---
title: 'Makani Galaxy: Extreme Starburst Winds'
url: https://www.emergentmind.com/topics/makani-galaxy
type: topic
---

# Makani Galaxy: Extreme Starburst Winds

Makani Galaxy, SDSS J211824.06+001729.4, is a compact, massive galaxy at $z = 0.459$ whose defining property is a starburst-driven galactic wind extending from the interstellar medium into the circumgalactic medium on $\sim 100$ kpc scales. It is described as a compact, massive post-starburst galaxy and as a compact merger remnant undergoing an extreme starburst, with stellar mass $M_* \approx 10^{11.1}\,M_\odot$, an effective radius reported as $r_e = 2.3$ kpc or $r_e \approx 2.5$ kpc, and a central starburst of radius $\approx 400$ pc that is consistent with being Eddington-limited. Makani is a benchmark system because its outflow can be decomposed into two star-formation and wind episodes separated in space and time, and because warm ionized gas, molecular gas, neutral gas, O VI/Ly$\alpha$-emitting coronal interfaces, and PAH-bearing warm dust have all been traced across the inner halo and CGM [2303.00194] [2503.20042] [2507.08098].

## 1. Host galaxy and star-formation history

Makani lies at a physical scale of $6.02$ kpc per arcsecond for a flat $\Lambda$CDM cosmology with $\Omega_m = 0.315$ and $H_0 = 67.4\ \mathrm{km\ s^{-1}\ Mpc^{-1}}$. It is massive, compact, and centrally concentrated. The central starburst is extremely compact, intense, and consistent with being Eddington-limited, while the broader system is identified as post-starburst and merger-remnant-like. These descriptions frame Makani as a transition object in which a recent or ongoing starburst coexists with evidence for rapid structural and gaseous transformation [2303.00194] [2507.08098] [2503.20042].

The star-formation rate depends strongly on tracer. Radio and infrared indicators give $\mathrm{SFR} \approx 224$–$300\,M_\odot\,\mathrm{yr}^{-1}$, whereas H$\alpha$ from the nuclear star-forming component gives a much smaller $\mathrm{SFR} \approx 14\,M_\odot\,\mathrm{yr}^{-1}$. The optical study explicitly interprets this discrepancy as suggesting very recent quenching and/or LyC leakage and/or heavy obscuration. A plausible implication is that Makani is being observed during a short-lived stage in which feedback has already altered the observable star-forming signatures without erasing the energetic imprint of the recent burst [2303.00194].

The temporal structure inferred for the central activity consists of two episodes: Episode I, approximately $0.4$ Gyr ago, and Episode II, approximately $7$ Myr ago. The wind properties strongly track this history, so the star-formation chronology is not merely a stellar-population inference but the organizing principle for the gaseous phenomenology seen from the nucleus to the halo [2507.08098].

## 2. Two-episode outflow architecture and phase structure

Makani’s outflow is explicitly resolved into two episodes. Episode I is the outer, older component, with age $\approx 400$ Myr and radial range $R_I \approx 20$–$50$ kpc. It has slowed substantially: projected speeds are $\sim 100\ \mathrm{km\ s^{-1}}$, the linewidth is $\sigma_{\rm wind} \approx 200\ \mathrm{km\ s^{-1}}$, and the mean bulk motion is consistent with ballistic travel over $400$ Myr, $\langle v \rangle \approx R_I/t_{*,I} \approx 120\ \mathrm{km\ s^{-1}}$. Episode II is the inner, recent component, with age $\approx 7$ Myr and radial range $R_{II} \approx 0$–$20$ kpc, although faint [O II] reaches $\sim 40$ kpc; its maximum speeds exceed $2000\ \mathrm{km\ s^{-1}}$, and $\langle v \rangle \approx R_{II}/t_{*,II} \approx 1400\ \mathrm{km\ s^{-1}}$ [2303.00194].

The phase structure is strongly stratified with radius. Warm ionized gas is detected to $r \approx 30$–$40$ kpc in Balmer lines and to $\sim 50$ kpc in [O II]. Neutral and molecular gas, together with dust traced by Balmer decrements, Na I D, Mg II, and CO, are prominent within $\sim 20$–$25$ kpc and are not detected beyond $\sim 25$ kpc in the optical/near-UV studies. The wind-to-galaxy size ratio $r_w/r_e \gtrsim 20$ proves that the outflow extends into the CGM, not merely the stellar body or a local superbubble environment [2303.00194].

