1. EAS Reconstruction with Cerenkov Photons Shower Simulation Reconstruction Algorithm Toy MC Study Two Detector Configuration Summary M.Z. Wang and C.C. Hsu National Taiwan Univ. Dept. of Physics Jan. 9, 2003 Palermo Workshop

2. CORSIKA-> Reconstruction Air Shower Č photons CORSIKA telesim Reconstruction Input card E, shower axis Read out Č photons and simple detector simulation photon weighting, detectors sampling

3. Event 0819 Primary particle: electron Incident angle: θ=0 Primary Energy : 10 14 eV φ : random Č photon wavelength : 290 ~390 nm Fix air density and n Incident distance: 40 km

4. Detector Configuration. 20 m 4 m 2m 40 X 50 detectors in a 4.5 x 5.0 km 2 area

5. Detector Parameters Focal length : 43 cm Pixel angular Span: 0.5 degree PMT QE : 0.3 Spot size : 0.04 cm < pixel size 0.375 cm 40x40 pixels in a 15x15 cm 2 area Lens area : 2m x 2m Transmittance 0.95

6. Č Photon Image X Y detector area : 10 km x 10km detector simulation area 4.5km x 5.0 km

7. Photon Density on the Ground 10 18 eV 10 16 eV 10 14 eV Peak position 680 – 750 m

8. Č Production Height Higher Energy Closer to the Ground 31-34 Km 10 18 eV 10 14 eV10 16 eV

9. Reconstruction Illustration MINUIT: minimize χ2 to get x,y,θ,φ and E (x2, y2 ) N2. t2   N1, t1 (x1, y1) E, (x, y,θ,φ)

11. Shower Height 10 14 eV

12. Č Photon density Photon density Double Gaussian cone + exponential decay

13. Fit t0 5 km Fit t0 5 km Similar PDF parameters

14. Reconstruction CORSIKA data 10 14 eV ΔxΔx ΔtΔt ΔθΔθ Er/Et

15. Reconstruction by Toy MC 10 14 eV Δθ Er/Et

16. Reconstruction CORSIKA data 10 16 eV Er/Et Δθ

17. Reconstruction CORSIKA data 10 18 eV Poorer results due to PDF and much bigger photon weights? Er/Et Δθ

18. ` d1d1 d2d2 d 1 d 2 Correlation

19. d1d1 χ2χ2 Rejection of Bad Reconstruction

20. Er/Et ΔθΔθ Correlation between Er/Et and Δ θ

21. Study of Two Detector Performance with Toy MC 10 15 eV electron shower One p.e. threshold, no background Shower starts 30 km away Mirror area 1 m 2 Field of view 12 o Pixel angular span 1 o Vary the distance with 100 m step

22. Sum of N p.e. 5km is the best?

23. Efficiency Within 2km, efficiency is flat. →

24. Reconstructed Angular Deviation 1 o can be achieved

25. Angular Resolution 2 o for 2 km separation

26. The Mean Reconstructed Energy Ratio OK, near 1

27. RMS of Reconstructed Energy Ratio OK, less than 100%

28. RMS of Reconstructed Position Deviation 1km for 2 o spread

29. Absolute Time Difference (ns) in Two Detectors 200 ns for separation distance 2km CORSIKA data Hard for coincidence?

30. We can reconstruct EAS (CORSIKA simulation) with Č photons by two detectors We can reconstruct EAS (CORSIKA simulation) with Č photons by two detectors The reconstruction method is simple and can be improved The reconstruction method is simple and can be improved The study done by a Toy MC indicates the detector separation of 3 o -4 o is optimal The study done by a Toy MC indicates the detector separation of 3 o -4 o is optimal Next Step: Next Step: Need to put in background and utilize pixel information for shower reconstruction Need to put in background and utilize pixel information for shower reconstruction Summary

31. Background Estimation incident flux: 200 ph/m 2 /srd/ns 0.12m 2 pupil area, 0.5 o pixel size, 4*64 pixels, 20ns gate width, Q.E.=0.2, 2 p.e. threshold, near-by 2-fold coincidence pixel trigger rate=2R1*R2*W R1=5.4/ms pixel trigger rate = 1.2 Hz local trigger rate = 307 Hz Addendum 1/10/03

32. 2km gate width 2μs with time tag 2nd level trigger rate 0.0015 Hz Offline selection based on pattern recognition (neighboring two cells, etc.) can reduce the rate further Expected acceptance for ν τ =10 16 eV is 38% (70% for 10 16.5 eV) Addendum 1/10/03

33. Tasks for Early March Discussion 1/10/03 Optimize detector configuration: number of detectors, number of pixels Compare: efficiency, angular resolution and energy resolution for 10 15 -10 17 eV neutrinos Č photon generation: Sigl model + CORSIKA Optics: 0.5m diameter or 1.0m diameter with F~1, Gaussian spot RMS 0.5 o Light collection: 10% (including Q.E.) except light guide simulation if needed Two different readout schemes: single ADC mode and 8 by 8 flag + energy sum mode Background should be added in pixel simulation assuming 200 ph/m 2 /srd/ns.