Monitored variables in biomass combustion devices, such as emissions or temperature, are usually characterized by their averaged values over the entire process, although they tend to present a pulsating behavior. This assumption induces instantaneous errors when experimental data are used in numerical simulation. The interest in obtaining high-quality experimental information for a proper model validation leads to find an alternative to describe them and to discern which operating variables are the main ones responsible for those fluctuations. The application of frequency-domain tools, such as coherence, is proposed as a useful solution for determining if the emission and temperature profiles are intrinsically related to any known periodical signal, such as the feeding cycle.