Rate-equation models and spectroscopy: are they compatible?

Co-doped laser performance is often modeled by rate equations. For thulium-holmium doped crystals, such equations can yield spurious results compared with experimental measurements. We show that the experimental results seem to invalidate the use of a single type of thulium and holmium pathway to describe the interaction of thulium and holmium. This can have major repercussions on the ability to predict the performance of these lasers when Q-switched or mode-locked.