In this A&A Letter to the Editor I show that the infamous [CII]-deficit (relative to the far-IR flux) can be explained by [CII] self-absorption without the need for secondary effects using SOFIA/upGREAT observations. We report the discovery that S144, a bubble‑shaped source embedded in the RCW79 HII region, is predominantly “filled” with ionized carbon and excited by a single O7.5–9.5 V/III star, indicating an early evolutionary stage before significant wind‑blown cavities form.

We modeled the [CII] emission with the SimLine non‑LTE radiative transfer code and decomposed it into a central fully ionized HII region and two high‑density (∼2500 cm⁻³) PDR layers. We found that the inner PDR shell expands at ∼2.6 km s⁻¹ and that the outermost shell exhibits a steep temperature and velocity gradient, leading to a high optical depth of τ ∼ 4 and self‑absorption features. We studied spatially averaged [¹²CII] and [¹³CII] F = 1–0 hyperfine spectra to derive a lower limit on the excitation temperature (Tₑₓ ≳ 54 K) and confirm significant optical depth effects.

I developed a procedure to reconstruct the missing [CII] flux by modeling the line wings and obtained correction factors ranging from 1.1 to 1.4. We examined the correlation between the corrected [CII] intensities and total FIR continuum and found that the restored data follow a linear relation without a [CII]‑deficit. We argue that alternative explanations for [CII]‑deficits, such as high dust optical depth or reduced photoelectric heating efficiency, are less likely under the moderate density and radiation field conditions of S144. We show that self‑absorption by cooler C⁺ layers along the line of sight can account for the apparent [CII]‑deficit in this PDR region. We conclude that self‑absorption must be accounted for when interpreting [CII]‑deficits in Galactic HII bubbles and that broader surveys are needed to quantify the prevalence of this effect.

Read the full A&A letter here or here.

The movies produced within this research project are listed below:

I found that the most massive associations reach ∼8×10⁶ M⊙ in dust and ∼5×10⁶ M⊙ in CO, but M33 lacks the very high‑mass (\(>10^6\,\mathrm{M_\odot}\)) GMC population seen in the Milky Way. Mean surface mass densities are \(22\,\mathrm{M_\odot\,pc^{-2}}\) (dust) and \(16\,\mathrm{M_\odot\,pc^{-2}}\) (CO), roughly an order of magnitude lower than typical Milky Way values, while beam‑averaged number densities remain low (∼5 cm⁻³ and ∼3 cm⁻³), reflecting unresolved substructure. Comparing to Milky Way catalogs, we find a similar upper limit of ∼150 pc on GMC/association size, suggesting a common regulating mechanism—perhaps disk scale height or stellar feedback. We observe no strong gradients in GMC properties with galactocentric radius, although central clouds are modestly denser and more massive, containing ∼30% of M33’s molecular mass. The GMC mass spectrum across the disk follows a power law of index α ≈ 2.3 (dust) and α ≈ 1.9 (CO), consistent with self‑similar structure and hierarchical fragmentation. Our results indicate that, while GMCs in M33 are broadly similar to those in the Milky Way in terms of size distribution and mass spectrum, M33’s lower overall metallicity and galaxy mass lead to systematically lower masses and densities and a deficit of the most massive clouds. These findings imply that basic GMC scaling relations hold across environments, but the upper end of the cloud mass function is sensitive to global galaxy properties.

Read the full A&A article here.

Both methods yielded consistent total column density maps which, after subtracting a VLA H I component, produced H₂ column density maps tracing the molecular gas distribution across M33. I found that the SED‑based method recovers more granular structure in the inter‑arm and outer disk regions, whereas the single‑band approach delivers smoother, more extended emission.

By dividing the IRAM CO(2–1) integrated intensity map by these dust‑derived H₂ column densities, I produced a spatially resolved CO‑to‑H₂ conversion factor (\(X_\mathrm{CO}\)) map, revealing strong local variations around a disk‑average of 1.8×10²⁰ cm⁻² (K km s⁻¹)⁻¹. I extracted column‑density probability distribution functions from the NH, NH₂, and H I maps and found predominantly log‑normal shapes in the diffuse ISM with emerging power‑law tails at high column densities, indicative of self‑gravitating gas. The variable‑β, variable‑κ₀ framework I developed allowed us to build an intrinsic dust‑to‑gas ratio map that captures metallicity‑driven trends without imposing fixed values. I assessed biases in both methods and compared our results to existing literature maps, demonstrating improved depth and resolution. The resulting high‑resolution products afford a refined view of giant molecular cloud complexes and their relation to the spiral‑arm structure in M33. We conclude that these new datasets provide a robust foundation for future studies of cloud matching, star formation, and ISM structure in nearby galaxies, and we publicly release all column density maps for community use.

Read the full A&A article here.

Recasting the D0 measurement of the dijet mass spectrum with \(\eta\) max up to 2.4 using 0.7 fb⁻¹ of data, we generated signal templates in MadGraph5 and applied the same kinematic selections and NLO background as the experiment. After combining statistical, systematic, and theoretical errors, we found that no meaningful bound on the Wilson coefficients can be derived from these data. We then projected the potential reach of a full 10 fb⁻¹ dataset under optimistic and conservative assumptions about systematic‐error scaling and again saw no sensitivity to new contact interactions. Our results demonstrate the general difficulty of obtaining model‐independent SMEFT constraints from relatively low‐luminosity, systematics‑limited hadron‐collider measurements. We conclude that Tevatron dijet data alone cannot probe four‑quark contact operators at the cutoff scales of interest, underscoring the need for higher‑precision or higher‑energy observables—such as those from the LHC—to place reliable EFT bounds.

Read the full JHEP article here.