Hamilton's Object Revisited: A challenging source redshift of a strong lensing configuration

Abstract

Low-resolution spectrographs used to have difficulties in determining redshifts of galaxies at $z\approx1$ and $z\approx3$. Spectral emission and absorption lines of magnesium and iron redshifted to $z\approx1$ fall close to hydrogen, silicon, and oxygen lines at $z\approx3$. Here, we demonstrate that even with modern, integrated field unit spectrographs, this task remains challenging. Hamilton's Object, a blue star-forming galaxy, gravitationally lensed into three multiple images by the galaxy cluster SDSS J223010.47-081017.8, is such a case. Using the Blue Keck Cosmic Web Imager (KCWI) its redshift was determined as $z=0.82$, while its MOIRCS spectrum hinted at $z=3.201$. To resolve the ambiguity, we completely reanalyse the Blue KCWI spectra of all three multiple images including the star-forming region in the outskirts. We employ a new data reduction pipeline, PypeIt, signal enhancement, and line fitting by Python-routines. The reevaluation confirms the previous result based on six absorption features, $z=0.820 \pm 0.001$, and four emission features, $z=0.821 \pm 0.002$. The alternative $z=3.199\pm 0.003$, based on six absorption and two emission lines, is a worse fit, also compared to other spectra. Moreover, we find the MOIRCS spectrum inconclusive: observations cover two of three multiple images, with the slit for image C only covering its central bulge; furthermore the pixel-to-wavelength calibration requires a nightsky emission-line calibration due to a missing calibration arc lamp. New MOIRCS observations are needed to verify that Hamilton's Object has the smallest separation in angular diameter distance between lensing cluster and source galaxy among the known cluster-scale strong lenses.