For this differential AJ measurement we use Run 7, 0-20% L2gamma data, compared to Run 6 p+p HT data.  p+p tracking efficiency is smeared to match Au+Au efficiency. Our hard-core di-jets (explained below) are selected using p_T^{lead jet} > 16 GeV/c, and p_T^{sublead jet} > 8 GeV/c, and are required to be back-to-back.

Figure 1:

Hard-core (clustering done with constituents above p_T^{hard const}) A_J for a grid of di-jet definitions, scanning both the jet radius and the p_T^{hard const}. We see that in no cases the Au+Au is balanced to the level of p+p. This implies there is some modification of the hard part of the jet for all of these di-jet definitions.

Figure 2:

Matched (jets clustered with all constituents with p_T > 0.2 GeV/c, and geometrically matched in dR to the hard-core jet) A_J for varying di-jet definitions (same definitions as in hard-core plot above). We see a relatively smooth evolution - at small p_T^{hard const} and small jet radius, Au+Au and p+p differ significantly, and as both variables are increased to larger values, the curve approach each other. At large jet radius and large p_T^{hard const}, the Au+Au is balanced to the level of the p+p reference.

Figure 3:

Here we estimate our sensitivity to correlated jet yield in the "soft" part of the jet (constituents with  0.2 GeV < p_T < p_T^{hard const}). We embed hard-core Au+Au di-jets into 0-20% central min-bias Au+Au events, and calculate the A_J for the corresponding matched jets (blue curve above). If all three curves sit on top of each other, then we have no sensitivity to the correlated jet yield, and our signal is dominated by the smearing from fluctuations in the energy density of the underlying event. However, we see that in all cases, both the Au+Au and the p+p are significantly more balanced than the embedded hard-core jets, implying we are sensitive to the physics we are interested in.

Figure 4:

Here, we do the same analysis as was done for the matched A_J. The difference is that now the matched jet radius is being changed, not the hard-core radius. We always start with a hard-core jet radius of 0.2, and scan out from 0.2 -> 0.4 in the matched jet. Essentially, this is scanning the radial energy distribution of the jet, with the goal of measuring at what radius do we start to recover a similar amount of energy in the Au+Au as we do in the p+p. Interestingly, we see here that unlike in the original differential matched A_J, once you've selected a narrow (R=0.2) hard-core jet, it does not matter what your p_T^{hard const} is, you always recover balance around R=0.35.