Kopytine's homepage

Sources of background

In this section the words ``backgrounds'', ``background effects'' should be understood in the technical and literal sense, as the causes responsible for the electrical signals produced by the experimental apparatus and recorded, but unrelated to the physics of the $PbPb$ collisions. In the course of data taking for this analysis, the following effects gave rise to the detector occupancy even in the absence of the $PbPb$ interactions:

1. nucleus-nucleus interactions between the incident $Pb$ and the material of the 1 cm thick styrofoam target holder (the ``empty target'')
2. imperfections of the detector: due to the non-zero pedestal width (seen on Fig. 6.1), the probability of detecting a ``hit'' even in the absence of a particle track is finite for any reasonable hit threshold. What is worse (and that has been seen in our detector) is the fact that the fake hits tend to arrange themselves in regular patterns. After an event-display study on a statistics of the order of $10^4$, it was concluded that the pathological patterns can be roughly categorized in two groups:
   1. Sometimes, several fake hits appear simultaneously in channels served by the same AMPLEX chip. Their amplitude can reach values typical of and exceeding those of the typical tracks. This problem is addressed by the sector-wise event mixing, as discussed in Subsection 6.6.2.
   2. A pattern of ``reversed $\,dN/\,d\eta$ shape'' was seen: a systematic increase of channel amplitude towards the outer rings (where physics multiplicity is lowest - hence the word ``reversed''). Subevent mixing is of little help here because these are large scale patterns. However, inspection of empty target runs showed that the same events exist there. This, on the one hand, rules out ``physics'' and pile-up explanations; on the other - enables a correction based on the empty target subtraction.
   The pathologies from group 2a affect power spectra components (PSC) of the azimuthal and pseudorapidity modes; those from group 2b - PSCs of the pseudorapidity mode.
3. $\delta$-electrons not deflected by the field and hitting the ``clean'' side of the detector.
4. random tracks whose origin is unrelated to our experiment

Items 1 and 4 are correctable by an empty target subtraction procedure unconditionally; items 2 and 4 can be corrected by event mixing, and by empty target subtraction if one assumes that it reproduces itself in the empty target events of the same total multiplicity as the given one. Item 3 has been shown by MC (described in Section 6.8) to be negligible.