Limits on isocurvature fluctuations from Boomerang and MAXIMA

TL;DR

Boomerang and Maxima-1 data constrain isocurvature fluctuations in a flat universe by comparing adiabatic and isocurvature initial conditions with uncorrelated spectra. The authors model the perturbations with independent spectral indices $n_{\rm ad}$ and $n_{\rm iso}$ and parameterize the isocurvature fraction by $\alpha$, computing CMB spectra with $\text{CMBFAST}$. Pure isocurvature models are decisively ruled out, while mixed models with a small $\alpha$ remain consistent, yielding $\alpha \leq 0.63$ at low multipoles and $\alpha_{200} \leq 0.13$ near the first peak. Calibration uncertainties and potential Planck polarization improvements are discussed, highlighting the enduring relevance of CMB temperature data for probing early-universe physics and particle-physics motivated isocurvature scenarios.

Abstract

We present the constraints on isocurvature fluctuations for a flat universe implied by the Boomerang and Maxima-1 data on the anisotropy of the cosmic microwave background. Because the new data defines the shape of the angular power spectrum in the region of the first acoustic peaks much more clearly than earlier data, even a tilted pure isocurvature model is now ruled out. However, a mixed model with a sizable isocurvature contribution remains allowed. We consider primordial fluctuations with different spectral indices for the adiabatic and isocurvature perturbations, and find that the 95% C.L. upper limit to the isocurvature contribution to the low multipoles is $α\leq 0.63$. The upper limit to the contribution in the $ł\sim 200$ region is $α_{200} \leq 0.13$.

Figures (6)

• Figure 1: The 68% (white), 95% (light gray), and 99.7% (medium gray) confidence level regions ($\Delta \chi^2 = 2.3$, 6.2, and 11.8) on the ($n_{\rm iso}$,$n_{\rm ad}$)-plane, and the best-fit values of $\alpha$ for each ($n_{\rm iso}$,$n_{\rm ad}$). The best-fit model (model 3) is marked with an asterisk ($\ast$).
• Figure 2: The 68% confidence level region ($\Delta \chi^2 \leq 3.5$) in the ($n_{\rm iso}$,$n_{\rm ad}$,$\alpha$)-space represented by contours of maximum $\alpha$. We do not show minimum $\alpha$, except that we show the region (between the dotted lines) where the minimum is $\alpha = 0$.
• Figure 3: Same as Fig. \ref{['fig:2']}, but for $\alpha_{200}$.
• Figure 4: The data points of COBE ($\diamond$), Boomerang ($\bullet$), and Maxima-1 ($\circ$), and the angular power spectra of five models: (a) Our best-fit mixed (adiabatic+isocurvature) model (model 3, the solid line with the maximum at $l \sim 200$). (b) The best fit adiabatic model (model 1, dashed). (c) The best-fit isocurvature model (model 2, the solid line with the maximum at $l \sim 300$). (d) A mixed model with the largest ($\Delta\chi^2 = 4$) allowed $\alpha = 0.63$ (model 5, dot-dashed). (e) A mixed model with the largest allowed $\alpha_{200} = 0.13$ (model 6, dotted).
• Figure 5: The angular power spectrum, $C_l$, and the adiabatic (dashed) and isocurvature (dot-dashed) contributions to it, $C_l^{\rm ad}$ and $C_l^{\rm iso}$, for model 5 with $(n_{\rm ad},n_{\rm iso}) = (1.20,0.90)$ and $\alpha = 0.63$.