Subsequent ultraviolet and infrared observations broadened this phase inventory. Deep HST imaging detects O VI and Ly$\alpha$ emission across the [O II] nebula with similar morphology and extent, out to $r \sim 50$ kpc, while JWST detects PAH-bearing warm dust to $\sim 35$ kpc. These later data show that the outer halo is not only ionized but also hosts coronal cooling interfaces and dust-bearing material, even where earlier slit-based reddening estimates suggested little or no extinction [2503.20042] [2507.08098].

The resulting picture is a temporally resolved, multiphase, CGM-scale wind. Episode II is fast, dusty, neutral, molecular, and strongly multiphase in the inner halo. Episode I is older, slower, and dominated observationally by warm ionized gas and coronal/interface tracers at larger radii. This suggests radial and temporal evolution in which a recently launched compact-starburst wind transitions into a more diffuse and CGM-coupled structure.

## 3. Optical spectroscopy, ionization state, and shocks

The foundational spectroscopic analysis used Keck II/ESI echellette observations with a $1''$ slit, $R \approx 4000$, and simultaneous rest-frame optical coverage from [O II] $\lambda\lambda 3726,3729$ through [S II] $\lambda\lambda 6716,6731$. Balmer lines were detected across the nebula, together with numerous collisionally excited lines including [O III] and its auroral $4363$ Å line, [O I] $6300$ Å, [N II] $5755$ Å, [S II], [Ne III], and [Ne V]. These line detections enabled extinction, density, temperature, and excitation diagnostics across both wind episodes [2303.00194].

The extinction profile is radially structured. Assuming Case B and $R_V = 3.1$, $E(B-V)$ peaks at $\sim 1.0$ around $r \approx 10$ kpc and declines to $\approx 0$ by $r \gtrsim 25$ kpc. Electron densities derived from [S II] $\lambda6716/\lambda6731$ and [O II] $\lambda3729/\lambda3726$ vary strongly with position: in the inner fast wind, $n_e \approx 3000\ \mathrm{cm^{-3}}$ with a $1\sigma$ range of $\approx 1200$–$10{,}000\ \mathrm{cm^{-3}}$, whereas outer apertures have $n_e \lesssim 10\ \mathrm{cm^{-3}}$, weakly constrained below $\approx 50\ \mathrm{cm^{-3}}$. Auroral-line ratios, specifically [O III] $4363/5007 \approx -1.1$ dex and [N II] $5755/6548 \approx -1.3$ to $-1.4$ dex, imply $T \gtrsim 2\times 10^4$ K in the shocked gas [2303.00194].

Line ratios were interpreted with MAPPINGS shock models, using Solar-metallicity fast-shock grids with $n_H = 1\ \mathrm{cm^{-3}}$ for the outer wind and $n_H = 10\ \mathrm{cm^{-3}}$ for the inner wind, at low magnetic parameter $B/n^{1/2} \lesssim 0.5\ \mu\mathrm{G\ cm^{-3/2}}$. Both shock-only and shock+precursor cases were considered, with precursor pre-ionization especially important in the inner wind. In diagnostic diagrams outside the nucleus, apertures lie in composite/LINER regions characterized by high [O I]/H$\alpha$ and [S II]/H$\alpha$, modest [O III]/H$\beta$, and low [O III]/[O II]. The broad nuclear component falls in the AGN region, but Makani shows no multiwavelength AGN signatures; in this system, the AGN-like optical ratios and [Ne V] are explained by fast shocks rather than by a luminous AGN [2303.00194].

The shock speeds inferred from the optical line ratios track the kinematics. Episode II requires $v_{\rm shock} \approx 300$–$400\ \mathrm{km\ s^{-1}}$, consistent with $\sigma_{\rm wind} \approx 400\ \mathrm{km\ s^{-1}}$, together with a hard radiation field and shock+precursor emission. Episode I requires $v_{\rm shock} \approx 200\ \mathrm{km\ s^{-1}}$, consistent with $\sigma_{\rm wind} \approx 200\ \mathrm{km\ s^{-1}}$, with lower ionization parameter and weaker precursor contribution. The empirical relation $v_{\rm shock} \approx \sigma_{\rm wind}$ in both episodes is one of the clearest arguments that shocks, rather than stellar photoionization, dominate the ionization of the extended nebula [2303.00194].

