Lyman-alpha Intensity Mapping of the High Redshift Intergalactic Medium – L-INTENSE

While IM is conceptually quite simple, there are several practical difficulties. The major challenge is the separation of the desired emission line signal from fore- and background continuum contamination and from interlopers of other lines. Sometimes such nuisance signals can be subtracted from the data prior to the application of IM, but often this is not possible. Different strategies have been developed to deal with this problem. In the simplest case the contamination is contained in discrete sources which can then be masked out (e.g., Gong et al. 2014). A more powerful method consists of cross- correlating the data with a sample of “tracers” with previously known redshifts, since uncorrelated fore- or background variance increases the errors but does not bias the determination (Switzer et al. 2019). A disadvantage of the cross-correlation approach is that it is insensitive to any potential emission line signal uncorrelated with the tracer population; it therefore delivers only a lower limit. Autocorrelation of the signal, on the other hand, gives an upper limit (Switzer et al. 2013) because of possible contamination by interlopers. Together, cross- and autocorrelation provide a robust bracket to the desired quantity.
Most IM experiments have been developed as an efficient alternative to galaxy surveys, with the purpose of measuring the evolution of the galaxy population and constraining the large scale structure of the Universe (Kovetz et al. 2017). Such observations generally employ large fields of view (which at least in the optical means imaging) and correspondingly have low spatial and spectral resolution. In other cases the spectroscopic data are extremely sparsely sampled in the sky (e.g. in SDSS). In our project we will be following a very differently designed IM approach. We will apply the method directly to a 3-dimensional integral-field spectroscopic dataset, which at the same time provides the tracer samples for cross-correlations. While the field of view is small, the total volume sampled is still substantial by virtue of the large redshift range covered. Finally, the main motivation for employing the IM formalism is different – we want to use it because of its sensitivity to detect spatially extended emission which would be too faint to be observable directly.

(1) We process and perform optimal source extraction, redshifts measurements and emission and absorption lines measurements for the 3 MUSE datacubes (Mosaic, UDF-10 and MXDF). These datasets serve as input of the 3 following studies.
(2) We calculate the median Lya cross-correlation function at 3< z<4, providing the most robust measurement of the Lya surface brightness profiles out to approximately 470 kpc. We demonstrate that the diffuse Lya haloes (LAHs) are dominated by neighboring Lya emitters (LAEs) at hundreds of kpc
(3) From the galaxy center to approximately 70 kpc around the galaxy, we observe an average blueshift of the Lya line. This is strong evidence of large-scale gas inflows from the galaxies.
(4) Within 15 kpc, we observe the bi-polar metal-enriched galactic outflows traced by anisotropic MgII emission. We prove this MgII outflow is prevalent among the massive galaxy population .

None

- The MUSE Hubble Ultra Deep Field surveys: Data release II, R. Bacon, …, Y. Guo, et al, 2023, A&A 670, A4
- Observational Evidence of the Prevalence of Bipolar Galactic Outflows out to 10 kpc at z~1 for Massive Galaxies, Y. Guo, R.Bacon, et al., Nature under review
- MEGAFLOW IX. The impact of gas flows on the relations between the mass, star formation rate, and metallicity of galaxies, I. Langan, …, Y. Guo et al. 2023, MNRAS, 521, 546

We propose to perform the first measurement of Lyman-alpha (Lya) emission from the average intergalactic medium at redshifts between 3 and 6.5, a crucial experiment in observational cosmology that so far was impossible because of a lack of sensitivity. We will combine two new and unique datasets obtained by us with the integral-field spectrograph MUSE, constructed under our leadership, with the Intensity Mapping (IM) formalism (adapted to our specific needs) to maximise the sensitivity of the measurement. The observational data comprise the revolutionary "MUSE eXtremely Deep Field" (MXDF) with 155 hours of exposure in a single pointing, augmented by several additional MUSE deep fields of 25-30 hours exposure each. Building on our recent successful work with MUSE we expect to reach an unprecedented surface brightness limit (1sigma for a single Lya emission line) of <1 x 10^(-21) erg s^-1 cm^-2 arcsec^-2 in the mean intensity mapping signal, lower by a factor of several compared to previous work. In order to achieve this top-level goal we have to address several related, but to some extent independent problems, which will produce novel and interesting results by themselves: (1) We have to remove the contribution from lower redshift emission lines by cross-correlating with appropriate tracer samples directly constructed from our MUSE data. This will include the application of the IM method to other strong lines such as [OII] or H-beta. (2) We must account for the contribution of individually undetected, extremely faint Lya-emitting galaxies, which requires measuring the redshift evolution of the faint end of the Lya emitter luminosity function. (3) We have to quantify and subtract the Lya contributions of individually detected galaxies, i.e.\ their interstellar and in particular their circumgalactic Lya emission. We will incorporate new dedicated cosmological numerical simulations produced by collaborators into the analysis, to help validating the IM approach, to guide us disentangling the genuine intergalactic component from the total measured signal, and to interpret the results in terms of physical properties of the intergalactic medium.
Besides these main goals, the results of our projects will be relevant for several other scientific questions, such as implications for cosmic reionisation from the faint end of the LAE luminosity function, and the properties of Lya haloes around galaxies at z>3. We underline that our project differs conceptually in several points from most other ongoing or planned IM experiments: (i) We operate directly on 3D spectroscopic datacubes with full area coverage. (ii) While the field of view of MUSE is small, the total volume sampled is still substantial by virtue of the large redshift range covered. (iii) Most importantly, we explicitly aim at removing the dominant part of the signal that can be related to individual galaxies, isolating only the weak glow from the extremely tenuous matter far away from galaxies.