BBC / VPD Vertex Investigation

One critical component of a dijet analysis using the trigger block data is finding a way to determine the Z vertex of an event. Partonic kinematics are accessible at leading order using dijets, but calculating these quantities requires knowing particle eta. The trigger block data gives us the detector eta of jets right away, but we need to know the z vertex of the event to get particle eta. Since the trigger block data doesn't contain tracking information, we have to use detectors like the BBC and VPD to get vertex position by looking at time differences. This page documents my first investigation into using the BBC and VPD for this purpose.

The basic philosophy is to take a number of already produced runs and compare the z vertex found from the TPC and compare it to the BBC or VPD time difference and find a correlation between the time difference and z vertex. For this investigation I used 9 production2009_500Gev_c runs (because these were already produced): 10085131, 10085134, 10085140, 10094095, 10094096, 10094098, 10102031, 10102033, and 10102036. I then run over all primary vertices in these MuDsts and extract the quantities I want. I also require a set of QA conditions on the primary vertices I look at: a 'good' primary vertex must have been reconstructed from at least two tracks and must have too 'good' tracks associated with it, where a good track is one with flag()>0, nHits>2, and nHits/nHitsPossible>0.51. A pdf showing various quantities of interest can be found here. (Note that the VPD plots show an odd spike at a time difference of around 3500, these events come from the runs on day 094).

Figures 1 and 2 show the Z vertex obtained from the TPC vs the time difference for the VPD and BBC respectivily. They also show the linear fits I used to correlate the time difference to a Z vertex.

Fig 1: PPV Z vertex Vs. VPD time difference

Fig 2: PPV Z vertex Vs. BBC time difference

As a final check I used the fitting parameters found above to calculate the Z vertex from the VPD and BBC time differences and then made several comparison plots between the PPV and VPD / BBC Z vertices. (Note that I did not use the runs from day 094 in these plots because of the odd behavior of the VPD). These comparison plots can be found here.