On the existence of bright IR galaxies at z > 2: Tension between Herschel and SCUBA-2 results?

Recent derivations of the galaxy star formation rate density (SFRD) obtained from submillimetre (sub-mm) surveys (e.g. SCUBA-2) show a tension with previous works based on Herschel and multi-wavelength data. Some of these works claim that the SFRD derived by pushing the Herschel surveys beyond z ≃ 2 are incorrect. However, the current sub-mm surveys obtained from SCUBA-2 data and the methods used to construct the total infrared (IR) luminosity function (LF) and the SFRD could be affected by some limitations. Here, we show how these limitations (i.e. selection bias and incompleteness effects) might affect the total IR LF, making the resulting dusty galaxy evolution of difficult interpretation. In particular, we find that the assumed spectral energy distribution (SED) plays a crucial role in the total IR LF derivation; moreover, we confirm that the long-wavelength (e.g. 850-μm) surveys can be incomplete against 'warm' SED galaxies, and that the use of a wide spectral coverage of IR wavelengths is crucial to limit the uncertainties and biases.