Orateur
Description
Robust tests of Lorentz invariance violation (LIV) require the combination of multiple datasets to disentangle intrinsic source effects from propagation-induced lags. In this talk, I present results and methods published in my PhD thesis, based on a joint LIV time-of-flight analysis of two BL Lacertae very-high-energy flares observed by imaging atmospheric Cherenkov telescopes: MAGIC (19–20 September 2020) and LST-1 (8–9 August 2021). For the 2020 MAGIC flare, an 88-second-binned light-curve template, modelled as the sum of five asymmetric pulses and a constant background, was used to capture sub-structure relevant to time-of-flight tests. The 2021 LST-1 flare is well described by two Gaussian pulses and a constant baseline, reflecting the observed double-peak structure on 10–20 minute timescales; non-uniform time bins are used in the light curve due to run subdivision. The study is implemented in the open-source maximum-likelihood framework gLike, which incorporates a dedicated unbinned LIV likelihood component and a hierarchical joint-likelihood combination.