Abstract This study addresses a key limitation of seismic noise interferometry for subsurface monitoring: the impact of non‐stationary sources on seismic velocity change estimates. Conventional methods rely on time‐period references, which can inherit source‐related effects when source conditions vary, biasing the estimated velocity changes. This is particularly relevant in urban and industrial environments, where quasi‐monochromatic sources are often present. We show that frequency variations in such signals can produce apparent seismic velocity changes of equivalent magnitude through analytical and numerical analyses. To address this issue, we introduce REFTION, a spatial‐reference approach based on synchronous comparison of correlation functions across seismic stations. We demonstrate its performance in an active open‐pit mine, successfully monitoring a tailings dam under complex noise conditions. By reducing sensitivity to source variability, REFTION enables reliable subsurface monitoring in challenging environments, where anthropogenic noise can be transformed into a valuable monitoring tool.