Pith. sign in

REVIEW 3 major objections 6 minor 34 references

Interloper lines in intensity maps are extra large-scale structure tracers, not only contaminants to clean away.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-31 05:56 UTC pith:AX3AYEVA

load-bearing objection Solid forecast Letter: interlopers as projected tracers plus a two-band BAO ladder are real; the keep-vs-clean Gp claim is softer than advertised because F_lost is never operationalized. the 3 major comments →

arxiv 2607.24917 v1 pith:AX3AYEVA submitted 2026-07-27 astro-ph.CO astro-ph.GAastro-ph.IM

Interlopers as Signal in Line Intensity Mapping

Anirban Roy This is my paper
classification astro-ph.CO astro-ph.GAastro-ph.IM
keywords line intensity mappinginterlopersbaryon acoustic oscillationsFisher forecastsSPHERExFYSTlarge-scale structureAlcock-Paczynski
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Line intensity mapping stacks the glow of many unresolved galaxies into 3D maps of the cosmos, but each observed frequency mixes several atomic and molecular lines from different redshifts. Those extra lines, called interlopers, are usually stripped out before cosmology is measured. This paper argues that stripping is not always best: once the lines and their redshifts sit in the model, each interloper is itself a map of structure at another epoch, only warped into the target line’s coordinates. The choice—clean versus jointly model—is cast as a Fisher decision that demands both a precision gain and a parameter bias below a set tolerance. Forecasts for SPHEREx-like Hα and FYST-like [C II] surveys show that matter density and BAO distances stay comparatively safe under astrophysical calibration error, while clustering amplitude and dark-energy parameters are more fragile. Modeling the interlopers turns two observed bands into a transverse BAO distance ladder spanning roughly 0.7 to 5.8 in redshift, so the same emission usually treated as contamination becomes calibrated cosmological signal.

Core claim

Once possible emitting lines and their redshifts are included in the model, interlopers become additional tracers of large-scale structure at other epochs, projected into the target line’s coordinate system. Under a Fisher decision rule that requires both improved precision and bias below a chosen tolerance, joint modeling is favored for Ω_m h² and BAO distance measures in the SPHEREx Hα and FYST [C II] forecasts, turning two observed bands into a transverse BAO ladder over 0.7 ≲ z ≲ 5.8. Interlopers are therefore a calibrated component of the cosmological signal model, not only contaminants to remove.

What carries the argument

The Fisher decision criterion: joint modeling is preferred when the precision gain G_p = σ_clean / σ_alt exceeds 1 and the induced bias stays below B_max (taken as 0.5). It is applied to the projected multi-line power spectrum that sums each line after Alcock–Paczynski warping into the target frame.

Load-bearing premise

The analysis assumes each rest-frame line’s brightness, bias, and shot noise can be captured by a few global calibration factors whose redshift evolution is fixed by one chosen luminosity–star-formation family, with external priors that may be wrong.

What would settle it

Apply the same clean-versus-joint comparison on real or end-to-end simulated SPHEREx or FYST maps: if joint modeling with realistic line-calibration priors fails to shrink errors on Ω_m h² or BAO dilation while keeping shifts under half a sigma, or if the multi-redshift BAO ladder does not appear, the central claim fails.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Future LIM pipelines should treat interloper handling as a per-parameter estimator choice, not a fixed cleaning step.
  • Two observed bands can supply transverse BAO anchors from z ≈ 0.7 to z ≈ 5.8 when lines are modeled rather than removed.
  • Ω_m h² and BAO dilation remain the safest cosmological targets under line-calibration error; σ_8, w_0, and w_a need tighter or better-centered astrophysical priors.
  • Gains are largest where the target line is faintest, so interloper-dominated bands become especially valuable once modeled.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Survey design could deliberately choose bands where bright interloper ladders supply the bulk of the distance information rather than only optimizing the nominal target.
  • The same decision rule could rank hybrid strategies—partial cleaning plus residual modeling, or external-tracer cross-spectra—against pure joint modeling before data freeze.
  • If line evolution is later shown to be more complex than a single global family, the bias-safe status of BAO may still hold while growth and dark-energy gains shrink first.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. This Letter argues that spectral-line interlopers in line intensity mapping should not be removed by default but treated as additional large-scale-structure tracers projected into the target-line coordinate system. The author formulates interloper treatment as a Fisher decision problem (Eqs. 1–4): joint modeling is favored over cleaning when the effective interloper information plus the target information lost to cleaning is positive, and when the induced parameter bias stays below a tolerance Bmax=0.5σ. Using a standard projected multi-line power spectrum (Eq. 5), a Gaussian Fisher matrix (Eq. 6), and the linearized Fisher-bias formula (Eq. 7), forecasts for SPHEREx Hα (with [O III], Hβ, [O II] interlopers) and FYST [C II] (with a CO ladder) show that Ω_m h² and BAO dilation parameters gain precision from joint modeling while remaining bias-safe under amplitude miscalibration, whereas σ8 and (w0, wa) are prior-sensitive. A concrete geometric realization is a transverse BAO distance ladder over 0.7≲z≲5.6–5.8 from two observed bands, with anchors fixed by rest wavelengths.