## 4. Mass, momentum, energy, and wind driving

For case B recombination and uniform density, the ionized-gas mass was estimated from the H$\alpha$ luminosity as
$$
M_{\rm HII} \approx \frac{1.4\,m_p\,L_{H\alpha}}{h\nu_{H\alpha}\,\alpha_{H\alpha}^{\rm eff}\,n_e}.
$$
In Makani, the H$\alpha$ luminosity was apportioned between shock precursor and post-shock regions using MAPPINGS guidance, with representative densities of $n_e \approx 10$ and $1000\ \mathrm{cm^{-3}}$ for Episode I and $n_e \approx 1$ and $100\ \mathrm{cm^{-3}}$ for Episode II. This procedure weights the mass toward the lower-density precursor component. The outflow rate was then bracketed with $\dot M \approx Mv/R$ and $\dot M \approx M/t_*$, using deprojected velocities, characteristic radii $R_I \approx 40$ kpc and $R_{II} \approx 15$ kpc, and both $v_{50\%}$ and $v_{98\%}$ to span the dynamics [2303.00194].

The total nebular luminosity is $L_{H\alpha,\rm tot} \approx (9.4^{+6.0}_{-3.0})\times 10^{42}\ \mathrm{erg\ s^{-1}}$, bootstrapped from ESI H$\alpha$ and KCWI [O II]. The star-forming component accounts for $L_{H\alpha} \approx 1.4$–$2.1\times 10^{42}\ \mathrm{erg\ s^{-1}}$, with an adopted value of $\approx 1.8\times 10^{42}\ \mathrm{erg\ s^{-1}}$. For the inner fast wind, the preferred estimates give $M_{II}(\mathrm{HII}) \approx (1$–$2)\times 10^9\,M_\odot$ and $\dot M_{II}(\mathrm{HII}) \approx 170$–$250\,M_\odot\,\mathrm{yr}^{-1}$, comparable to the molecular phase with $M_{II}(\mathrm{H_2}) \approx (2.4 \pm 0.6)\times 10^9\,M_\odot$ and $\dot M_{II}(\mathrm{H_2}) \approx 245\,M_\odot\,\mathrm{yr}^{-1}$. Representative kinematics are $\langle v_{\rm rad}\rangle \approx 2100$–$2300\ \mathrm{km\ s^{-1}}$; for the ionized component alone, $\dot p_{II}(\mathrm{HII}) \approx 2.6\times 10^{36}\ \mathrm{dyn}$ and $\dot E_{II}(\mathrm{HII}) \approx 8.8\times 10^{44}\ \mathrm{erg\ s^{-1}}$ [2303.00194].

For the outer slow wind, the preferred estimates give $M_I(\mathrm{HII}) \approx 5\times 10^9\,M_\odot$ and $\dot M_I(\mathrm{HII}) \approx 10$–$13\,M_\odot\,\mathrm{yr}^{-1}$, with representative kinematics $\langle v_{\rm rad}\rangle \approx 90$–$100\ \mathrm{km\ s^{-1}}$. The corresponding rates are much smaller than in Episode II: $\dot p_I(\mathrm{HII}) \approx 6.9\times 10^{33}\ \mathrm{dyn}$ and $\dot E_I(\mathrm{HII}) \approx 4.2\times 10^{41}\ \mathrm{erg\ s^{-1}}$. Makani therefore combines a very massive outer ionized reservoir with a relatively small present-day outer-wind mass flux [2303.00194].

