Kopytine's homepage

Correcting for the electron veto inefficiency.

The problem of electrons misidentified as pions is non-negligible in the low field settings only, because the spectrum of electrons falls rapidly with momentum. The electron veto correction factor was determined using supplementary PID provided by the UCAL electromagnetic/hadronic ratio. The question of how tight a cut one needs to apply in order to get the supplementary PID is of little importance, since the electron veto inefficiency determination operates entirely within the sample which satisfies this tough PID criterion. This is true as long as devices used for the veto and for the supplementary PID are uncorrelated. The requirement of $UCHAD/UCEM>10$ identifies track as a reliable hadron and constitutes the tough, clean supplementary PID. Fig. 4.9 shows how the ratio of tracks that exceed the pion $C2$ veto cut to those that do not (let me denote this as $N(\bar{C2})/N(C2)$) changes with $UCHAD/UCEM$. Such figures were obtained for all low field settings and different centralities.

Figure: $N(\bar{C2})/N(C2)$ (see text of subsection 4.4.3) as a function of $UCHAD/UCEM$ for the weak field, high angle, positive polarity setting.

They all have the same characteristic pattern: the ratio falls with $UCHAD/UCEM$, and then becomes flat after $UCHAD/UCEM$ exceeds a certain threshold. From this we conclude that

• a high signal in $C2$, just as a low ratio $UCHAD/UCEM$, both characterize the same group of tracks - obviously, the $e^+(e^-)$ ones
• contribution of the $e^+(e^-)$ goes down to zero when $N(\bar{C2})/N(C2)$ gets flat, and thus one chooses the safe $UCHAD/UCEM$ cut
• for true hadrons, the reasons that cause $C2$ to fire are unrelated to UCAL since $N(\bar{C2})/N(C2)$ is flat above 10.

Thus we have justified the electron veto correction obtained via UCAL. It is the factor

$$1 + N(\bar{C2})/N(C2)$$ (29)

with the counting errors in both $N$ counts propagating into the systematic error of the pion yields.