 Evaluating PSF Error Pisa, May 16 th, 2006 PSF 68 68% • The number of events within 68% is N 68 • The error on N 68 is ΔN 68=sqrt(NTOT× 0. 68×(1 -0. 68)) [Binomial] • ΔPSF 68 is obtained computing the quantiles corresponding to a number of events equal to N 68 ±ΔN 68 C. Cecchi S. Germani M. Pepe 17
Beam Test Worksop 2 Pisa, May 16 th, 2006 PSF Error one beam energy vs all Vtx 0 deg Goal Error Level C. Cecchi S. Germani M. Pepe 18
Beam Test Worksop 2 Pisa, May 16 th, 2006 Conclusions • 1 and 2. 5 Ge. V e- simulated on PG-farm 27 M analysed • 2 Ge. V SLAC production 1. 6 M (e-) analysed • • No difference Tag energy Simulated and Selection with selecting 1 or M difference only on scaled PSF at low energies reconstructed energies gives same results on PSF only the first two silicon higher photon energy • Is 800 K e-/angle/energy enough to reach 1% error on PSF? No at low energies. . . At higher energies may depend on true rejection • Cross check results using only not scaled PSF C. Cecchi S. Germani M. Pepe 19
- Slides: 19