The shock-radiated power was estimated from
$$
\frac{L_{\rm shock}}{L_{H\alpha}} = 107\,v_{s,100}^{0.59}\left(1 + 1.32\,v_{s,100}^{-0.13}\right)^{-1},
$$
with $v_{s,100} = v_{\rm shock}/(100\ \mathrm{km\ s^{-1}})$. For Episode II, $v_{\rm shock} \approx 400\ \mathrm{km\ s^{-1}}$ implies $L_{\rm shock,II} \approx 1.15\times 10^2 L_{H\alpha,II} \approx 6.6\times 10^{44}\ \mathrm{erg\ s^{-1}}$, while the mechanical power in the ionized+molecular wind, $\dot E_{II}(\mathrm{HII+H_2}) \approx 1.8\times 10^{45}\ \mathrm{erg\ s^{-1}}$, is sufficient to power the observed line emission. For Episode I, $v_{\rm shock} \approx 200\ \mathrm{km\ s^{-1}}$ implies $L_{\rm shock,I} \approx 7.3\times 10^1 L_{H\alpha,I} \approx 1.6\times 10^{44}\ \mathrm{erg\ s^{-1}}$, whereas $\dot E_I(\mathrm{HII}) \approx 0.003\,L_{\rm shock,I}$. The optical analysis therefore concludes that much of the energy driving Episode I shocks must reside in an unseen hotter phase or has propagated to larger CGM radii [2303.00194].

The driving mechanism is cast in terms of the momentum boost,
$$
\mathrm{boost} \equiv \dot p_{\rm out}/(L/c).
$$
Using $L_{\rm bol} \approx 2L_{\rm IR} \approx 1.16\times 10^{46}\ \mathrm{erg\ s^{-1}}$, so that $L/c \approx 3.8\times 10^{35}\ \mathrm{dyn}$, the inner outflow has a momentum boost of $\approx 7$. The interpretation advanced for Makani is a momentum-driven flow supplied jointly by hot ejecta and radiation pressure from the compact Eddington-limited starburst, without requiring a clear AGN [2303.00194].

## 5. Ultraviolet line emission and coronal cooling in the CGM

Deep HST/ACS-SBC imaging with the F150LP and F165LP long-pass filters provided a differential narrow-band method for separating O VI from Ly$\alpha$. At Makani’s redshift, O VI $\lambda\lambda 1032,1038$ shifts to $\sim 1505$ and $1514$ Å and is included in F150LP but excluded from F165LP, whereas Ly$\alpha$ at $\sim 1774$ Å lies in both filters. The critical practical result is that F150LP traces O VI+Ly$\alpha$, F165LP traces Ly$\alpha$, and the filter difference isolates O VI with minimal Ly$\alpha$ leakage. After dark-current control, background modeling, drizzling to the KCWI scale, and Voronoi binning, O VI and Ly$\alpha$ emission were detected across the [O II] nebula with similar morphology and extent, out to $r \sim 50$ kpc [2503.20042].

The morphology is highly constraining. The F150LP emission reproduces the [O II] hourglass, including four lobes and the northern cavity, to $r \approx 8$–$9''$, or $48$–$54$ kpc. The inner $r \lesssim 1$–$1.5''$ region is dominated by continuum and/or Ly$\alpha$ from the compact starburst, but beyond $r \approx 1.5''$ the emission is line-dominated. In the extended nebula the measured count-rate ratio is $R \equiv C_{150}/C_{165} = 1.91 \pm 0.13$ at $1.5''$–$5''$ and $2.33 \pm 0.62$ at $5''$–$7''$, implying that O VI contributes $\approx 44$–$51\%$ of F150LP counts in the inner extended zone and $42$–$66\%$ in the outer zone [2503.20042].

