JER uncertainties

Check with JetMET how to split systematic uncertainties into various sources (ask Andrea Malara).

assigned to @pkosmogl

Recommendations from JetMET --> This would be a minimal scenario. A full breakdown would be preferrable.

added On hold label

changed milestone to %Understanding JERC (UL FR2)

added to epic &2

changed epic to &8

From JetMET TWiki:

Our recommendation is to treat JER uncertainties as uncorrelated across all years in Run2.

NEW For analysis that are sensitive to JER, i.e. analyses that are able to constrain the single JER nuisance parameter per year w.r.t. their assigned uncertainty (e.g. using control regions), we provide a preliminary recipe to split uncertainties. A full breakup into sources was studied on 2018 data and is available for testing on request. For full Run2 analyses, we provide a simple breakup into 6 nuisance parameters per year (correlated among all jets in all events per year, but uncorrelated across years):

• in region |eta| < 1.93 use one pT bin: (0,Inf)
• in region 1.93 < |eta| < 2.5 use one pT bin: (0,Inf)
• in region 2.5 < |eta| < 3 use two pT bins: (0,50) and (50,Inf)
• in region 3 < |eta| < 5 use two pT bins: (0,50) and (50,Inf)

mentioned in merge request !174 (merged)

added changepatch typephysics labels