Significance. If the quantitative conclusions hold, the paper usefully reframes interloper treatment from a fixed preprocessing step into an estimator choice, and gives a cheap Fisher-level decision rule that survey teams can run before committing to a cleaning pipeline. Strengths worth crediting: the projected multi-line model (Eq. 5) with AP Jacobian and Kaiser RSD is standard and internally consistent; the decision criterion couples precision to an explicit, checkable bias test (Eqs. 4, 7) rather than asserting robustness; the BAO-anchor picture (Fig. 4) is physically transparent and its redshift locations are set by atomic/molecular rest wavelengths, making the ladder geometry falsifiable and model-independent; and stability to finite-difference steps, (k,μ) binning, k_max, noise, beam, and priors is reported. The timing is good, with SPHEREx taking data and FYST under construction. The work is forecast-only, and its central quantitative claim rests on an undefined cleaned baseline (below), which currently limits its force.

major comments (3)
  1. [Eqs. (1)–(3), Fig. 3] The decision Gp>1 is, by Eq. (3), the statement F_eff_interloper + F_lost > 0, yet F_lost is never given an operational definition. No cleaning prescription behind the 'cleaned target-only analysis' entering σ_clean in Fig. 3 is specified anywhere in the text or End Matter — no masking scheme, deprojection mode count, or external-tracer subtraction. F_lost varies by orders of magnitude across real cleaning strategies: aggressive masking of contaminated channels makes modeling win almost trivially, while the partial-cleaning/deprojection options the Letter itself lists can leave F_lost small. As written, the Gp values in Fig. 3 are not reproducible and the Gp=1 boundary could be crossed from the other side for an efficient cleaner. Please specify a concrete cleaned baseline (with its mode/volume loss) and show how the decision boundary shifts with the assumed F_lost.
  2. [Eq. (4) and Fig. 3] The decision rule is applied asymmetrically. The joint-modeling branch is subjected to the Fisher-bias test of Eq. (7) (offset line amplitudes), but the cleaned branch enters as pure information loss with zero residual bias: σ_clean contains no Δp term from residual interloper leakage, which any real cleaning (masking, deprojection, external-tracer subtraction) leaves at some level. Since Eq. (4) requires both Gp>1 and |Δp|/σ_alt<Bmax, the comparison should either include a residual-bias estimate for the cleaned branch or justify explicitly why residual leakage bias is negligible for the specified cleaning method. This is load-bearing because the flagship conclusion — modeling favored for Ω_m h² and BAO dilation — is exactly where the two branches are compared.
  3. [Fig. 3; End Matter Eqs. (13)–(16)] The robustness claim ('Ω_m h² and BAO dilation are comparatively robust to astrophysical calibration errors') is tested only against coherent global amplitude offsets — 'the true line amplitude is offset from the assumed calibration by half of that prior width.' But the forecast assumes a single luminosity–SFR family supplies the full redshift evolution of each rest-frame line across bins (global ln I_L, ln b_L, ln P_shot,L per line). A redshift-dependent miscalibration — wrong SFR(M,z) evolution in Eq. (16), evolving dust attenuation, or evolving scatter S_sc in Eq. (15) — is a different and plausible failure mode that would correlate across bins and could mimic distance/growth evolution. Please add at least one redshift-dependent mismatch test (e.g., perturbed ϵ(z) or M_p(z) slope) to the bias analysis, or narrow the stated robustness claim to amplitude errors only.
minor comments (6)
  1. [Modeling the signal / End Matter] Parameter-set inconsistency: the main text constrains 'four cosmological parameters (Ω_m h², σ8, w0, wa)' while the End Matter forms finite differences in (Ω_m, σ8, h, w0, wa) — five. Please state the full varied parameter vector and which parameters (Ω_b h², n_s, τ) are held fixed, and whether any external cosmological priors are applied. This materially affects the Fig. 2 headline factors (≈4 and ≈190 on σ(σ8), up to ≈50 on w0, wa).
  2. [End Matter, Experiment specifications] Survey-range inconsistency: the main text describes the flagship SPHEREx Hα survey as spanning 0.7<z<1.6 split into five redshift bins, while the End Matter targets Hα at z=1 in 'a patch spanning 0.9<z<1.1.' Please reconcile; the binning enters V_z in Eq. (6) directly.
  3. [Abstract / Conclusions / Fig. 4] The ladder's upper edge is quoted as z≈5.6 in the abstract and conclusions, while the FYST channel places [C II] at z=5.79 (End Matter) and Fig. 1(b) uses z=5.8. Please use a consistent value (5.8) or clarify which anchor sets the edge.
  4. [Fig. 2 discussion] The factors ≈190 (σ8) and ≈50 (w0, wa) are relative to a target-only fit with unmodeled interlopers in the covariance, where the faint [C II] target is nearly unconstrained; Fig. 3 instead uses the cleaned baseline. Please flag explicitly that these two baselines differ so readers do not compare the numbers directly.
  5. [Eq. (5), k_max] k_max = 0.25 Mpc⁻¹ is used with linear Kaiser RSD and (presumably linear) CLASS P_m; at z≈1 this is mildly nonlinear. State whether any nonlinear prescription is included, or note the optimism; the reported stability to k_max variation partly addresses this but should be summarized quantitatively.
  6. [Fig. 3] The caption should describe the color scale for B=|Δp|/σ and its cap at Bmax=0.5 explicitly, and state the Bmax=0.25 robustness check result alongside the figure rather than only in the text.

