Data driven approach to background subtraction for jet substructure measurements

In high-energy hadron collisions, accurate signal reconstruction requires effective background subtraction to mitigate underlying event and pileup effects. The p_T of the detector level leading jet is estimated by subtracting the average event density (ρ), but the measured substructure observables are still contaminated by background particles. The perpendicular cone method estimates this background contribution by using energy deposits in azimuthally displaced regions to the leading jet, assuming that this region is unlikely to contain particles from the hard scattering processes.

Recent studies reveal residual biases in the background subtracted signal when applied to jet substructure measurements in lower p_T regimes. We introduce a simple multiplicative correction factor that restores closure and robustly improves background estimation across diverse kinematic ranges. We demonstrate its application in multiple jet subtructure analyses, including recent jet hadrochemistry and energy-energy correlator measurements.