Having indirectly probed the nature of matter in the Universe using the previous estimates, it is now time to turn to direct constraints that have been derived in the past decade. Here, perhaps more than any other area of observational cosmology, new observations have changed the way we think about the Universe.
4.1. The Baryon Density: a re-occuring crisis?:
The success of Big Bang Nucleosynthesis in predicting in the cosmic abundances of the light elements has been much heralded. Nevertheless, the finer the ability to empirically infer the primordial abundances on the basis of observations, the greater the ability to uncover some small deviation from the predictions. Over the past five years, two different sets of observations have threatened, at least in some people's minds, to overturn the simplest BBN model predictions. I believe it is fair to say that most people have accepted that the first threat was overblown. The concerns about the second have only recently subsided.
i. Primordial Deuterium: The production of primordial deuterium during BBN is a monotonically decreasing function of the baryon density simply because the greater this density the more efficiently protons and neutrons get processed to helium, and deuterium, as an intermediary in this reactions set, is thus also more efficiently processed at the same time. The problem with inferring the primordial deuterium abundance by using present day measurements of deuterium abundances in the solar system, for example, is that deuterium is highly processed (i.e. destroyed) in stars, and no one has a good enough model for galactic chemical evolution to work backwards from the observed abundances in order to adequately constrain deuterium at a level where this constraint could significantly test BBN estimates.
Three years ago, the situation regarding deuterium as a probe of BBN changed dramatically, when David Tytler and Scott Burles convincingly measured the deuterium fraction in high redshift hydrogen clouds that absorb light from even higher redshift quasars. Because these clouds are at high redshift, before significant star formation has occurred, little post BBN deuterium processing is thought to have taken place, and thus the measured value gives a reasonable handle on the primordial BBN abundance. The best measured system  yields a deuterium to hydrogen fraction of
|(D/H) = (3.3. ± 0.5)×10-5 (2)||(3)|
This, in turn, leads to a contraint on the baryon fraction of the Universe, via standard BBN,
|B h2 = .0190 ± .0018 (2)||(4)|
where the quoted uncertainty is dominated by the observational uncertainty in the D/H ratio, and where H0 = 100h. Thus, taken at face value, we now know the baryon density in the universe today to an accuracy of about 10%!
When first quoted, this result sent shock waves through some of the BBN community, because this value of B is only consistent if the primordial helium fraction (by mass) is greater than about 24.5%. However, a number of previous studies had claimed an upper limit well below this value. After the dust has settled, it is clear that these previous claims are likely to under-estimated systematic observational effects. Recent studies, for example, place an upper limit on the primordial helium fraction closer to 25%.
In any case, even if somehow the deuterium estimate is wrong, one can combine all the other light element constraints to produce a range for b h2 consistent with observation:
|B h2 = .016 - 0.025||(5)|
ii. CMB constraints: Beyond the great excitement over the observation of a peak in the CMB power spectrum at an angular scale corresponding to that expected for a flat universe lay some excitement/concern over the small apparent size of the next peak in the spectrum, at higher multipole moment (smaller angular size). The height of the first peak in the CMB spectrum is related to a number of cosmological parameters and thus cannot alone be used to constrain any one of them. However, the relative height of the first and second peaks is strongly dependent on the baryon fraction of the universe, since the peaks themselves arise from compton scattering of photons off of electrons in the process of becoming bound to baryons. Analyses of the two first small-scale CMB results produces a claimed constraint :
|B h2 = .032 ± .009 (2)||(6)|
Depending upon how you look at this, this is either a stunning confirmation that the overall scale for B predicted by simple BBN analyses is correct, or a horrible crisis, in which the two constraints, one from primordial deuterium, and one from CMB observations, disagree at the two sigma level. Given the history of this subject, the former response was perhaps most appropriate. In particular, the Maxima and Boomerang results were the very first to probe this regime, and first observations are often suspect, and in addition, the CMB peak heights do have a dependence on other cosmological parameters which must be fixed in order to derive the above constraint on B. Fortunately, more recent data has confirmed the view that one should not assume the sky is falling the first time around. Boomerang, along with several other experiments have most recently reported new data in which the second peak fits precisely where one would expect it to be based on BBN predictions.
This gives additional support for the assumption that the Burles and Tytler limit on B h2 is correct, and taking the range for H0 given earlier, one derives the constraint on B of
|B = .045 ± 0.15||(7)|