Circularity Check

0 steps flagged

No significant circularity: Fisher forecasts compare joint vs cleaned estimators using external line models and CLASS cosmology; results are not forced by definition or self-citation.

full rationale

The paper’s load-bearing chain is a standard Fisher forecast, not a closed derivation. Observed power (Eq. 5) is a projected sum of target plus interloper lines with AP factors and Kaiser RSD; Pm(k,z) comes from CLASS; line moments I_L, b_L, P_shot,L are built from external halo-SFR and empirical luminosity relations (UniverseMachine, Gong optical scalings, De Looze [C II], Li/Mashian CO). Cosmological parameters are constrained after marginalizing nuisance amplitudes with external priors; bias is tested by offsetting true amplitudes from the model (Eq. 7). The decision rule Gp>1 and |Δp|/σ<Bmax compares two estimators on that model—it does not define cosmology as the fitted line amplitudes, nor does it rename a fit as a prediction. Self-citations (Roy & Battaglia; Roy et al. line/dust papers) supply methodology and related context but are not uniqueness theorems that force the central claim. Weaknesses such as an idealized F_lost or asymmetric bias testing of cleaning vs modeling are methodological/correctness issues, not circular reductions of outputs to inputs. The BAO ladder geometry follows from known rest wavelengths and AP projection, which is independent content.

Axiom & Free-Parameter Ledger

6 free parameters · 7 axioms · 0 invented entities

The claim rests on standard LIM projection and Fisher forecasting plus adopted empirical line–SFR families and survey specs. No new physical entities. Load-bearing modeling choices are the shared global line nuisances, external prior widths, linear Kaiser+Gaussian covariance, and the hand-set bias tolerance Bmax. Free parameters are analysis knobs and adopted normalizations, not cosmology fitted to new data.

