The Simons Observatory: forecasted constraints on primordial gravitational waves with the expanded array of Small Aperture Telescopes

TL;DR

This paper updates forecasts for Simons Observatory primordial B-mode constraints by detailing an expanded small-aperture telescope program (three additional SATs, including two MF and one LF) and a ten-year survey through 2035. Using Pipeline A multi-frequency forward modeling, Gaussian likelihoods, and simulations, the authors quantify how delensing with the SO LAT and foreground treatment affect the tensor-to-scalar ratio limit $r$. Under conservative assumptions with 50% delensing and decorrelation, they forecast $\sigma_r \,=\,1.4\times10^{-3}$ after 10 years; with optimistic noise and 70% delensing (no decorrelation), $\sigma_r$ improves to $7\times10^{-4}$. The expanded program promises roughly a 2.6× improvement over the original plan, with final results strongly dependent on foreground decorrelation and instrument noise, motivating careful modeling of systematics and delensing potential. The work also outlines commissioning progress and the timeline for the new instruments beginning in 2027, highlighting the enhanced capability to probe inflationary models through $r$ and related parameters.

Abstract

We present updated forecasts for the scientific performance of the degree-scale (0.5 deg FWHM at 93 GHz), deep-field survey to be conducted by the Simons Observatory (SO). By 2027, the SO Small Aperture Telescope (SAT) complement will be doubled from three to six telescopes, including a doubling of the detector count in the 93 GHz and 145 GHz channels to 48,160 detectors. Combined with a planned extension of the survey duration to 2035, this expansion will significantly enhance SO's search for a $B$-mode signal in the polarisation of the cosmic microwave background, a potential signature of gravitational waves produced in the very early Universe. Assuming a $1/f$ noise model with knee multipole $\ell_{\rm knee} = 50$ and a moderately complex model for Galactic foregrounds, we forecast a $1σ$ (or 68% confidence level) constraint on the tensor-to-scalar ratio $r$ of $σ_r = 1.2\times10^{-3}$, assuming no primordial $B$-modes are present. This forecast assumes that 70% of the $B$-mode lensing signal can ultimately be removed using high resolution observations from the SO Large Aperture Telescope (LAT) and overlapping large-scale structure surveys. For more optimistic assumptions regarding foregrounds and noise, and assuming the same level of delensing, this forecast constraint improves to $σ_r = 7\times10^{-4}$. These forecasts represent a major improvement in SO's constraining power, being a factor of around 2.5 times better than what could be achieved with the originally planned campaign, which assumed the existing three SATs would conduct a five-year survey.

Figures (4)

• Figure 1: Timeline of the Simons Observatory expanded SATs programme. Two additional "mid-frequency" SATs, operating at 93/145 GHz, are due to begin operations in 2027. A sixth "low-frequency" SAT, operating at 27/39 GHz is due to commence operations on a similar timescale. The additional telescopes, combined with an extension of the SO SAT survey to the mid-2030s, will significantly enhance the scientific reach of the observatory. This is demonstrated here by quoting forecasts for the errors achieved on the tensor-to-scalar ratio, $\sigma_r$, by 2028 and by 2035. For clarity, forecasted errors are quoted only for the "pessimistic" $1/f$ noise model, assuming a moderately complex model for astrophysical foregrounds (i.e. allowing for foreground frequency decorrelation -- see text in Section \ref{['sec:science']} for details). See Figure \ref{['fig:sr']} and Table \ref{['tab:sr']} for the detailed forecasting results, covering a wider range of assumptions for the levels of noise, foreground complexity and delensing achieved.
• Figure 2: The $B$-mode power spectrum and uncertainties achievable by SO, accounting for the expansions described in Section \ref{['sec:infrastructure']}. Bandpowers of width $\Delta\ell=10$ are shown after 10 years of observations assuming no delensing (orange bands) and 70% delensing (blue bands), and assuming the "pessimistic" $1/f$ noise model (see Section \ref{['sec:forecast_assumptions']} for details). The solid line shows the lensing $B$-mode power spectrum in the absence of tensor modes ($r=0$). The signal for $r=0.01$ is shown in dashed black. Current measurements from BICEP/ Keck2021PhRvL.127o1301A, SPT-3G 2025arXiv250502827Z, and Polarbear2022ApJ...931..101A are also shown for reference.
• Figure 3: The forecasted 68% constraints on the tensor-to-scalar ratio $r$, achievable by SO, as a function of time, accounting for the expansions described in Section \ref{['sec:infrastructure']}. "Y$N$" denotes the $N$-th year after the start of observations. Results are shown assuming 50% and 70% delensing (blue and red, respectively), for different assumptions regarding $1/f$ noise and foreground decorrelation. Each band covers the range of $\sigma_r$ between the worst- and best-case scenarios (the "pessimistic" $1/f$ noise model with foreground frequency decorrelation as described by the moment expansions vs. the "optimistic" $1/f$ noise model and no decorrelation). The constraints that the nominal SO configuration would have been able to attain, with the same foreground and noise assumptions, and $A_{\rm lens}=0.5$, are shown as the gray hashed band.
• Figure 4: The forecasted 95% constraints on the tensor-to-scalar ratio $r$ and the scalar spectral index $n_s$ that could be achieved by SO. Results are shown for the full SO SAT configuration, including the enhancements detailed in Section \ref{['sec:infrastructure']}, assuming 10 years of observation and 70% delensing. These forecasts are presented for two cases: (i) assuming the "pessimistic" $1/f$ noise model and allowing for foreground frequency decorrelation (yellow), and (ii) assuming the "optimistic" $1/f$ noise model and no frequency decorrelation (purple). The black line shows the prediction from $R^2$ inflation (Starobinsky model) for a total number of $e$-folds in the range $45<N_*<75$, with the range limits marked by the small and large black circles. We have assumed the true value of $r$ to correspond to the Starobinsky prediction given the best-fit value of $n_s$ preferred by Planckplanck_cosmo:2018. For comparison, the red contours show the 95% constraints achievable by the nominal SO SAT configuration (assuming a 5-year survey and 50% delensing, as well as optimistic $1/f$ noise and no frequency decorrelation), while the green contours show the current 68% and 95% constraints from BICEP/ Keck, Planck, and WMAP data 2021PhRvL.127o1301A.