The biases of Spitzer-selected clusters and the discovery of almost 200 z > 1.3 (proto)clusters in the LSST deep drilling fields.

The most efficient means of detecting high redshift (proto)clusters is by identifying galaxy overdensities in a Spitzer/IRAC colour-selected sample that effectively selects z > 1.3 galaxies, regardless of galaxy age or type. Its simplicity and efficacy have resulted in it being one of the most widely used (proto)cluster detection methods to date (e.g., ISCS, IDCS, CARLA, Papovich 2008), with the majority of known high redshift clusters discovered using it. While this method has been used extensively in the literature, its biases are poorly understood, bringing into question how representative this sample of (proto)clusters is compared to the wider population. We therefore study its biases through comprehensive testing on a mock catalogue in the form of a lightcone. To examine the Spitzer selection method, we matched the lightcone to the IRAC data to a depth of 22.75 mag, and optimised the method to give us the purest sample. We find that the Spitzer-selected (proto)cluster sample is highly incomplete (~4%) and heavily biased towards larger, richer, more massive, and more centrally concentrated (proto)clusters. This bias may explain the disparity in quenched fractions and galaxy scaling relations found in (proto)clusters selected by different methods. Finally, we applied the optimised Spitzer-selection method on a set of LSST's deep drilling fields, covering ~30 deg^2, locating 189 candidate (proto)clusters. We estimate that 60-80% of the identified candidates are likely to be genuine (proto)clusters, which is corroborated by a ~4sigma stacked X-ray signal from these structures. With knowledge of their selection biases and their homogeneity, in addition to the LSST deep drilling fields soon to become the most studied fields in astronomy, the protoclusters in our sample are the ideal laboratories for future cluster formation and galaxy evolution studies.