free parameters (6)
  • Bmax (bias tolerance) = 0.5 (reference); 0.25 checked
    Decision threshold |Δp|/σ_alt < Bmax; paper adopts 0.5 (checks 0.25). Directly controls which parameters ‘favor modeling.’
  • External prior widths on line astrophysics = 5%–100% scan; 5% in Fig. 2; 20% for BAO ladder
    Gaussian priors on (ln I_L, ln b_L, ln P_shot,L) scanned from 5% to 100%; set Gp and bias colors in Fig. 3 and BAO errors at 20%.
  • k_max and (k,μ) binning = k_max=0.25 Mpc^-1 (SPHEREx flagship)
    Fisher information integral cutoff; SPHEREx example uses k_max=0.25 Mpc^-1. Changes total information though paper claims decision stability.
  • Optical and sub-mm luminosity–SFR normalizations = Literature values as in End Matter Eqs. 17–24
    Adopted coefficients (e.g. L_Hα=3.31e7 SFR, De Looze [C II], α_CO=1.37, β_CO=-1.74, r_J1 SLED) fix fiducial I_L, b_L, P_shot relative weights of target vs interlopers.
  • UniverseMachine-like SFR double power-law parameters = As written in End Matter Eq. 16
    Mp(z)=3e11[(1+z)/3]^-0.4 M_sun, ε(z)=2[(1+z)/3]^1.2 M_sun/yr drive redshift evolution of all lines under the global-calibration assumption.
  • Instrument noise powers and beams = SPHEREx PN~3.4e2 or 6.5; FYST N_white=4.9e9 Mpc^3 Jy^2 sr^-2
    SPHEREx PN and FYST N_white set S/N and whether maps are target- or interloper-dominated.
axioms (7)
  • domain assumption Observed LIM power is a projected sum over lines with AP Jacobian (q_perp, q_parallel) and linear Kaiser RSD, plus white noise and Poisson shot noise (Eq. 5).
    Standard in interloper LIM literature; neglects nonlinear RSD, fingers-of-god, and mode coupling that real analyses face.
  • domain assumption Gaussian Fisher matrix on P_obs(k,μ) with independent (z,k,μ) bins gives inverse marginalized parameter variances (Eq. 6).
    Forecast standard; non-Gaussian covariance and survey window are omitted.
  • domain assumption Linearized Fisher bias from ΔP_obs = P_true - P_model predicts parameter shifts under miscentered line calibration (Eq. 7).
    Valid for small model error; large astrophysical mismodeling can violate linearity.
  • ad hoc to paper Each rest-frame line’s redshift evolution is fixed by one luminosity–SFR family; only global amplitude/bias/shot factors are free per line.
    Stated explicitly: ‘the forecast therefore assumes that the chosen luminosity-SFR family supplies the redshift evolution.’ Central to joint-model information and bias tests.
  • domain assumption Sheth–Tormen mass function and bias plus halo luminosity moments define I_L, b_L, P_shot,L.
    End Matter; common halo-model LIM practice.
  • ad hoc to paper Cleaning yields F_clean = F_target - F_lost with F_lost≥0, while joint modeling yields F_target + F_eff_interloper after nuisance marginalization.
    Decision-problem framing (Eqs. 1–3); quantitative F_lost depends on an idealized cleaned baseline not matched to a named pipeline.
  • domain assumption CLASS linear/matter power spectrum and flat w0waCDM parameter set {Ω_m h², σ8, w0, wa} capture the cosmological dependence.
    Standard Boltzmann + dark-energy parametrization for the forecasts.

pith-pipeline@v1.2.0-grok45-kimik3 · 16766 in / 4627 out tokens · 96020 ms · 2026-07-31T05:56:37.304805+00:00 · methodology

0 comments
read the original abstract

Line intensity mapping measures the combined emission of atomic and molecular lines from unresolved galaxies, offering a way to map cosmic structure across enormous volumes. A single observed frequency, however, contains emission from several spectral lines at different redshifts. These interlopers are usually removed before cosmological inference. We show that this removal is not always optimal. Once the possible emitting lines and their redshifts are included in the model, interlopers become additional tracers of large-scale structure at other epochs, projected into the coordinate system of the target line. We formulate interloper treatment as a Fisher decision problem, requiring both improved precision and parameter bias below a chosen tolerance. In SPHEREx forecasts for H$\alpha$ with [O III], H$\beta$, and [O II] interlopers, and in FYST forecasts for [C II] with CO interlopers, $\Omega_m h^2$ and baryon acoustic oscillation (BAO) distance measurements are comparatively robust to astrophysical calibration errors, whereas the clustering amplitude $\sigma_8$ and the dark energy parameters $w_0$ and $w_a$ are more sensitive to the adopted line model and priors. The BAO distance measurement gives a simple physical picture: modeling interlopers turns two observed bands into a transverse BAO distance ladder over $0.7\lesssim z\lesssim5.8$. Interlopers are therefore not only contaminants to remove, but a calibrated component of the cosmological signal model for future LIM surveys.

