Unveiling the evolution of the CO excitation ladder through cross-correlation of CONCERTO-like experiments and galaxy redshift surveys

This study demonstrates that cross-correlating CONCERTO-like millimeter-wave line-intensity mapping data with galaxy redshift surveys can accurately recover the cosmic CO excitation ladder and molecular gas density up to $z=3$, although the specific CONCERTO experiment lacks the sensitivity to detect the required cross-power spectra.

The Gist

Here is an explanation of the paper, translated from complex astrophysics into everyday language using analogies.

The Big Picture: Listening to the Cosmic Hum

Imagine the universe isn't just a collection of distinct stars and galaxies, but a giant, foggy room filled with a low, constant hum. This "hum" is made of light emitted by cold gas clouds (mostly hydrogen) where new stars are being born.

Astronomers want to understand this gas because it's the fuel for stars. However, looking at individual galaxies is like trying to hear a single person whisper in a crowded stadium. It's hard, and you miss the people who are too quiet to be seen.

Line Intensity Mapping (LIM) is a new technique that doesn't try to hear the whispers; instead, it listens to the entire hum of the stadium at once. It measures the total glow of the gas across huge patches of sky.

The Problem: The "Static" and the "Ghost"

There are two big problems with listening to this cosmic hum:

1. The Static: The signal is very faint and gets drowned out by other sources of noise (like dust or other types of light).
2. The Ghosts (Interlopers): Imagine you are trying to listen to a specific note played by a violin. But, a trumpet playing a different note happens to sound exactly the same pitch because of a trick of the ear (redshift). In astronomy, light from a different galaxy or a different type of atom can "mimic" the signal you are looking for. These are called interlopers. They mess up your data, making it look like there is more gas than there actually is.

The Solution: The "Guest List" Cross-Check

This paper proposes a clever trick to solve the "Ghost" problem. It's like trying to find a specific group of people in a massive, dark crowd.

• The LIM Data: This is the blurry photo of the whole crowd's glow.
• The Galaxy Survey: This is a detailed "guest list" (a catalog of known galaxies with precise locations).

The authors suggest cross-correlating these two. They ask: "Where the guest list says a galaxy exists, does the blurry photo also show a glow?"

If the glow appears exactly where the guest list says a galaxy is, it's real. If the glow appears in a spot where the guest list says "no one is here," it's likely a "ghost" (an interloper) and can be ignored. By matching the two, they can filter out the noise and isolate the true signal of the cold gas.

What They Did: The Simulation Test

Since we can't actually build a perfect telescope yet, the authors built a virtual universe (a computer simulation called SIDES).

1. They created 12 different "patches" of this virtual sky, complete with billions of virtual galaxies and gas clouds.
2. They simulated what a telescope (like the CONCERTO experiment) would see, including all the "ghosts" and noise.
3. They applied their "Guest List" cross-check method to this fake data.

The Results: What Did They Learn?

1. It Works (Mostly): They successfully reconstructed the "SLED" (Spectral Line Energy Distribution). Think of the SLED as the musical fingerprint of the gas. It tells us how "excited" the gas is. Are the gas clouds calm and cool, or hot and turbulent? Their method could read this fingerprint accurately up to a certain level of complexity (up to $J_{up}=6$).
2. The "Starburst" Surprise: They found that rare, violent galaxies (called "starbursts") act like loud rock bands. They don't make up many of the galaxies, but they contribute a huge amount of the high-energy "noise" in the signal. However, because the method averages everything out, these loud bands didn't ruin the overall measurement of the total gas in the universe.
3. The "Ghost" Trap: They found that one specific type of gas signal (CO 7-6) is almost impossible to separate from a "ghost" signal (Carbon I). It's like trying to distinguish a violin from a flute when they are playing the exact same note in the same room. The method couldn't tell them apart.
4. The Telescope is Too Weak: Finally, they checked if the current CONCERTO telescope is strong enough to do this in real life. The answer is no. The telescope isn't sensitive enough yet to hear the signal over the noise, even with the "Guest List" trick. It's like trying to hear a whisper in a hurricane; you need a much quieter room (a more sensitive telescope) or a louder whisper (a bigger survey).

