Home » Groups » STAR Protected

Run 9 200GeV Dijet Cross Section Time Variation Investigation

Updated on Tue, 2013-09-24 11:50. Originally created by pagebs on 2013-09-06 15:35.

Here I look into the dependance of the dijet cross section on the time into the run ...

Figure 1: In this figure, the dijet yield is broken into 10 day periods and normalized by the integrated luminosity for that period. The upper left plot shows the spectra and the upper right plot shows the ratio (10 Day Period / Full Run) for each 10 day period. The bottom panels show the same things except the runs have been grouped differently.

Figure 2: This figure shows the ratio (10 day period / Full run) but now the samples are normalized in different ways. The upper left panel is normalized using the integrated luminosity (same as in figure 1), the upper right panel is normalized using the number of BBCMB-Cat2 triggers as reported in the run log, the bottom left panel is normalized using the number of ZDCMB triggers as reported in the run log, and the bottom right panel is normalized using the number of VPDMB triggers as reported in the run log.

Figure 3: This figure shows the averages of several quantities as a function of run index for all events/jets which are used in the dijet cross section.

Figure 4: This figure shows the ratios of various scalar quantities as a function of run. Note: due to the way my code is set up, it was much easier to make these plots with index running from 1-773.

 
Table 1: This table shows the average value for each ratio in each 10 day period
  L2/BBC L2/ZDC L2/VPD BBC/ZDC BBC/VPD ZDC/VPD Dijets/BBC
12 14.11 22.107 0.0221 1.567 0.00156 0.000998 1.1899
13 13.798 22.0459 0.0219 1.598 0.00159 0.000993 1.2247
14 12.748 21.341 0.0212 1.674 0.00166 0.000994 1.1252
15 12.360 21.0260 0.0214 1.701 0.00173 0.00102 1.0339
16 11.649 19.669 0.0200 1.688 0.00171 0.00102 1.0028
17 11.529 18.766 0.0206 1.628 0.00178 0.00110 0.9673
18 11.686 19.407 0.0207 1.661 0.00177 0.00106 0.9645
 
Figure 5: This figure shows the fully unfolded dijet cross section for the RFF and FF run periods. The top panel shows the spectra and the bottom panel shows the ratio (Field-Full)/Full where 'Field' is either RFF or FF.
 
Figure 6: This figure shows the (Data-Theory)/Theory ratio for the RFF and FF data sets where the theory has been corrected for UEH. The correction is done seperately for each field polarity.
 
Figure 7: This figure shows the average quantities but zoomed in and fit with 3 p0's corresponding to the RFF, FF before spin rotator change, and FF after spin rotator change.
 
Table 2: This table shows the p0, error, and Chi2/ndf for the three fits to each of the profile plots.
  RFF Fit: p0 Val; p0 Err; Chi2/NDF FF Before Rot: p0 Val; p0 Err; Chi2/NDF FF After Rot: p0 Val; p0 Err; Chi2/NDF
Dijet Mass 24.2808; 0.0049; 741/363 24.4848; 0.0047; 563/332 24.4533; 0.0099; 118/73
Jet Pt 11.6379; 0.0020; 1053/363 11.7361; 0.0019; 782/332 11.7207; 0.0041; 163/73
# Tracks 4.1487; 0.0010; 682/363 4.1346; 0.0010; 725/332 4.1057; 0.0021; 100/73
# Towers 9.1582; 0.0017; 802/363 9.1550; 0.0017; 599/332 9.1663; 0.0035; 133/73
Sum Track Pt 6.3893; 0.0019; 987/363 6.4271; 0.0018; 852/332 6.3770; 0.0039; 274/73
Sum Tower Pt 5.4073; 0.0015; 467/363 5.4666; 0.0015; 431/332 5.5003; 0.0031; 90/73
Track Pt 1.5399; 0.0004; 2306/363 1.5544; 0.0004; 1206/332 1.5533; 0.0008; 333/73
Tower E 0.6484; 0.0002; 737/363 0.6566; 0.0002; 519/332 0.6601; 0.0003; 114/73
Tower ADC 79.7662; 0.0102; 1105/363 79.5555; 0.0100; 594/332 79.6377; 0.0208; 116/73
 
Figure 8: This figure shows the ADC spectra for all towers in the dijets. The zoomed out version can be seen here.

Figure 9: This figure compares the full cross section and the cross section for various parts of the run. In all cases, the cross section is unfolded using the same embedding runs as the data runs used for a given part of the run. The first two panels compare the RFF and FF cross sections to the cross section from the full run. In the first panel, the cross sections are normalized using the luminosity calculated using the BBC and in the second panel, the cross sections are normalized using VPDMB counts. The second two panels compare the cross sections split into 4 periods again using the BBC and VPDMB for normalization



Figure 10: This figure shows the average number of tracks in the jets of a dijet for the data (Top) and the embedding sample (Bottom) with the same three fits used in figure 7. Note that the spread in points is much greater in simulation than in the data. Also the errors in the simulation are much smaller (I don't think I took into account the weights properly in the error). Similar plots for all other quantities shown in figure 7 can be found in this PDF.

Although the spread in points in the simulation is large, there do not seem to be any systematic shifts in the averages over the course of the run as there appears to be in the data.

In a message to list, Jamie D. suggested that I look at dead areas in the TPC to see if they correlate with shifts in the various detector quantities. These dead areas should appear in certain sectors only as specific RDO boards die, so I have broken up the average number of tracks and average track pt plots by sector.

According to Gene's blog, there were two sectors which had RDO boards die during the period of my analysis: sector 2 and sector 18 (there were also a couple of boards which died on day 180). The board in sector 2 died durring run 10162027 which corresponds roughly to index 721. Looking at the average number of tracks plots, there is no obvious feature in sector 2 at index 721, but there is a broad dip between index 500 and 700 which I don't have an explanation for. Sector 18 had two boards die, one on run 10158014 (index 650) and one on run 10174044 (index 912). There is a definite decrease around index 912 and a decrease a short while after index 650.

Figure 11: This figure shows the systematic associated with the time variation of the cross section. The size of the systematic was taken as the magnitude of the difference between the RFF and FF cross sections as calculated and shown in figure 9. The first panel shows the sys/cross sec ratio, the middle panel shows the old cross sec ratio, and the bottom panel shows the cross sec ratio with the new systematic.



Groups: