Real-time dual-telescope normalization for improved density estimation in absorption muography

Absorption muography infers path-averaged density from the attenuation of atmospheric muons. Transient flux variations driven by weather, solar activity, and geomagnetism bias sequential measurements that rely on a non-simultaneous background. Here we show a real-time dual-telescope normalization in which two identical plastic-scintillator telescopes operate in parallel, one viewing open sky as a reference and one viewing the target. Their count-rate ratio cancels common-mode flux changes and tracks detector drift. A 140-h no-target run yields a stable calibration, and stacked acrylic slabs give an empirical transmission–thickness curve. Applied to blind targets, the dual-telescope workflow recovers the densities of 20-cm aluminum and 10-cm stainless-steel blocks within uncertainties, whereas a single-telescope workflow shows systematic biases of +18.95% and −22.82%. A vertical stack of the two telescopes further improves common-mode rejection, reducing calibration ratio scatter from 7.6% to 3.3%. The method is model-independent, removes systematic density biases in any campaign where a simultaneous flux reference is feasible, and is directly applicable to field deployments.