The Takeaway

This paper is a proof of concept. It says: "If we build a better telescope, this 'Guest List' method will allow us to map the cold gas of the universe, understand how stars form, and measure the total amount of fuel available for the cosmos, without getting confused by the ghosts in the machine."

It's a blueprint for how to listen to the universe's background noise and turn it into a clear story about how galaxies grow and evolve.

Technical Summary

Here is a detailed technical summary of the paper "Unveiling the evolution of the CO excitation ladder through cross-correlation of CONCERTO-like experiments and galaxy redshift surveys."

1. Problem Statement

Line Intensity Mapping (LIM) is a powerful technique for mapping the global distribution of molecular gas (traced by Carbon Monoxide, CO) across cosmic time. However, millimeter-wave LIM experiments face two major challenges:

1. Contamination: The observed signal is a superposition of CO lines from different redshifts and interloping lines (e.g., [CII], [CI]) from other sources, making it difficult to isolate specific transitions.
2. Excitation Uncertainty: To derive the cosmic molecular gas density ($\rho_{H_2}$) from high-$J$ CO transitions (where $J_{up} \ge 2$), one must know the CO Spectral Line Energy Distribution (SLED) to correct for the excitation state of the gas. Current models often predict total signals but struggle to constrain individual line contributions or the specific SLED evolution without direct detection of faint sources.

The paper addresses whether cross-correlating LIM data with spectroscopic galaxy surveys can isolate individual CO line signals, reconstruct the background SLED, and accurately measure $\rho_{H_2}(z)$ up to $z=3$, while assessing the feasibility of this method with current/future instruments like CONCERTO.

2. Methodology

Data Simulation (SIDES-Uchuu)

The authors utilized the SIDES (Simulated Infrared Extragalactic Sky) simulation coupled with the Uchuu N-body simulation.

• Scope: 12 independent light cones of $9 \text{ deg}^2$each, covering redshifts$z=0.5$to$3.0$.
• Galaxy Population: Includes Main Sequence (MS) and Starburst (SB) galaxies, with physical parameters (SFR, stellar mass, dust) derived from abundance matching and empirical relations.
• SLED Modeling:
  • MS galaxies: Modeled using a combination of clumpy and diffuse gas components.
  • SB galaxies: Modeled using specific templates (Birkin et al. 2021).
  • Interlopers: [CII], [CI], and other CO lines were included to simulate realistic contamination.
• Mock Observations: Generated spectral cubes for CO transitions ($J_{up}=1$ to $8$) and galaxy number-density contrast cubes (simulating a spectroscopic survey with$K_s \le 24.0$, complete to$10^{10} M_\odot$).

Analysis Framework

1. Cross-Power Spectrum ($P_{cross}$): The authors computed the angular cross-power spectrum between the CO intensity maps and the galaxy density maps. This isolates emission from the same cosmic volume, effectively suppressing uncorrelated interloper noise (which acts as variance rather than bias).
2. SLED Reconstruction: By taking the ratio of cross-power spectra between different CO transitions (e.g., $P_{J_{up}-gal} / P_{J_{ref}-gal}$), the clustering bias and matter power spectrum terms cancel out, allowing direct recovery of the background intensity ratio (SLED).
3. Bias-Weighted Intensity: The cross-power amplitude was used to derive the product of the effective clustering bias ($b_{eff}$) and the mean background intensity ($B_\nu$).
4. $\rho_{H_2}$ Estimation: The derived intensities were converted to molecular gas density using a two-component $\alpha_{CO}$ conversion factor (different for MS and SB galaxies) and excitation corrections relative to the CO(1-0) line.
5. Feasibility Study: The required Noise-Equivalent Intensity (NEI) for a CONCERTO-like experiment to detect these cross-signals was calculated using a spherical power spectrum formalism, accounting for spectral resolution limits.

