1 of 38

2 of 38

HI in ISM & CGM/IGM for galaxy studies

MW-like

LMC-like

The x=y line

Important and major component of galaxies

Guo+23

Zwaan+05

The high-column (N_HI>10²⁰ cm^-2) ISM contains most of the HI

3 of 38

The important N_HI range of 10¹⁷ ~10²⁰cm^-2

(Peroux+20)

ISM within disk

ISM-CGM or disk-halo interface or cosmic web within halo

van de Voort+12

Mostly in halo and carries most of the inflow

Historically it was detected via QSO absorptions, termed as the Lyman Limit Systems

4 of 38

FEASTS: FAST Extended Atlas for Selected Targets Survey

Data release: all realsed

5 of 38

Combining FEASTS and interferometry to obtain full HI

Procedure and flow chart modified based on Stanimirovic (2002).

Wang+2024b

Linear combination: optimized for NHI deviation

Joint deconvolution: optimized for extracting kinematics

Resi of VLA

Resi of FAST

model

Lin+2025

6 of 38

Around-galaxy distribution & kinematics

7 of 38

Possibility of HI distribution in the 10^17.7<N_HI<10^19.5 cm^-2regime

Low-i

view

Edge-on

view

Filamentary

Truncated

Extending the inner disk

Floating clouds

Halo-like

Flat

8 of 38

The diffuse HI morphology

9 of 38

The surprising regularity of 10^17.7-10²¹ cm^-2HI

Wang+25

2016 @ 1 Msun/pc²～10²⁰ cm^-2

2025 @ 0.01 Msun/pc² ～10¹⁸ cm^-2

Strong constraint to the galactic baryonic flow from such regular and disky morphology.

The inner disk (10²⁰cm^-2) size-mass relationships extends to 10¹⁸cm^-2 (Wang+16,25)。

See further from the disk thickness at 10¹⁸ cm^-2 from Dong’s talk, and a simulation comparison for all these in Xuchen’s talk.

Much better aligned than when radius normalized by other things

10 of 38

The diffuse HI velocity field

11 of 38

More extended rotation curve and dynamically cold HI disk

Li et al. in prep

Yi et al. in prep

~100 kpc

12 of 38

Wang+2024

The general dynamically cool nature of the diffuse HI

The strong implication for HI survival and gas cooling through HI-CGM interaction

A pressure weighted thickness, different from the detectable “maximum” thickness in Yang’ talk

13 of 38

Lin & Wang*+ 2025

See Xuchen

Critcal NHI to survive thermal evaporation

Wang+2023

Possible cooling scenarios

Turbulent mixing cooling, when velocity dispersion is low and N_HI>~10¹⁸ cm^-2

Thermal conduction cooling, when velocity dispersion is high and N_HI >~10¹⁹ cm^-2.

14 of 38

The f(NHI)

Wang+ in prep

l: Length incidence = galaxy density * HI cross section

HI cross section = True HI cross section * covering factor at the resolution

For N_HI>10^17.2cm^-2HI

Not deviating far from z=3, despite the different universe

15 of 38

Short summary

HI at 0.01 Msun/pc² ～10¹⁸ cm^-2follows a tight size-mass relation, and the radial profile normalized with this radius shows the smallest scatter among all normalizing radius tested.

The HI there is dynamicall cold, and fast moving at v_rotin the hot gas halo (CGM)

The HI there is pushed by weight/pressure of the CGM,

possibly reaching a relatively high and almost constant volume density,
possible playing a siginificant role in hot gas cooling and ISM gas accretion,
possible doing that through turbulent mixing and thermal conduction.

16 of 38

The Extraplanar HI revealed by FEASTS

PHISCC2025

2025-9-22

16

Hello everyone, I’m Dong Yang, I’m very glad to come here and introduce our recent work about the extraplanar HI gas. As introduced by Jing, the HI beyond the disk participates in the galaxy baryon cycle and associates with gas accretion, fountain and so on. To study the extraplanar HI gas, we select seven perfectly edge-on galaxies from FEASTS. The edge-on geometry makes it easy to spatially disentangle the HI gas in the disk and outside it, so that we could subtract the contamination from HI disk and focus on the extraplanar HI. We follow the linear-combination method as Jing. As one case, here shows the FEASTS HI moment zero map of the famous edge-on galaxy, NGC 891. We calculate the excess HI by subtracting the Interferometry HI convolved with FAST beam from FESTS data, and we add the excess HI with the interferometry HI to get the full HI distribution. One step further, as we focus on the extraplanar HI, we calculate the HI mean column density profile along the vertical direction, as shown in the left figure. It clearly shows the extraplanar HI detected by FEASTS, which extending to larger vertical distances compared to the HALOGAS.

