Prompt Photon Signal in Pythia

This page was made purely for the purpose of discussion during the FMS meeting!

Using the result from full Pythia 6.222 + GSTAR simulation, we look at the prompt photon signal.

The included processes are all normal QCD involving light quarks, all prompt photon productions, and doubly diffractive scattering.

Fig1. Pi0 vs. Single Photon in full simulation

Fig 2. Process contritbutions to the single photon signal

The glaring problem here is what I labeled as "unknown". These events have peculiarities that makes me worry that they may have been corrupted. In the least, these are events that I don't know how to interpret. The features of these events are,

1. As far as I can understand, in the Pythia record, the main hard interaction occurs between the line 4 and 5(initial state after ISR) and line 6 and 7(final state before FSR). Line 6 and 7 tehrefore do not have parent particle defined, and the energy sum of line 4 and 5 is an exact match to the energy sum of line 6 and 7. For these "unknown" events, however, the initial state energy and final state energy differ by a wide margin, and often there is almost no energy in the final state.

2. Most of these "unknown" photons have a parent that is a Pi0. And the parent Pi0 also has a daughter that is the observed photon, so it's self consistent up to the particle association. However, for these events the parent Pi0 never has nearly enough energy to produce the daughter photon. In fact, most cases the Pi0 carries less than a few GeV, even when the daughter photon has ~70GeV of energy. Naturally, if I sum the two Pi0 daughters according to the event record, it gives me invariant mass that is all over the map and as high as 100GeV.

3. The main processes that generate these oddball events are fairly democratic. It's rarely prompt photon production, but it's mostly regular QCD and sometimes diffractive interaction.

4. Finally, when I was looking at the origin of Pi0s in Pythia, there were a group of events in which I couldn't figure out the parentage of the Pi0. Turns out thoese events also has the symptom number 1, that is the initial and final state energy in the hard interaction does not match up.

I am still not sure what to make of these events. For now, I tried omitting these events to see how things change.

Fig 3. Pi0 vs. Single Photon in full simulation, without the "unknown" events

Fig 4. Pi0 vs. Single Photon in Run6T data

Whatever it's worth, without the odd events the single photon slope looks closer to that in the data.

Fig 5. Process contritbutions to the single photon signal, without the "unknown" events

Here we see that indeed the crossover between Pi0 decay photons and Prompt Photon occurs at ~55GeV. (This is with the Center Cut, which functions as kind of a veto for single photon) We also see that the 2 photon merging seems not to be an issue, although it may have been somewhat underestimated if my track-photon association is inefficient. (Should be ok, I associate every track that is in 2-cell radius from the center of a cluster, which for very high every Pi0s should comfortably include both daughters)

Fig 6. MC vs. Data Single Photon to Pi0 Ratio, without the "unknown" events

So I could boldly claim that the data may after all have enough room to have at least as much prompt photons there are in Pythia, with lots and lots of assumption.