Figures

Figures reproduced from arXiv: 2607.24917 by Anirban Roy.

Figure 1
Figure 1. Figure 1: FIG. 1. Projected monopole [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Marginalized 68% and 95% constraints for one FIG. 2. Marginalized 68% and 95% constraints for one [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Transverse BAO distance [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

34 extracted references · 30 linked inside Pith

  1. [1]

    E. D. Kovetzet al., arXiv e-prints (2017), arXiv:1709.09066 [astro-ph.CO]

    Pith/arXiv arXiv 2017

  2. [2]

    Fonseca, M

    J. Fonseca, M. B. Silva, M. G. Santos, and A. Cooray, Mon. Not. R. Astron. Soc.464, 1948 (2017), arXiv:1607.05288

    Pith/arXiv arXiv 1948

  3. [3]

    Y. Gong, X. Chen, and A. Cooray, Astrophys. J.894, 152 (2020), arXiv:2001.10792

    Pith/arXiv arXiv 2020

  4. [4]

    J. L. Bernal and A. Baleato Lizancos, Phys. Rev. D111, 043539 (2025), arXiv:2406.12979 [astro-ph.CO]

    Pith/arXiv arXiv 2025

  5. [5]

    Karoumpis, B

    C. Karoumpis, B. Magnelli, E. Romano-D ´ ıaz, M. Haslbauer, and F. Bertoldi, Astron. Astrophys. 691, A262 (2024), arXiv:2410.17330

    Pith/arXiv arXiv 2024

  6. [6]

    Roy and N

    A. Roy and N. Battaglia, Astrophys. J.969, 2 (2024), arXiv:2312.08471 [astro-ph.CO]

    Pith/arXiv arXiv 2024

  7. [7]

    Aghanimet al.(Planck), Astron

    N. Aghanimet al.(Planck), Astron. Astrophys.641, A6 (2020), [Erratum: Astron.Astrophys. 652, C4 (2021)], arXiv:1807.06209 [astro-ph.CO]

    Pith/arXiv arXiv 2020

  8. [8]

    Chevallier and D

    M. Chevallier and D. Polarski, International Journal of Modern Physics D10, 213 (2001), arXiv:gr-qc/0009008 [gr-qc]

    Pith/arXiv arXiv 2001

  9. [9]

    E. V. Linder, Phys. Rev. Lett.90, 091301 (2003), arXiv:astro-ph/0208512 [astro-ph]

    Pith/arXiv arXiv 2003

  10. [10]

    A. G. Adameet al.(DESI), JCAP07, 028, arXiv:2411.12022 [astro-ph.CO]

    Pith/arXiv arXiv

  11. [11]

    Lidz and J

    A. Lidz and J. Taylor, Astrophysical Journal825, 143 (2016), arXiv:1604.05737 [astro-ph.CO]

    Pith/arXiv arXiv 2016

  12. [12]

    W. E. Ballinger, J. A. Peacock, and A. F. Heavens, Monthly Notices of the Royal Astronomical Society282, 877 (1996), arXiv:astro-ph/9605017

    Pith/arXiv arXiv 1996

  13. [13]

    Seo and D

    H.-J. Seo and D. J. Eisenstein, Astrophysical Journal 665, 14 (2007), arXiv:astro-ph/0701079

    Pith/arXiv arXiv 2007

  14. [14]

    Kaiser, Monthly Notices of the Royal Astronomical Society227, 1 (1987)

    N. Kaiser, Monthly Notices of the Royal Astronomical Society227, 1 (1987)

    1987

  15. [15]

    D. Blas, J. Lesgourgues, and T. Tram, Journal of Cosmology and Astroparticle Physics2011(7), 034, arXiv:1104.2933 [astro-ph.CO]

    Pith/arXiv arXiv

  16. [16]

    Aravenaet al.(CCAT-Prime), Astrophys

    M. Aravenaet al.(CCAT-Prime), Astrophys. J. Suppl. 264, 7 (2023), arXiv:2107.10364 [astro-ph.CO]

    Pith/arXiv arXiv 2023

  17. [17]

