Optimizing spectral stacking for 21-cm observations of galaxies: accuracy assessment and symmetrized stacking

We present an assessment of the accuracy common operations performed in $21$-cm spectral line stacking experiments. To this end, we generate mock interferometric data surveying 21-cm emission at frequency $1310<\nu<1420$ MHz ($0.005<z<0.084$) and covering area $\sim 6$ deg$^2$ sky, mimicking observational characteristics real MeerKAT observations. find that primary beam correction accounts for just few per cent ($\sim8\%$ 0 power, 3\%$ 0.6 power) deviations from true $M_{\rm HI}$ signal, weighting schemes based on noise properties provide unbiased results. On contrary, distance can account significant systematic mass differences when applied to a flux-limited sample ($\Delta M_{\rm HI}\sim 40-50\%$ studied case). no difference final $\braket{M_{\rm HI}}$ obtained spectroscopic redshift uncertainties are accounted procedure ($ \Delta z\sim 0.00035$, i.e. $\Delta v \sim 100\,{\rm km\, s}^{-1}$). also novel technique increase effective size galaxy by exploiting geometric symmetries cubelets, potentially enhancing SNR factor $\sim\sqrt{2}$ analyzing stacked spectrum (a 4 cubelet). This is found be robustly unbiased, while efficiently increasing SNR, as expected. argue appropriate framework employing detailed realistic simulations required exploit upcoming datasets SKA pathfinders accurate reliable manner.