TL;DR
This paper uses WMAP data to constrain the fine-structure constant ($\alpha$) at the epoch of cosmic decoupling. They find that $\alpha$ has not varied significantly since then, setting new upper bounds on its variation. The study also highlights how future polarization data from Planck will dramatically improve these constraints, particularly when combined with information about the reionization epoch.
What Problem Does This Paper Address?
The fundamental constants of nature, like the fine-structure constant ($\alpha$), are assumed to be constant throughout cosmic history. However, some theoretical frameworks, particularly those involving extra dimensions or modified gravity, predict that these constants might vary over time. Evidence for such variations has been hinted at in quasar absorption systems at redshifts of around 2-3. This paper aims to test these theories by looking for variations in $\alpha$ at a much earlier epoch: the time of cosmic decoupling (redshift $z \sim 1100$), when the cosmic microwave background (CMB) radiation was released. This is a crucial test because if $\alpha$ is a non-decreasing function of time, any variation should be larger at earlier times.
Methods
The authors employ a standard Bayesian inference technique using the Wilkinson Microwave Anisotropy Probe (WMAP) data. They compare the observed CMB temperature and polarization anisotropies to a set of flat cosmological models. A modified version of the CMBFAST code is used to compute these models, with parameters sampled over a range of values. The key parameter of interest is $\alpha_{dec}/\alpha_0$, the ratio of the fine-structure constant at decoupling to its present-day value. They perform a marginalized likelihood analysis, meaning they consider the probability distribution of $\alpha_{dec}/\alpha_0$ while averaging over the uncertainties in other cosmological parameters. They also explore the impact of reionization on these constraints.
Datasets
The primary dataset used in this paper is the first-year data from the Wilkinson Microwave Anisotropy Probe (WMAP) satellite, specifically its temperature and cross-polarization measurements. The authors also refer to experimental parameters for the Planck satellite (Table I) and discuss future "cosmic variance limited" (CVL) experiments, which represent an ideal scenario for CMB measurements.
Key Insights and Ideas
- WMAP provides tight constraints on $\alpha$ at decoupling: This is the first constraint on $\alpha$ at decoupling derived from a single CMB measurement, and it's already competitive with previous results that relied on multiple datasets.
- $\alpha$ appears to be constant since decoupling: The WMAP data is consistent with no significant variation in $\alpha$ between decoupling and today.
- Reionization is crucial for future constraints: The paper emphasizes that the reionization epoch, a period when the early universe became re-ionized, plays a significant role in improving constraints on $\alpha$. The "reionization bump" in the polarization spectrum is a key feature that can be exploited.
- Polarization data is more powerful for $\alpha$ constraints: While temperature data provides some constraints, polarization measurements, especially when combined with temperature data, offer much tighter limits on $\alpha$.
Key Results
- At 95% confidence level, the WMAP data alone constrains the variation of $\alpha$ at decoupling to be within 2% (for a positive variation) and 6% (for a negative variation) relative to its present-day value.
- The tightest possible constraint on $\alpha$ using CMB data alone is estimated to be around 0.1%.
- Future experiments like Planck are expected to improve constraints on $\alpha$ by a factor of 5, reaching a precision of about 0.34%.
Limitations and Future Work
The authors acknowledge that achieving constraints tighter than 0.1% will require incorporating additional (non-CMB) priors on other cosmological parameters. They also point out that while WMAP data alone is powerful, future, more accurate cosmic microwave background polarization data will be essential for significantly improving these constraints. The paper suggests that a post-Planck polarization experiment would be particularly valuable for this purpose. They also highlight that the current data is consistent with no variation, but the likelihood is skewed towards smaller values of $\alpha$ at decoupling, hinting at a potential direction for future investigation.
Summary updated