3. Key Contributions

• Methodological Validation: Demonstrated that cross-correlation with galaxy surveys can successfully recover the CO background SLED and bias-weighted intensities up to $J_{up}=6$ with uncertainties $\le 20\%$, even in the presence of significant line interlopers.
• Excitation Correction: Showed that the effective clustering bias ($b_{eff}$) is consistent across different CO transitions ($J_{up}=1-8$), validating the assumption that the ratio of cross-power spectra directly yields the SLED ratio without needing complex bias modeling for each line.
• Starburst Impact: Quantified the contribution of rare starburst galaxies to the CO background. While they contribute significantly to high-$J$ transitions ($J_{up} \ge 6$), they do not significantly bias the estimation of $\rho_{H_2}$ because their contribution is naturally accounted for in the mean SLED derived from the cross-correlation.
• Instrumental Constraints: Provided a rigorous sensitivity analysis showing that current CONCERTO parameters are insufficient for this specific measurement, highlighting the need for higher sensitivity or larger survey areas.

4. Key Results

• SLED Recovery: The method accurately recovers the CO background SLED up to $J_{up}=6$ with relative differences of $\le 20\%$ compared to the intrinsic simulation values.
  • Limitation: CO(7-6) is heavily contaminated by the CI line (contributing ~52% of the signal), leading to overestimation. CO(8-7) suffers from high variance due to faintness and interloper noise.
• Molecular Gas Density ($\rho_{H_2}$):
  • The reconstructed $\rho_{H_2}(z)$ matches the intrinsic simulation values within 7% relative difference for $J_{up}=2$ to $6$.
  • The 1$\sigma$ uncertainty on the relative difference is $<20\%$.
  • The contribution of starbursts to the total $\rho_{H_2}$ is negligible (<1%) because the MS population dominates the gas mass, despite SBs having higher specific SFRs.
• Effective Bias: The effective bias $b_{eff}$ is found to be independent of the CO transition ($J_{up}$), confirming that the clustering properties of the CO-emitting gas trace the underlying dark matter distribution similarly across the ladder.
• CONCERTO Feasibility:
  • Current Status: CONCERTO lacks the necessary sensitivity to detect the CO $\times$ galaxy cross-power spectrum at relevant scales ($k \le 0.35 \text{ Mpc}^{-1}$), even under ideal conditions.
  • Required Sensitivity: To achieve a Signal-to-Noise (S/N) of 3, CONCERTO would need an NEI roughly an order of magnitude lower than its current projected performance.
  • Mode Loss: The study highlights that spectral resolution limits ($\delta\nu = 1.5$ GHz) cause a loss of radial modes, significantly reducing the S/N. Fully exploiting 3D Fourier space (radial and transverse) is critical for noise-dominated LIM experiments.

5. Significance

This paper establishes cross-correlation with galaxy surveys as a robust, model-independent method to:

1. Decontaminate LIM signals from interlopers.
2. Measure the CO SLED evolution directly from the background, providing insights into the physical conditions (density, temperature) of the interstellar medium (ISM) in galaxies.
3. Constrain the cosmic molecular gas density ($\rho_{H_2}$) up to $z=3$ without relying on the detection of individual faint galaxies.

While current instruments like CONCERTO are not yet sensitive enough to execute this specific analysis, the methodology provides a clear roadmap for future high-sensitivity millimeter-wave LIM experiments (e.g., CCAT-prime, TIME, or future generations of CONCERTO). It suggests that combining LIM with deep spectroscopic surveys (like those from Euclid or the Vera C. Rubin Observatory) will be essential for unlocking the full potential of CO intensity mapping in understanding galaxy evolution and the cosmic gas cycle.