E781 Results of Sigma0 Reconstruction

 1.0 The first obvious test of lambda reconstruction in V0 package
     is the reconstruction of neutral Sigma Hyperon decay. Indeed, 
     we do a test for V0 and Photon packages.
     First for Selex, the Sigma0 decay in Vertex spectrometer was
     analyzed by Victor Kurshetsov (see his home page).
     He suggested sigma0 to lambda rate equal to 1:2 and calculated
     single photon efficiency to be about 40%.
     I did Sigma0 reconstruction in order to compare different types
     of lambdas we do have in V0 package and to obtain the acceptance
     to reconstruct Sigma0 decay.
      
     We have 3 types of lambdas:
     - lambda in Vertex;
     - lambda in M1 spectrometer (2p reconstruction);
     - lambda in M1 spectrometer (1p reconstruction);
      
     We have 2 types of photon hits:
     - all photon hits   (any hit in adcgam_bk)
     - identified gammas (id_adcgam==1)

     The second sample gives us more than less natural gammas which
     do have expected energy distribution in led glass cluster, do
     not have second maximum or charged particle entry in corresponding
     cluster etc. But we have lost something  (e.g. due to overlapping
     photon and charged particle clusters).
     I do not intend to suppress any background here using pi0
     reconstruction, sigma0 kinematics etc, but this should be done
     for real physics.
     I demand a save-full energy cut: E gamma > 2 GeV.

      All the PAW fits to Sigma0 mass were done by Grisha Davidenko.      
      Thank you, Grisha.
      
 1.1 Lambda in Vertex.
     The mass distributions for all photon hits and identified gammas
     combinations with lambda in vertex selected in soft cut mode and
     in mass interval +-5 MeV/cc  (3000 lambdas) are shown in   fig.1a 
     and in   fig.1b 
     respectively.

                   one million filtered interactions
                   ---------------------------------
                   Sigma0 events    mass mean value   GeV/cc     rms
  all photon  hits   443 +- 92       1.193 +- 0.0016           0.0081
  identified gammas  324 +- 73       1.192 +- 0.0016           0.0075
  
     We see that we have right peak at right place, we lost 25% of
     sigma(lambda,gamma) decays due to cut on gamma ID. The number
     443 is in an agreement with sigma/lambda = 1:2 rate and single
     photon efficiency of 44%.
     
 1.2 Lambda in M1 spectrometer, 2p topology.
     First, I took the lambdas selected with default V0 setup in mass
     range +- 25 MeV/cc and any photon hits. The mass distribution 
     of Lambda+gamma combinations is shown in   fig.2 .
     Next, I did Lambda+gamma with identified gammas and lambdas
     selected in soft cut mode (see   fig.3 ).

     One can see a nice peaks at sigma0 mass.
     
                   Sigma0 events    mass mean value   GeV/cc    rms
  default setup  2439 +- 187      1.191 +- 0.0009           0.0114
  soft cut mode  1346 +- 112      1.192 +- 0.0010           0.0105

     Applying the cuts we win about 10% in the mass resolution and
     loose 45% in statistics. The default V0 setup seems to be good
     to extract sigma0.  

 1.3 Lambda in M1 spectrometer, 1p topology.
     I used soft cut mode, identified gammas and in addition, 
     uncoleaniarity cut of 2 mr (maximal angle between momentum vector
     and geometrical vector of primary vertex and lambda decay vertex).
     This works against helps cascade lambdas and artificial vertices.
   
     The mass distributions for 1p topology is presented in   fig.4a 
     and that for 2p topology in  fig.4b .
     
     
   (unco cut = 2)   Sigma0 events    mass mean value   GeV/cc    rms
   1p topology       376 +-  67.3    1.193 +- 0.0020           0.0101
   1p topology      1168 +- 106.2    1.192 +- 0.0010           0.0102

     We may expect 442 sigma0 in vertex, 2439 in M1 with 2p topology and
     376 in M1 with unmeasured pion momentum.

     all together: 

       0.33% sigmas per 1 filtered interaction or 18% per lambda
       =========================================================
       (the multiple lambda entries might be in the sigma0 peak)
       

vmatveev@fnal.gov