Calibration of light-flavour $b$-jet mistagging rates using ATLAS proton-proton collision data at $\sqrt{s}=13$~TeV

The identification of jets containing $b$-hadrons is important for the physics programme of the ATLAS experiment at the Large Hadron Collider. Two evaluations of the misidentification rate of light-flavour jets for the $b$-tagging algorithm MV2c10, used in the LHC Run 2 ATLAS analyses, are described. The evaluations are performed in various ranges of jet transverse momenta and pseudorapidities with proton-proton collision data collected at a centre-of-mass energy of ${\sqrt{s} = 13}$~TeV during the years 2015 and 2016. The first evaluation is based on a data sample enriched in light-flavour jets thanks to the application of a dedicated algorithm with much reduced capabilities in tagging $b$-jets and similar performance in mistagging light-flavour jets when compared to the standard $b$-tagging algorithm. The second evaluation is based on a bottom-up approach where the underlying tracking variables in the simulation are adjusted to match performance observed in data; the $b$-tagging algorithm is then re-evaluated to assess the change in the light-flavour jet mistagging probability. The results of both methods are in good agreement and are used to calibrate the mistagging rate predicted by the nominal ATLAS simulation. Calibration factors are in the range of about 1.5 to 3 with uncertainties up to 50%.