The adopted line-separation relations are
$$
C_{150} = \mathcal{C}\left(F_{\rm OVI}T^{150}_{\rm OVI} + F_{\rm Ly\alpha}T^{150}_{\rm Ly\alpha}\right), \qquad
C_{165} = \mathcal{C}\left(F_{\rm Ly\alpha}T^{165}_{\rm Ly\alpha}\right),
$$
with
$$
F_{\rm OVI} = (7.68\times 10^{-15}\ \mathrm{erg\ s^{-1}\ cm^{-2}})\,C_{150}\,[1 - 1.07\,R^{-1}],
$$
$$
F_{\rm Ly\alpha} = (3.20\times 10^{-14}\ \mathrm{erg\ s^{-1}\ cm^{-2}})\,C_{150}\,R^{-1},
$$
and
$$
F_{\rm OVI}/F_{\rm Ly\alpha} = 0.20\,(0.93R - 1).
$$
After correction for Milky Way extinction with $E(B-V)=0.0587$, the extended-nebula ratios are $F_{\rm OVI}/F_{\rm Ly\alpha} = 0.16 \pm 0.03$ at $1.5''$–$5''$ and $0.24 \pm 0.12$ at $5''$–$7''$. Summing the detected bins gives $L_{\rm OVI} = (2.42 \pm 0.09)\times 10^{42}\ \mathrm{erg\ s^{-1}}$, while model extrapolation gives $L_{\rm OVI} = 3.67\times 10^{42}\ \mathrm{erg\ s^{-1}}$ integrated to $60$ kpc and $4.06\times 10^{42}\ \mathrm{erg\ s^{-1}}$ integrated to infinity; the adopted value is $L_{\rm OVI} \approx 4\times 10^{42}\ \mathrm{erg\ s^{-1}}$. This is comparable to $L([\mathrm{O\,II}]) = 3.3\times 10^{42}\ \mathrm{erg\ s^{-1}}$ [2503.20042].

The radial surface brightness of O VI is well described by
$$
\Sigma(r) = \Sigma_0^{\rm in}\,e^{-r/r_{\rm exp}^{\rm in}} + \Sigma_0^{\rm out}\,e^{-r/r_{\rm exp}^{\rm out}},
$$
with $\Sigma_0^{\rm in} = 3.42\times 10^{-16}\ \mathrm{erg\ s^{-1}\ cm^{-2}\ arcsec^{-2}}$ (fixed), $r_{\rm exp}^{\rm in} = 4.29 \pm 0.34$ kpc, $\Sigma_0^{\rm out} = (0.78 \pm 0.10)\times 10^{-16}\ \mathrm{erg\ s^{-1}\ cm^{-2}\ arcsec^{-2}}$, and $r_{\rm exp}^{\rm out} = 16.68 \pm 1.01$ kpc. The O VI half-light radius is $r_{\rm eff} \approx 19$–$21$ kpc, similar to [O II] at $\approx 18$ kpc. Only $\sim 13\%$ of the total O VI flux lies within the compact central $1''$ aperture, while the remaining $\sim 85\%$ is extended, which the UV study takes as evidence for in-situ O VI excitation rather than dominant resonant scattering of nuclear light [2503.20042].

Physically, O VI is interpreted as tracing radiative cooling at $T \approx 10^{5.5}$ K in hot-cold interfaces, where the $T \gtrsim 10^7$ K CGM or hot wind exchanges mass with $T \approx 10^4$ K clouds. Using a metal-line cooling coefficient $\Lambda^{Z_\odot}_{\rm OVI} \approx 5\times 10^{-23}\ \mathrm{erg\ s^{-1}\ cm^3}$ and
$$
M_{\rm OVI} = (1.32\times 10^5\,M_\odot)\left(\frac{L_{\rm OVI}}{10^{42}\ \mathrm{erg\ s^{-1}}}\right)\left(\frac{n_e}{1\ \mathrm{cm^{-3}}}\right)\left(\frac{Z_\odot}{Z}\right),
$$
the paper notes that $L_{\rm OVI} \approx 4\times 10^{42}\ \mathrm{erg\ s^{-1}}$ and $Z \approx Z_\odot/3$ imply $n_e \approx 0.5\ \mathrm{cm^{-3}}$ if the O VI mass equals the ensemble halo O VI mass of $\approx 3\times 10^6\,M_\odot$. With $t_{\rm cool} \approx 1$ Myr for O VI-bearing coronal gas, the resulting oxygen-phase cooling rate is $\dot M^{\rm OVI}_{\rm cool} \approx 3\,M_\odot\,\mathrm{yr}^{-1}$, and scaling to total gas gives $\dot M_{\rm cool} \approx 1500\,M_\odot\,\mathrm{yr}^{-1}$ for $Z = Z_\odot/3$. This suggests very strong mass exchange and cloud growth in the outer wind [2503.20042].