17 of 38

The extension of the extraplanar HI

PHISCC2025

Marasco+2019

2025-9-22

17

18 of 38

The correlation between the extension and MHI

PHISCC2025

only & tightly correlates with MHI

--> self-similar distribution radially and vertically

partial pearson correlation (r/p):

MHI: 0.83 / 0.002
SFR: 0.41 / 0.2
Mstar: 0.19 / 0.6

2025-9-22

18

19 of 38

Comparison to face-on galaxies

PHISCC2025

2025-9-22

19

Besides the vertical extension, we try to measure the morphology of these diffuse HI down to 10 to the 18. To do so, we compare the full HI distribution of our edge-on sample to the radial profile from face-on galaxies. They show a clear bimodality along these two directions, indicating they are associated with different physical processes. By calculating the slope ratio of the mean profile, we get a very rough estimation of the axis ratio , which is roughly 0.5. We also try two other independent methods to get the axis ratio and the results are similar. Due to the small sample size,, the exact value is not important, but it emphasizes that the HI extends to much larger radii along radial direction rather than vertical direction. in other words, the distribution is much flattener than a spherical model.

20 of 38

The kinematics of the excess HI

PHISCC2025

lagged rotation

consistent with EPG revealed by HALOGAS

Marasco+2019

2025-9-22

20

21 of 38

Comparison to absorption results

PHISCC2025

Large scatter in absorption results
projection effects?

= Ly-alpha absorption

2025-9-22

21

22 of 38

Comparison to GBT results for N891 and N4565

PHISCC2025

Das+2020,2024

2025-9-22

22

23 of 38

Comparison to GBT results for N891 and N4565

PHISCC2025

Beam error / sidelobe?

Pointing error?

2025-9-22

23

24 of 38

Take-home message

PHISCC2025

2025ApJ...984...15Y

2025-9-22

24

25 of 38

25

22 Sept 2025 PHISCC

(Lin+submitted)

(MHONGOOSE vs. simus, Marasco+25, after coffee)

26 of 38

26

22 Sept 2025 PHISCC

TNG (in Diemer+19)

Auriga (in Marinacci+17)

EAGLE (in Bahé+16)

27 of 38

27

22 Sept 2025 PHISCC

28 of 38

28

22 Sept 2025 PHISCC

(Lin+submitted)

29 of 38

29

22 Sept 2025 PHISCC

Irregular

Extended

(Lin+submitted)

30 of 38

30

22 Sept 2025 PHISCC

(Lin+submitted)

31 of 38

31

22 Sept 2025 PHISCC

(Lin+submitted)

X = most important factor

All

Consistent Half

Inconsistent Half

32 of 38

32

22 Sept 2025 PHISCC

(Lin+submitted)

All

Consistent Half

Inconsistent Half

33 of 38

Short summary

33

22 Sept 2025 PHISCC

(van de Voort+19)

(Ramesh+23)

34 of 38

Turbulent-mixing accretion scenario (Lin+25)

34

10/12/2024

v_s

v_ent

Ṁ_in

∼100 pc

(Ji+19, Mandelker+19, 20, Tan+21, Yang & Ji 23)

v_s

h, σ_v

(Ji+19)

35 of 38

Combine FEASTS w/ interferometry (THINGS/HALOGAS):�High sensitivity + high resolution

35

22 Sept 2025 PHISCC

MEM�(mosmem)

dirty�map

SD�map

default�image

model

redisuals

1st iter

correction

output

…

36 of 38

To summarize

37 of 38

Unexpected regularity of 10^17.7~10^19.5 cm^-2 HI at z=0

Low-i

view

Edge-on

view

Filamentary

Truncated

Extending the inner disk

Floating clouds

Halo-like

Flat

✔️

❌

38 of 38

The expectation of FLOWER