For SBBN, where S = 1 (N = 3) and e = 0, the primordial abundances of D, 3He, 4He, and 7Li are predicted as a function of only one free parameter, the baryon density parameter ( or B h2 B). As described above (see Section 4.1), D is the baryometer of choice. From SBBN and the adopted relic abundance of deuterium, yD = 2.6 ± 0.4, 10 = 6.1+0.7-0.5 (B h2 = 0.022 ± 0.002).
Having determined the baryon density to ~ 10% using D as the SBBN baryometer, it is incumbent upon us to compare the SBBN-predicted abundances of the other light nuclides with their relic abundances inferred from the observational data. For this baryon density, the predicted primordial abundance of 3He is y3 = 1.04 ± 0.10, in excellent agreement with the primordial value of y3 = 1.1 ± 0.2 inferred from observations of an outer-Galaxy HII region (Bania et al. 2002). Within the context of SBBN, D and 3He are completely consistent.
The first challenge to SBBN comes from 4He. For the SBBN-determined baryon density the predicted 4He primordial mass fraction is YP = 0.248 ± 0.001, to be compared with our adopted value from extragalactic HII regions (Olive, Steigman & Walker 2000) of YPOSW = 0.238 ± 0.005. Agreement is only at the ~ 2 level. Given the unresolved systematic uncertainties in determining YP from the HII region data, it is not clear at present whether this is a challenge to SBBN or to our understanding of HII region recombination spectra. As will be seen below, this tension between SBBN D and 4He can be relieved for nonstandard BBN if the assumption that S = 1 (N = 3) is relaxed.
The conflict with the inferred primordial abundance of lithium is even more challenging to SBBN. For yD = 2.6 ± 0.4, [Li] = 2.65+0.10-0.12. This is to be compared to the estimate (see Figure 4) of [Li] = 2.2 ± 0.1 based on a sample of metal-poor, halo stars. The conflict is even greater with the [Ryan et al. (2000)] estimate of [Li] = 2.09+0.19-0.13 derived from an especially selected data set. Unlike the tension between SBBN and the D and 4He abundances, the conflict between D and 7Li cannot be resolved by a nonstandard expansion rate (nor, by an electron neutrino asymmetry). Most likely, the resolution of this conflict is astrophysical since the metal-poor halo stars from which the relic abundance of lithium is inferred have had the longest time to mix their surface material with that in their hotter interiors, diluting or destroying their prestellar quota of lithium (see, e.g. [Pinsonneault et al. (2002)] and references to related work therein).
At present SBBN in combination with the limited data set of QSOALS deuterium abundances yields a ~ 10% determination of the baryon density parameter. Consistency between the inferred primordial abundances of D and 3He lends support to the internal consistency of SBBN, but the derived primordial abundances of 4He and 7Li pose some challenges. For 4He the disagreement is only at the ~ 2 level and the errors in the observationally inferred value of YP are dominated by poorly quantified systematics. However, if the current discrepancy is real, it might be providing a hint at new physics beyond the standard model (e.g. nonstandard expansion rate and/or nonstandard neutrino physics). Before considering the effects on BBN of a nonstandard expansion rate (S 1; N 3), we will compare the SBBN estimate of the baryon density parameter with that from the CBR.