Reinvestigation of the Multiepoch Direct Detections of HD 88133 b and Upsilon Andromedae b

Abstract

We reanalyze the multiepoch direct detections of HD 88133 b and ups And b that were published in Piskorz et al. (2016) and Piskorz et al. (2017), respectively. Using simulations to attempt to reproduce the detections, we find that with the six and seven L-band Keck/NIRSPEC epochs analyzed in the original works, the planets would not have been detectable unless they had unreasonably large radii. HD88133 and ups And both have fairly large stellar radii, which contributed to the difficulty in detecting the planets. We take this opportunity to consider how these planets may have been detectable with the small number of epochs originally presented by running simulations both with the upgraded NIRSPEC instrument and with near-zero primary velocities, as recommended by Buzard et al. (2021). While seven L-band NIRSPEC2.0 epochs with near-zero primary velocities would have allowed a strong (10.8σ) detection of ups And b, many more than six L-band epochs would have been required for a strong detection of HD88133b, which could be due in part to both this system's large stellar radius and low stellar temperature. This work stresses the importance of careful analytic procedures and the usefulness of simulations in understanding the expected sensitivity of high-resolution spectroscopic data.