Table of Contents
Fetching ...

The MeerKAT Fornax Survey VI. The collapse of the galaxy HI Mass Function in Fornax

D. Kleiner, P. Serra, A. Loni, S. H. A. Rajohnson, F. M. Maccagni, W. J. G. de Blok, P. Kamphuis, R. C. Kraan-Korteweg, M. A. W. Verheijen

TL;DR

This work delivers the deepest HI mass function measured outside the Local Group by exploiting the MeerKAT Fornax Survey to map HI in the Fornax cluster down to $\log(M_{HI}/M_{\odot})\approx6$ and detect 35 HI-detected galaxies plus 44 HI clouds. Deep optical data from the Fornax Deep Survey reveal that the HI clouds form a distinct, optically ultra-faint population, with most clouds spatially and kinematically linked to the two HI-rich galaxies NGC 1365 and NGC 1427A. Using a modified maximum likelihood approach and Rauzy completeness tests, the FORNAX HIMF shows an abrupt departure from a Schechter function below $\log(M_{HI}/M_{\odot})=7$, with a low-mass slope of $\alpha=-1.31\pm0.13$ for $\log(M_{HI}/M_{\odot})\ge7$ and a knee at $\log(M_*/M_{\odot})=10.52\pm1.89$ in a volume of 2.7 Mpc$^3$. The collapse at $\log(M_{HI}/M_{\odot})=7$ is interpreted as rapid HI removal in low-mass cluster galaxies, while the high-mass end remains consistent with field HIMFs, highlighting a strong environmental influence on gas content and galaxy evolution in clusters. These results imply that environment-driven gas stripping shapes the low-mass end of the HIMF and motivate similar deep HIMF studies in other clusters with current and future radio facilities such as SKA.

Abstract

We present the deepest HI mass Function (HIMF) ever measured, outside the Local Group. The observations are part of the MeerKAT Fornax Survey and cover a 4 x 4 deg^2 field, corresponding to ~ Rvir. The 3$σ$ detection limit is log(MHI/Msun) = 5.7 for a 50 km/s-wide point source. We detect HI in 35 galaxies and 44 clouds with no optical counterparts. Using deep optical images from the Fornax Deep Survey, we show that the clouds are a distinct population, separated by a four magnitude gap from the faintest HI-detected galaxies. The majority (33 out of 44) of the clouds are associated with the two galaxies with the most HI in the cluster -- NGC 1365 and NGC 1427A, although the clouds contribute a negligible amount to the total MHI budget. By performing a SNR analysis and computing the Rauzy statistic on the HI detections, we demonstrate that our catalogue is complete down log(MHI/Msun) = 6, and we are therefore able to probe the HIMF down to this level. We find an abrupt drop of the number density of HI-detected galaxies at log(MHI/Msun) = 7, signifying a clear absence of galaxies between 6 < log(MHI/Msun) < 7. We use the modified maximum likelihood method to fit a Schechter function down to log(MHI/Msun) > 7, the range where the HIMF follows a power-law. The measured low-mass slope is $α$ = -1.31 $\pm$ 0.13, with a characteristic knee mass of log(M*/Msun) = 10.52 $\pm$ 1.89. The low-mass slope matches the slope in the field, while the knee is defined by a single galaxy and is unconstrained. Below log(MHI/Msun) = 7, there is a sharp departure from a Schechter function, and we report the first robust measurement of the collapse of a HIMF. For the HIMF below log(MHI/Msun) = 7 to follow a power-law, tens of galaxies are needed -- a factor ~ six higher than what is observed. The collapse of the Fornax HIMF is likely due to the rapid removal of HI from low-mass galaxies.

The MeerKAT Fornax Survey VI. The collapse of the galaxy HI Mass Function in Fornax

TL;DR

This work delivers the deepest HI mass function measured outside the Local Group by exploiting the MeerKAT Fornax Survey to map HI in the Fornax cluster down to and detect 35 HI-detected galaxies plus 44 HI clouds. Deep optical data from the Fornax Deep Survey reveal that the HI clouds form a distinct, optically ultra-faint population, with most clouds spatially and kinematically linked to the two HI-rich galaxies NGC 1365 and NGC 1427A. Using a modified maximum likelihood approach and Rauzy completeness tests, the FORNAX HIMF shows an abrupt departure from a Schechter function below , with a low-mass slope of for and a knee at in a volume of 2.7 Mpc. The collapse at is interpreted as rapid HI removal in low-mass cluster galaxies, while the high-mass end remains consistent with field HIMFs, highlighting a strong environmental influence on gas content and galaxy evolution in clusters. These results imply that environment-driven gas stripping shapes the low-mass end of the HIMF and motivate similar deep HIMF studies in other clusters with current and future radio facilities such as SKA.