## 6. Warm dust, PAHs, and dust survival to tens of kiloparsecs

JWST NIRCam and MIRI imaging exploited a coincidental redshift match between Makani’s PAH features and standard imaging filters. At $z = 0.459$, PAH $3.3\ \mu$m falls in NIRCam F480M, PAH $7.7\ \mu$m in MIRI F1130W, PAH $(11.3+12.2)\ \mu$m in MIRI F1800W, H$_2$ $0$–$0$ S(1) $17.03\ \mu$m in MIRI F2550W, and Pa$\beta$ in NIRCam F187N. “Off-band” continuum windows were provided by MIRI F770W and F2100W. After PSF subtraction with STPSF and photutils PSFPhotometry, PAH $7.7\ \mu$m was detected to $\sim 35$ kpc, PAH $(11.3+12.2)\ \mu$m to $\sim 30$ kpc, and PAH $3.3\ \mu$m to $\sim 20$ kpc [2507.08098].

The spatial relation to other phases is important. Warm dust extends well beyond the inner CO(2–1) and Mg II emission at $\sim 20$ kpc, but does not reach as far as the outer [O II] nebula at up to $\sim 50$ kpc. Within $\sim 20$ kpc, the flux ratios F1130W/CO(2–1), F1800W/CO(2–1), F1130W/Mg II, and F1800W/Mg II remain roughly constant, suggesting co-spatial dust with cool gas in the inner wind at similar relative strengths. Beyond $\sim 20$ kpc, PAHs persist where CO and Mg II are not detected in the current ALMA/KCWI data. A NW halo cloud complex is evident in F1130W and F1800W out to $30$–$35$ kpc, and the extended emission is clumpy and asymmetric [2507.08098].

The primary quantitative dust diagnostic is
$$
R_{\rm PAH} \equiv \frac{F1800W}{F1130W} \approx \frac{I_{11.3}+I_{12.2}}{I_{7.7}}.
$$
Measured values are $R_{\rm PAH} = 1.40 \pm 0.02$ in the nucleus, $0.830 \pm 0.004$ in the inner halo at $R = 10.5$–$17.6$ kpc, $0.70 \pm 0.10$ in the outer halo cloud CGM-E, and $\lesssim 0.11$ in CGM-W. Because the $11.3$ and $12.2\ \mu$m bands are stronger in neutral and larger PAHs, whereas $7.7\ \mu$m strengthens in ionized or smaller PAHs under harder radiation fields, the observed radial decline in $R_{\rm PAH}$ was modeled as indicating decreasing starlight intensity, decreasing PAH sizes, and increasing PAH ionization fractions with increasing distance from the nucleus [2507.08098].

The modeling compared measured filter ratios to the PAH+dust spectral library of Draine (2021), redshifted to $z = 0.459$ and convolved with the exact JWST throughputs, across starlight intensity $U$, PAH size distribution $f_{\rm size}$, and PAH ionization fraction $f_{\rm ion}$. In the nucleus, high $R_{\rm PAH} \approx 1.4$ and elevated continuum ratios are consistent with large $U$, larger grains, and lower PAH ionization fractions. In the inner and outer halo, lower $R_{\rm PAH}$ values require lower $U$ with $\log U \lesssim 0$, smaller average PAH sizes than the standard distribution, and elevated PAH ionization fractions relative to the nucleus. No dust temperatures or masses were reported [2507.08098].

The dust-survival problem is acute. For representative values $R \approx 30$ kpc and $v \approx 300\ \mathrm{km\ s^{-1}}$, the travel time is $t \approx 0.1$ Gyr, consistent with $R/v \approx 10^8$ yr. By contrast, the adopted thermal sputtering time,
$$
t_{\rm sp} \approx 0.17\ \mathrm{Gyr}\left(\frac{a}{0.1\,\mu\mathrm{m}}\right)\left(\frac{10^{-27}\ \mathrm{g\ cm^{-3}}}{\rho}\right)\left[\left(\frac{10^{6.3}\ \mathrm{K}}{T}\right)^\omega + 1\right],
$$
with $\omega = 2.5$, gives $t_{\rm sp} \approx 10^4$–$10^5$ yr for PAH-sized grains with $a \approx 0.001$–$0.01\ \mu$m under the densities and temperatures adopted for the hot phase, far shorter than the travel time. The observed radial decline in $R_{\rm PAH}$, reduced F480M detections, and low halo continuum are therefore interpreted as evidence that PAHs survive to $30$–$35$ kpc but are eroded and processed during transport. A possible survival mechanism proposed in the JWST study is shielding in cloud-wind mixing layers, where dust can survive for $\gtrsim 80\,t_{\rm sp}$ if condensation and cooling are efficient [2507.08098].

