A Sample of Radiative Tracks at 4.03 GeV

The selection criteria for this sample are based on the assumption that one of the charged tracks has the full beam energy, while the other one shares the beam energy with the photon. This photon should have more energy then any other photons (split-offs, noise) in the event.

The main purpose of this sample is to get a pure source of electrons for PID studies. It could also be used as a source of photons and to investigate calorimeter noise and split-offs.

Cuts:

NCHRG=2
NNEU>0
TRKSUB(1,*)>0 and TRKSUB(3,*)>0 for both tracks
Qtot=0
Closest approach: 0<R<0.01 and |Z|<0.1 for both tracks
larger momentum |P-2.015|<0.15
smaller momentum P<1.25 Higher momenta are of no use for PID studies.
E(photon)>0.3
Shower module(photon)=1 (Photon hits the barrel shower counter)
cos(alpha)>0.998 (alpha is the angle between the predicted photon momentum and the direction of the largest photon)

No cut on the angle between the 2 charged tracks (acollinarity) is made. Only one out of 1237 tracks has MFIT different from 2, so a cut seems not neccesary.

Plots:

The plots were made from the first 856116 events of the '92 Ds data. 1237 events passed all cuts.

The momentum distribution of the tracks with the lower momentum is flat. The need for the cut on the photon energy is evident in the plot of electron momentum vs photon energy.

The angular distribution however, is awkward. Most tracks have cos (theta) > 0.6. The distribution can be appraximated by a polynomial of even powers in cos (theta):

      x=cos(theta)**2
      even=1.7473 + 33.705*x -279.27*x*x +1443.4*x**3 -1503.3*x**4

      if (ranf ( ).lt.10.0/abs(even)) call evlog(i)

The previous line logs out all events with |cos(theta)|<0.55, and some of the others, so that a more flat distribution results. It is shown as the shaded histogram in the plot.