Hadronic Component

The component of the background rates coming from untagged hadronic interactions of the beam in the target has been ignored in this discussion. I will argue here that for the purposes of calculating trigger and event rates it can be ignored. The hadronic interactions of a bremsstrahlung beam with a beryllium target must be analyzed over the full range of photon energies. By far the most important hadronic interaction from the point of view of sheer rate is nuclear photodisintegration. This cross section reaches a peak on the order of mb in the region of the giant resonances (10-30MeV) in most stable nuclei. Beryllium has a photodisintegration peak at very low energy (1MeV) [3] that has important implications for nuclear reactions inside stars. However for our purposes these interactions are not important; in spite of their copious production inside our target, the protons and alpha particles from photodisintegration are stopped before they escape the target. The neutrons that are released can escape the target, but their interaction probability in the thin trigger scintillators is small.

The most important hadronic interactions of the beam that contribute to background rates in the detectors is nucleon resonance photoproduction followed by pion emission. The $\Delta(1230)$ is the most important, amounting to approximately 70KHz at a tagged rate of $5\cdot 10^7 \gamma$/s. The higher lying $N^*$ resonances contribute about 60KHz in addition. This covers the photon energy range up to about 2GeV. Above 2GeV the cross section becomes approximately constant at about 120$\mu$b. Here the interactions are diffractive in character and contribute about 30KHz to the total rate in the detector, 6KHz of which is tagged. These numbers include only reactions with a recoil proton; to obtain an inclusive rate that covers reactions with a recoil neutron as well, they should be multiplied by 9/4.

Comparison with the singles rates in Table 3 shows that the hadronic component is a small fraction of the total, of order 10% at best. Most of the hadronic tracks in an inclusive sample come from charged pions, and these are suppressed at the same time as the electromagnetic background by raising the threshold on the recoil proton trigger scintillators. With a $dE/dx$ suppression factor of 5 and a tagger accidental probability of 40%, this brings the total hadronic rate to about 30KHz. Delta resonance decays favour neutral pion over charged pion production 2:1, so averaged over the entire spectrum this gives a further reduction in the inclusive hadronic trigger rate of a factor of about two. This leads to 15KHz of level 1 triggers from hadronic events, to be compared with 150KHz in Table 3 for electromagnetic sources. Only a fraction of this hadronic rate is within the acceptance of our recoil proton trigger.

Beyond level 1, most of these events are eliminated by their low energy deposited in the LGD. What remains at this point is essentially just the diffractive neutral events from the tagged end of the photon spectrum. The rates calculation presented in this report included 3KHz of tagged hadronic events in the trigger at level 1, two thirds of which contain charged tracks in the forward region. These events are kept at level 2 but get reduced at level 3 by about a factor of 10 by the requirement total energy requirement in the forward calorimeter. All of these ratios were extracted from the data taken during 1998 running. That one sees 3KHz at level 1 instead of 6KHz follows from the fact that the RPD trigger saw only a fraction of the total recoil protons. With the BSD one might expect this number to increase closer to the full acceptance limit of 6KHz. This would have a negligible impact on the rate of events written to tape because the events passing the level 3 trigger are 90% accidentals at full beam intensity. I conclude that the uncertainty in these predictions from hadronic backgrounds that have not been taken into account amounts to about 5% which justifies having disregarded them in this study.