Here I show the effect of reweighting the simulation so the z-vertex distribution matches that from data ...

Figrue 1: This figure shows the z-vertex spectra from data and simulation for L2JetHigh and JP1 triggers as well as the data/simu ratio.

Figure 2: This figure shows the z-vertex spectra from data and the simulation which has been reweighted using the parameters from figure 1.

The reweighting is done by taking the data/simu ratio and fitting it with a 4th order polynomial. The simulation partonic weight factor is multiplied by the value of this fit at the z-vertex of the simulated event so that the simulation vertex distribution matches the data vertex distribution. Note: the reweighting is only done for the matched simulation (at det, particle, and parton levels), not the unbiased simulation from the pythia.root files.

Figure 3: This figure shows how the reweighting changes the L2JetHigh response matrix. The ratio (reweighted-original)/original is shown for each bin.

Figure 4: This figure shows how the z-vertex reweighting changes the jet eta distribution in the (detector level) simulation.

Figure 5: This figure shows the jet eta data / simu comparison using the simulation which has not been reweighted (top 4 panels) and the simulation which has had the vertex reweighting applied (bottom 4 panels). The agreement between data and simulation is better for the vertex reweighted simulation, most noticeably at the ends of the spectra.

Figure 6: This figure shows how the z-vertex reweighting changes the dijet invariant mass spectrum in the (detector level) simulation.

Figure 7: This figure shows the ratio of the L2 cross section calculated using the reweighted simulation to the L2 cross section calculated using the original unreweighted simulation. Note that the vertex reweighting is only applied to the simulation which makes up the response matrix.

Plots similar to figure 7 have also been made for the cross sections calculated with the various efficiency and track/tower shifts applied. They can be seen in the following links. 4pct track inefficiency, 7pct track inefficiency, -1pct track pt shift, +1pct track pt shift, -3.7pct tower energy shift, +3.7pct tower energy shift.

Figure 8: This figure shows the L2 dijet (data-theory)/theory curve with systematics on theory and data included using the simulation which had the z-vertex reweighting applied. This plot can be compared to figure 8 on this page which is the same except it was generated using the simultion which did not have the z-vertex reweighting applied.

As expected, the vertex reweighting only changes the final cross section slightly. The most dramatic change can be seen in the jet eta distribution where reweighting improves the data simu agreement at the edges of the spectrum.