Improvement in the ECAL Energy Reconstruction

The signal saturation in the electronics is insignificant over the energy range of interest, the remaining saturation is in the calorimeter fibers. It is related to conversion of ionization to light. As illustrated in Figure 1 the effect is maximal near the shower peak, whereas for the rest of the shower, the cells are not affected. Using non-saturated cells allows an accurate accounting for the amount of saturation.

Example of a 1.6 TeV electron shower in the ECAL.
Figure 1. Example of a 1.6 TeV electron shower in the ECAL. Most of the shower energy is deposited along the shower axis near the shower maximum. For high energy showers (>500 GeV), some of the amplitudes are saturated.

This effect was studied in detail and calibrated using electron showers collected on orbit. First, it was verified that this effect is not related to the electronics — the amount of saturation for large unattenuated signals near the PMT is the same as for attenuated signals far away from the PMT.

Then the saturation effect is calibrated using the energy dependence of the maximum cell amplitude in the shower. This is illustrated in Figure 2 using a sample of high-energy electrons. There is good agreement between the data and the Monte Carlo simulation if saturation effects are taken into account. On the contrary, if saturation effects are not included in the Monte Carlo simulation, its prediction is significantly above the observation.

The maximal cell amplitude in a shower as a function of the reconstructed shower energy ?0: the data (full red circles) from the AMS sample of cosmic ray electrons and the Monte Carlo simulation with (open green squares) and without (open blue circles) the saturation effects in fibers.
Figure 2. The maximal cell amplitude in a shower as a function of the reconstructed shower energy $E_0$: the data (full red circles) from the AMS sample of cosmic ray electrons and the Monte Carlo simulation with (open green squares) and without (open blue circles) the saturation effects in fibers. The Monte Carlo simulation that does not take into account saturation effects shows a significant discrepancy with the data.

Correction for the saturation effects (see Figure 2) is verified by excluding the amplitudes above a threshold from the fit to the cell amplitudes. Figure 3 shows the effect on the reconstructed energy as a function of the threshold value. As seen, the reconstructed shower energy does not depend on the threshold value.

Shower energy reconstructed excluding amplitudes above a threshold energy per cell.
Figure 3. Shower energy reconstructed excluding amplitudes above a threshold energy per cell ($E_{\rm thres}$). As seen the reconstructed energy does not depend on the threshold value. This is an important verification that this method correctly reconstructs the shower energy.

The new technique provides AMS with a precision energy measurement of electrons and positrons up to multi-TeV.