Intrinsic galaxy alignments in CAMELS simulations and the significant impact of baryon model

TL;DR

The paper provides a model-agnostic detection of intrinsic galaxy alignments in the CAMELS simulations and demonstrates that alignment amplitudes depend on cosmological parameters ($Ω_m$, $σ_8$) and supernova feedback ($A_{\text{SN1}}$, $A_{\text{SN2}}$), while showing no clear AGN impact in the current small-volume box. By analyzing projected cross-correlations $w_{m+}$ and their ratio to the matter autocorrelation $w_{mm}$, the work finds that $σ_8$-driven dependence largely cancels in the ratio as predicted by linear alignment models, yet supernova feedback leaves a robust imprint. The study further reveals that quiescent galaxies exhibit much stronger alignments than star-forming ones and that normalizing ellipticity magnitudes (orientation-only) preserves SN sensitivity, indicating multiple alignment mechanisms across galaxy types. These results highlight the need to jointly model cosmology and baryonic physics when interpreting weak-lensing signals and offer practical diagnostics for mitigating IA systematics in future surveys.

Abstract

We present a detection of the intrinsic galaxy alignments in the CAMELS suite of hydrodynamic simulations. We find that the alignment amplitude depends significantly on cosmological and supernova feedback parameters - specifically $Ω_m$, $σ_8$, $A_{\text{SN1}}$, $A_{\text{SN2}}$- while no dependence on AGN feedback is observed (due to the limited simulation volume $(25\,h^{-1}\,\text{Mpc})^3$). The dependence on $σ_8$ vanishes when projected correlation functions $w_{m+}$ are normalized by matter density correlations $w_{mm}$, consistent with predictions from linear alignment models. We find alignment amplitudes in quiescent galaxies to exceed those in star-forming galaxies by an order of magnitude. Moreover, examining orientation-only correlation functions from ellipticity-normalized galaxies $\tilde w_{m+}$, we confirm that alignment signals retain sensitivity to supernova feedback across full, star forming, quiescent, and ellipticity-normalized samples. Finally, we find evidence that supernova feedback impacts alignment signals differently in star-forming versus quiescent populations, suggesting that distinct alignment mechanisms operate across galaxy types. Our results offer key insights for understanding galaxy formation and alignment models for future weak gravitational lensing analyses.

Figures (12)

• Figure 1: A $1.67\,h^{-1}\,\text{Mpc}$ thick slice of the CAMELS LH-310 simulation viewed along the $y$-axis. The background shows the log-normal dark matter density with stellar density overlaid. Galaxy shapes are represented as ellipses with eccentricity and orientation encoding the magnitude and angle of the measured ellipticity $\varepsilon$. On average, galaxy shapes tend to align preferentially with the principal axes of the dark matter distribution.
• Figure 2: Top: tangential-component correlation function averaged over all simulated catalogues $w_{m+}(r_\perp)$ for three galaxy samples: all galaxies, quiescent galaxies, and star forming galaxies. Bottom: dark matter auto-correlation function $w_{mm}(r_{\perp})$ averaged over all simulated catalogues. $r_\perp$ values are plotted at the logarithmic centre of each of the six bins and a linear interpolation is added to guide the eye. All error bars are the standard error from $N_{\text{cat}}=3000$ catalogues.
• Figure 3: Orientation-only correlation functions averaged over all simulated galaxy catalogues $\tilde{w}_{m+}(r_\perp)$, who's catalogues have the magnitudes of galaxy ellipticities set to constant $\langle |\varepsilon|\rangle$. Full, quiescent and star forming galaxy samples are displayed. Correlation functions here take the same shape as Figure \ref{['fig:mean_corr1']} indicating that the magnitudes of the ellipticities predominantly impact alignment amplitude. Linear interpolation is added to guide the eye.
• Figure 4: Cross-component correlation function averaged over all catalogues $w_{m\times}(r_\perp)$ for full, quiescent and star forming galaxy samples. This null-test measurement is consistent with zero, as expected. Linear interpolation is added to guide the eye.
• Figure 5: Correlation functions weighted and averaged over transverse distances $0.3\le r_\perp\le20$, $\langle w^{(i)}_{m+} \rangle_{r_\perp}$, versus each of six simulation parameters for all simulation catalogues $i$. A density histogram overlays the scatter plot and a linear fit with Pearson value $\rho$ measures linear correlation. The $p$-value $p<10^{-4}$ suggests a statistically significant correlation between gravitational alignments and simulation parameter. Note: the $\langle w^{(i)}_{m+} \rangle_{r_\perp}$ distribution over $i$ has a long tail to high values, so the red linear fit positioned at the mean is not aligned with the mode.