Prefiring for AK8

In https://indico.cern.ch/event/1450930/#2-comparison-of-dijet-mass-in, differences have been observed between EOY16 and UL18 only for AK8 in the forward region. One possible explanation (although one should keep in mind that it is not the only one, e.g. dead regions in the detector only on one side) is that the prefiring corrections are provided only in bins of AK4 jet

p_\mathrm{t}

.

We need to investigate:

• impact on other observable, e.g. inclusive jet (closer to what happens in the detector)
• clarify the level of JEC correlations among the years and the time stability of the JECs (this changes the level of agreement)
• prefiring corrections as a function AK8 jet
  p_\mathrm{t}
  • ideally, one should rederive the prefiring correction as a function of AK8 jet
    p_\mathrm{t}
    ;
  • in practice, it might be sufficient to derive a mapping from AK4 jet
    p_\mathrm{t}
    to AK8 jet
    p_\mathrm{t}
    .

added to epic &2

changed the description

About this discrepancy, @cversteg pointed out that another source of difference could be the VFP effect, which was not simulated in the pre-UL 2016 samples.

I think that we will be able to check by repeating @asingla's studies with UL samples.

added changepatch statusto do typephysics labels