The JWST detections also revise the earlier radial dust picture. Optical slit data had found $E(B-V) \approx 0$ by $r \gtrsim 25$ kpc along the slit, but the PAH maps directly reveal dust in the outer warm-ionized wind and note that the slit may have missed parts of the NW PAH clouds. This is not a contradiction in the strict sense; it is a demonstration that slit-based extinction measurements did not fully sample the three-dimensional dust distribution [2507.08098].

## 7. Uncertainties, interpretive limits, and broader significance

Each observational window carries substantial systematics. In the optical, uncertainties arise from extinction corrections, especially at large radius, from the $\approx 0.3$ dex RMS scatter in the H$\alpha$–[O II] bootstrap, from difficult low-S/N density measurements in the outer wind, from the adopted two-phase precursor/post-shock density model, from shock-model degeneracies involving precursor fractions and magnetic parameter, and from the geometry and deprojection assumptions used to derive $\dot M$, $\dot p$, and $\dot E$. Flux calibration differences between ESI and SDSS/KCWI and the assumed $40\%$ subtraction of star-forming H$\alpha$ in the inner nebula are additional systematic terms [2303.00194].

In the ultraviolet, the main limitations are internal dust attenuation, which is poorly constrained spatially in the UV; Ly$\alpha$ radiative transfer, which can depress intrinsic O VI/Ly$\alpha$ ratios by adding scattered Ly$\alpha$; the noise-dominated nature of the pure F150LP–F165LP difference image beyond $r \approx 2.5''$; residual dark-current gradients in ACS-SBC at extremely low surface brightness; and the possibility that up to $\sim 20\%$ of the total O VI light could arise from dust-scattered UV continuum if the entire inner exponential were due to scattering. The UV analysis therefore treats shock ionization as plausible rather than uniquely established, with the favored shock parameters depending sensitively on extinction and Ly$\alpha$ transport [2503.20042].

In the infrared, the dominant issues are PSF-subtraction residuals, especially at long wavelength; lower-than-expected F2550W sensitivity; the fact that filters integrate across broad PAH features and were analyzed without explicit continuum subtraction; and the possibility that current ALMA/KCWI limits are too shallow to exclude co-spatial cool gas beyond $\sim 20$ kpc. The JWST study also notes that radial $R_{\rm PAH}$ trends, while favoring erosion and increased ionization, could be influenced by varying radiation hardness or local shock conditions [2507.08098].

Despite these caveats, Makani occupies an unusual position in feedback studies. The optical work describes it as the poster child of a galactic wind on scales of the circumgalactic medium; the UV work presents the first imaging detection of spatially extended O VI emission coextensive with a $100$ kpc wind in a massive galaxy’s CGM and only the second resolved O VI emission halo imaged for any galaxy; and the JWST work reports direct PAH detections to $\sim 35$ kpc, providing strong evidence that ejected dust can survive to CGM scales while being processed in transit [2303.00194] [2503.20042] [2507.08098].

Taken together, the Makani data define a physically coherent but still incomplete feedback case study. Episode II is a fast, massive, dusty, multiphase inner-halo wind whose warm ionized mass and outflow rate are comparable to those of the molecular phase and whose momentum budget is consistent with a momentum-driven flow from a compact Eddington-limited starburst. Episode I is an older, slower CGM-scale outflow containing a large ionized reservoir, coronal O VI interfaces, and PAH-bearing dust, but with evidence that much of its energy and momentum reside in a hotter phase or at still larger radii. This suggests that Makani is best understood not as a single outflow snapshot but as a resolved feedback sequence linking compact-starburst launch physics, shock-powered ionization, multiphase mass exchange, CGM cooling, and dust transport over $\sim 10^8$ yr.

Source: https://www.emergentmind.com/topics/makani-galaxy