Probing Dark Energy Microphysics with kSZ Tomography

TL;DR

This work targets the dark-energy microphysics by probing perturbations through kSZ tomography in combination with galaxy clustering. Using a Fisher-mmatrix forecast for LSST and CMB-S4–like surveys, it quantifies how velocity–density cross-spectra $P_{gv}$ and velocity auto-spectra $P_{vv}$ add information beyond background probes, tightening constraints on $(w_0,w_a)$ and offering a different degeneracy structure than geometric measurements. The authors introduce a simple two-parameter model of dark-energy perturbations, $(\alpha,\ell_s)$, to assess detectability: for canonical $c_s=1$ the perturbation signal is sub-percent and horizon-limited, while smaller $c_s$ shifts the signal to observable scales. They conclude that near-term kSZ measurements will test consistency between background and perturbations, whereas future low-noise, high-resolution surveys could begin to reveal the microphysical properties of dark energy, provided systematic modeling of $P_{ge}$ and $b_v$ is controlled.

Abstract

The accelerated expansion of the Universe is well established by geometric probes, yet its physical origin remains poorly understood. Most constraints on dark energy arise from background observables -- supernovae, baryon acoustic oscillations, and the cosmic microwave background -- which mainly test the homogeneous expansion history. To move beyond this limitation, we examine how kinetic Sunyaev--Zel'dovich (kSZ) tomography, combined with galaxy clustering, can probe perturbative effects of dark energy and improve constraints on its background parameters. Using a Fisher-matrix analysis of the joint power spectra for LSST- and CMB-S4-like surveys, we quantify the additional information kSZ tomography contributes to dark-energy inference. Including kSZ data tightens constraints on $w_0$ by 15 % and on $w_a$ by 32 %, with parameter degeneracies distinct from those of geometric probes. We also assess the detectability of dark-energy perturbations through a two-parameter model, finding that for canonical sound speed ($c_s=1$) the effects are sub-percent and confined to horizon scales, while smaller sound speeds shift them to accessible $k$-ranges. Near-term kSZ measurements will primarily serve to test the consistency between background and perturbative signals, while future low-noise, high-resolution surveys may begin to uncover the microphysical properties of dark energy.

Figures (9)

• Figure 1: Evolution of quintessence perturbations $\delta \varphi$ for three representative scales. Modes quickly decay once they enter the sound horizon, while super-horizon modes remain frozen. The resulting transfer function is set by equation \ref{['eq:Decay']}.
• Figure 2: Scale-dependent modifications to the matter power spectrum induced by dark energy perturbations. We plot the ratio of the matter power spectrum in a $w_0w_a$ model to that of $\Lambda$CDM, normalized on small scales where dark energy perturbations are negligible. We plot this ratio for all values of $w_0w_a$ sampled from the DESI BAO, PantheonPlus SNe, and Planck CMB posterior. Best-fit values predict a $\sim 1\%$ effect on horizon scales. We used the PPF formalism as implemented in $\mathtt{camb}$Lewis_2000Fang_2008 to compute these and subsequent matter powerspectra.
• Figure 3: Constraints on $(w_0, w_a)$ from different background probes and their combinations (see text for details). For the BAO and SNe analyses, we impose a Gaussian BBN prior on the baryon density deside25 in order to break the degeneracy between baryonic and cold dark matter components. Here, and throughout, posteriors were sampled using the $\mathtt{cobaya}$Torrado_20212019ascl.soft10019T2021JCAP...05..057T and $\mathtt{pocomc}$karamanis2022acceleratingkaramanis2022pocomc python packages. The samples were analyzed using the $\mathtt{getdist}$Lewis_2025 package.
• Figure 4: Forecasted measurements of the galaxy–velocity cross-spectrum $P_{gv}$ based on LSST and CMB-S4 specifications, with error estimates from smith2018ksztomographybispectrum. Top: scale-independent modifications to the spectrum, illustrated using small-scale modes. Bottom: scale-dependent modifications, shown for large-scale modes, normalized on small scales. In green we also show how the SNR increases if the CMB survey parameters $(s_w, \theta_{\rm FWHM})$ are reduced by a factor of two relative to those described in Eq. \ref{['eq:CMBspec']}.
• Figure 5: Constraints on $(w_0, w_a)$ from a joint Fisher analysis of background (BAO, CMB and SNe) and kSZ ($P_{gg}$, $P_{gv}$ and $P_{vv}$) probes. The background posterior is taken from the Gaussian approximation to the MCMC posteriors derived in previous sections. Background and kSZ constraints are combined by adding the corresponding Fisher matrices. We show the full Fisher matrix, as well as constraints derived using more futuristic CMB specifications, in the appendix.