    J. J. Bocket al., Astrophys. J.999, 139 (2026), arXiv:2511.02985 [astro-ph.IM]

    arXiv 2026

  18. [18]

    Tegmark, Phys

    M. Tegmark, Phys. Rev. Lett.79, 3806 (1997), arXiv:astro-ph/9706198

    Pith/arXiv arXiv 1997

  19. [19]

    Seo and D

    H.-J. Seo and D. J. Eisenstein, Astrophys. J.598, 720 (2003), arXiv:astro-ph/0307460

    Pith/arXiv arXiv 2003

  20. [20]

    Amara and A

    A. Amara and A. Refregier, Monthly Notices of the Royal Astronomical Society391, 228 (2008), arXiv:0710.5171

    Pith/arXiv arXiv 2008

  21. [21]

    L. Knox, R. Scoccimarro, and S. Dodelson, Physical Re- view Letters81, 2004 (1998), arXiv:astro-ph/9805012

    Pith/arXiv arXiv 2004

  22. [22]

    Huterer and M

    D. Huterer and M. Takada, Astroparticle Physics23, 369 (2005), arXiv:astro-ph/0412142

    Pith/arXiv arXiv 2005

  23. [23]

    A. Roy, D. Valent ´ ın-Mart ´ ınez, K. Wang, N. Battaglia, and A. van Engelen, Astrophys. J.957, 87 (2023), arXiv:2304.06748 [astro-ph.GA]

    Pith/arXiv arXiv 2023

  24. [24]

    Y. Gong, A. Cooray, M. B. Silva, M. Zemcov, C. Feng, M. G. Santos, O. Dore, and X. Chen, Astrophys. J.835, 273 (2017), arXiv:1610.09060

    Pith/arXiv arXiv 2017

  25. [25]

    B. P. Crill, M. Werner, R. Akeson,et al., inSpace Telescopes and Instrumentation 2020: Optical, Infrared, and Millimeter Wave, Proc. SPIE, Vol. 11443 (2020) p. 114430I

    2020

  26. [26]

    A. Roy, R. S. Somerville, and A. R. Pullen, arXiv e-prints (2026), arXiv:2607.19504 [astro-ph.GA]

    Pith/arXiv arXiv 2026

  27. [27]

    R. K. Sheth and G. Tormen, Mon. Not. R. Astron. Soc. 308, 119 (1999), arXiv:astro-ph/9901122

    Pith/arXiv arXiv 1999

  28. [28]

    P. S. Behroozi, R. H. Wechsler, and C. Conroy, Astro- phys. J.770, 57 (2013), arXiv:1207.6105

    Pith/arXiv arXiv 2013

  29. [29]

    Behroozi, R

    P. Behroozi, R. H. Wechsler, A. P. Hearin, and C. Con- roy, Mon. Not. R. Astron. Soc.488, 3143 (2019), arXiv:1806.07893

    Pith/arXiv arXiv 2019

  30. [30]

    Kennicutt, Robert C., Annu

    J. Kennicutt, Robert C., Annu. Rev. Astron. Astrophys. 36, 189 (1998), arXiv:astro-ph/9807187

    Pith/arXiv arXiv 1998

  31. [31]

    De Loozeet al., Astron

    I. De Loozeet al., Astron. Astrophys.568, A62 (2014), arXiv:1402.4075

    Pith/arXiv arXiv 2014

  32. [32]

    T. Y. Li, R. H. Wechsler, K. Devaraj, and S. E. Church, Astrophys. J.817, 169 (2016), arXiv:1503.08833

    Pith/arXiv arXiv 2016

  33. [33]

    for the multi-JCO ladder [31–33]. Experiment specifications Instrumental sensitivity enters the forecast through a single number, the white noise power spectrum,P N = σ2 voxVvox, whereV vox is the comoving voxel volume de- termined by the beam and the spectral channel size. SPHEREx is a wide field spectrophotometer whose 6.2 ′′ pixels andR= 41 spectral ch...

  34. [34]

    Mashian, A

    N. Mashian, A. Sternberg, and A. Loeb, J. Cosmol. As- tropart. Phys.2015(11), 028, arXiv:1507.02686

    Pith/arXiv arXiv 2015