1. MSSM H 0  3 prong using vertexing at CMS Helsinki b/tau workshop 30.05.-01.06.2002 Lauri A. Wendland / HIP

2. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland2 Outline Motivation Tools & methods Secondary vertex (SV) reconstruction Charge reconstruction Effective mass reconstruction Conclusions Future plans

3. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland3 Motivation If 3-prong  events can be used in addition to 1-prong  decays, a factor of 1.7 of signal events are gained

4. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland4 Tools & methods Software –ORCA_5_4_1 –CMSIM 123 Reconstruction –Lvl1,2 calo selections (L2cut = 3.4 GeV (lowlumi)) –Lvl3 pixel selection (R S = 0.07, R M = 0.1, R I = 0.4) –full tracker reconstruction (Reco) pixel lines from Lvl3 used as seeds at least 8 hits per track demanded

5. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland5 Events produced with ORCA_4_5_0, L=2·10 33 cm -2 s -1 Used event samples

6. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland6 SV reconstruction SV RECO PV MC SV MC  -jet axis  axis  axis The residual of the SV is divided into components along and transverse to the jet axis A double gaussian fit is used to estimate the tails of the residuals

7. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland7 SV residual along  -jet axis  = 5.4 ± 0.5 mm  s = 1.3 ± 0.1 mm ±2  = 62 % H500 2, 3 tracks H200 2, 3 tracks  = 1.6 ± 0.2 mm  s = 0.65 ± 0.03 mm ±2  = 69 %  = 5.9 ± 2.4 mm  s = 1.0 ± 0.1 mm ±2  = 73 % H500 3 tracks H200 3 tracks  = 1.4 ± 0.4 mm  s = 0.60 ± 0.04 mm ±2  = 80 % cm

8. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland8 Why are there tails? cm  R = sqrt(  2 +  2 )  Tails come predominantly from jets with low  R  Hit mismatching because of narrowness of jet? H500 2, 3 tracks H500 2, 3 tracks

9. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland9 Track reconstruction efficiency number of tracks H500H200 number of tracks Initial number of prongs Number of pixel lines Number of reconstructed tracks (at least 8 hits) 0.5

10. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland10 SV residual transverse to  -jet axis  = 50 ± 11 µm  s = 25 ± 3 µm ±2  = 78 % H500 2, 3 tracks H200 2, 3 tracks  = 43 ± 66 µm  s = 22 ± 2 µm ±2  = 87 %  s = 21 ± 1 µm ±2  = 93 % H500 3 tracks H200 3 tracks  s = 20 ± 1 µm ±2  = 93 % cm

11. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland11 Flightpath reconstruction 3D reconstructed flightpath in cmBackground efficiency Signal efficiency Signed flightpath is fine for making cuts D-mesons could be a concern

12. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland12 Charge reconstruction qq Low reliability, because currently tracks are lost H500, 3-prong H500 Reconstructed charge - MC charge Total charge of two  -jets bin 0: 55 % others: 45 %

13. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland13 Effective mass reconstruction m eff = sqrt((  sqrt(p i 2 + m  2 )) 2 -  p i 2 ) QCD background should contain heavier particles m  cut-off GeV mm mm H500, 3-prong QCD, 80-120 GeV

14. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland14 Conclusions SV resolution (without tails) of –600-1000 µm can be reached in jet axis direction for 200 and 500 GeV Higgs events respectively –20-21 µm can be reached transverse to jet axis –resolution dominated by high boost of the  -jets More efficient track reconstruction method has to be used –The signal resolution as well as background rejection will improve if efficiency of reconstructing all 3 prongs is high

15. Helsinki b/tau workshop 30.05.-01.06.2002: MSSM H 0  3 prong using vertexing at CMS - Lauri A. Wendland15 Future plans Move to ORCA_6_x_x To use regional track finding instead of pixel seeds To use deterministic annealing filter (DAF)