Image Details
Caption: Figure 3.
Left panels: photometric redshift results from simulating a 10 filter MINERVA survey from UNCOVER/MegaScience and CANUCS/Technicolor to PRIMER/MINERVA depth and including the PRIMER filters (18 JWST filters + 3 HST filters). This is plotted on the y-axis, and the ground truth from UNCOVER/MegaScience and CANUCS/Technicolor is plotted on the x-axis. In this case, MINERVA’s zphot are estimated to have had a ∼1.26× improvement in σNMAD and 1.18× reduction in outlier fraction. Right panels: the same simulation but for a MINERVA survey that would include five filters in the SW (four medium bands + F070W) and all eight LW medium bands at simulated MINERVA depth, including PRIMER broad bands (21 JWST filters + 3 HST filters). With all medium bands included, MINERVA’s zphot would have had a 1.86× improvement in σNMAD and 1.30× reduction in outlier fraction. Given that the required observing time increase for each dataset would be a factor of 1.25 and 1.75, respectively, this shows that the improvement in zphot scales approximately linearly with exposure time.
© 2026. The Author(s). Published by the American Astronomical Society.