Abstract

We present the deepest HI mass Function (HIMF) ever measured, outside the Local Group. The observations are part of the MeerKAT Fornax Survey and cover a 4 x 4 deg^2 field, corresponding to ~ Rvir. The 3 detection limit is log(MHI/Msun) = 5.7 for a 50 km/s-wide point source. We detect HI in 35 galaxies and 44 clouds with no optical counterparts. Using deep optical images from the Fornax Deep Survey, we show that the clouds are a distinct population, separated by a four magnitude gap from the faintest HI-detected galaxies. The majority (33 out of 44) of the clouds are associated with the two galaxies with the most HI in the cluster -- NGC 1365 and NGC 1427A, although the clouds contribute a negligible amount to the total MHI budget. By performing a SNR analysis and computing the Rauzy statistic on the HI detections, we demonstrate that our catalogue is complete down log(MHI/Msun) = 6, and we are therefore able to probe the HIMF down to this level. We find an abrupt drop of the number density of HI-detected galaxies at log(MHI/Msun) = 7, signifying a clear absence of galaxies between 6 < log(MHI/Msun) < 7. We use the modified maximum likelihood method to fit a Schechter function down to log(MHI/Msun) > 7, the range where the HIMF follows a power-law. The measured low-mass slope is = -1.31 0.13, with a characteristic knee mass of log(M*/Msun) = 10.52 1.89. The low-mass slope matches the slope in the field, while the knee is defined by a single galaxy and is unconstrained. Below log(MHI/Msun) = 7, there is a sharp departure from a Schechter function, and we report the first robust measurement of the collapse of a HIMF. For the HIMF below log(MHI/Msun) = 7 to follow a power-law, tens of galaxies are needed -- a factor ~ six higher than what is observed. The collapse of the Fornax HIMF is likely due to the rapid removal of HI from low-mass galaxies.

Paper Structure

This paper contains 8 sections, 1 equation, 6 figures, 2 tables.

Figures (6)

  • Figure 1: Cumulative histogram of the RMS values. The black and blue histograms show the cumulative distribution of the median pixel value measured along the line-of-sight in the mosaic, and the RMS at the location of each H i-detection. The grey dotted lines at 0.28 and 0.34 mJy beam$^{-1}$ show the 80$^{\text{th}}$ and 90$^{\text{th}}$ percentiles of the cumulative distribution of H i detections, respectively. The overwhelming majority of H i detections detected near the minimum RMS of the mosaic, with only a small variation.
  • Figure 2: $M_{r}$-$M_\mathrm{HI}$ plane. The main figure shows galaxies detected in H i as filled blue circles, and the H i cloud 5$\sigma$ optical flux upper limit are shown as orange left pointing arrows. The grey line and shaded region is the mean and scatter of the scaling relation of CVn galaxies Kovac2009, the red solid line represented the lowest $M_\mathrm{HI}$ detection limit of the mosaic, and the dashed green line at $M_{r}$ = $-11.7$ denotes the completeness limit of the FDS catalogue. The top panel shows absolute and apparent $r$-band histogram of all cluster galaxies (black), H i-detected galaxies (blue), and the optical flux 5$\sigma$ upper limits of the H i clouds. The H i-detected galaxies and clouds are clearly two distinct populations.
  • Figure 3: Optical, H i emission and velocity field of NGC 1427A (top) and NGC 1365 (bottom). The left shows the FDS $g$-band image, with 3$\sigma$$N_\mathrm{HI,25\,km\,s^{-1}}$ = 2.7 $\times$ 10$^{18}$ cm$^{-2}$ contour overlaid. The H i-detections are coloured according to their integrated $M_\mathrm{HI}$, where galaxies are labelled and are shown with open contours, and H i clouds are filled. On the right are the velocity fields of the respective fields. In all images, the 43$\times$ 40 beam is shown in the bottom left and a 20 kpc scale bar on the bottom right. With respect to NGC 1427A and NGC 1365, the H i clouds have similar velocities, and show a continuous gradient.
  • Figure 4: The $M_\mathrm{HI}$ and statistical SNR. In the main panel, the blue points and orange crosses show the H i-detected galaxies and clouds respectively. Dashed lines have been placed at log($M_\mathrm{HI}$$M_{\odot}$) = 6 and 7, and SNR = 5 and 20. The top panel shows the number of detections for all H i-sources (black) and the galaxies and clouds using the same colours as the main panel. The top panel shows a severe truncation of H i-detected galaxies at log($M_\mathrm{HI}$/$M_{\odot}$) = 7, where the typical statistical SNR is 20.
  • Figure 5: The modified Rauzy test. The Rauzy completeness statistic ($T_{c}$) as a function of integrated flux. The blue points represent $\langle T_{c} \rangle$ from 1000 runs varying the line-of-sight distance of each detection within 3$R_\mathrm{vir}$ of the Fornax cluster, the uncertainties are computed as the variance and the grey lines show each individual test. The red, short-dashed and black, long-dashed lines represent the 97.7% ($T_{c}$ = –2) and 99.4% ($T_{c}$ = –3) confidence intervals.
  • ...and 1 more figures