Large-Scale Distribution of Total Mass versus Luminous Matter from Baryon Acoustic Oscillations: First Search in the Sloan Digital Sky Survey III Baryon Oscillation Spectroscopic Survey Data Release 10

Baryon acoustic oscillations in the early Universe are predicted to leave an as yet undetected signature on the relative clustering of total mass versus luminous matter. A detection of this effect would provide an important confirmation of the standard cosmological paradigm and constrain alternatives to dark matter as well as nonstandard fluctuations such as compensated isocurvature perturbations (CIPs). We conduct the first observational search for this effect, by comparing the number-weighted and luminosity-weighted correlation functions, using the SDSS-III BOSS Data Release 10 CMASS sample. When including CIPs in our model, we formally obtain evidence at $3.2\sigma$ of the relative clustering signature and a limit that matches the existing upper limits on the amplitude of CIPs. However, various tests suggest that these results are not yet robust, perhaps due to systematic biases in the data. The method developed in this Letter used with more accurate future data such as that from DESI, is likely to confirm or disprove our preliminary evidence.

Figure 1

Our measurement of ${\xi }_{\mathrm{L}}$ (blue) and ${\xi }_{\mathrm{n}}$ (red) [times $s^2$], using 31 radial bins, and our best-fit maximum likelihood model (allowing all parameters to be nonzero). The best fit corresponds to ${\chi }^2/\text{d.o.f}\mathrm{.}=2.51$, where d.o.f. is the number of degrees of freedom in the fit. This high value of ${\chi }^2/\text{d.o.f.}$ is partially due to the highly correlated errors among the various binned measurements and perhaps systematic errors (it drops to $\sim 1.5$ when using 21 bins), which also make the fits difficult to judge visually.

Figure 2

Our measurement of the difference ${\xi }_{\mathrm{L}}-{\xi }_{\mathrm{n}}$ (times $s^2$), using 31 radial bins, and the same quantity in our best-fit model. The red line corresponds to our full model, the blue line corresponds to a model with $B_{\mathrm{CIP}}=0$, and the green line corresponds to a model with $B_{\mathrm{CIP}}=B_{\mathrm{L},\mathrm{\Delta }}=0$.