Near Flat Curvature

The tension between the CMB-alone preference for slight negative curvature (Ω_K ≈ −0.04) and the flat universe demanded by standard inflation represents a consistency check that ΛCDM fails. When BAO measurements are added to Planck CMB data, they pull Ω_K toward zero — but this combination creates an internal tension between the CMB and BAO datasets themselves, with the two probes preferring different values of the curvature at the 2–3σ level. A self-consistent ΛCDM fit should find that all data sets agree on a single cosmological parameter set; the curvature parameter instead shifts depending on which data is included.

More fundamentally, the 'near flatness' problem is that ΛCDM's favored value |Ω_K| < 0.005 requires either inflation (which predicts flatness but introduces its own theoretical problems) or miraculous fine-tuning of initial conditions. The model offers no mechanism that would naturally produce a universe with curvature indistinguishable from zero at our current epoch without being exactly zero for all time. Any small deviation from flatness grows with time, so the early universe would have needed to be flat to one part in ~10⁶⁰ at the Planck era. This is the flatness problem restated in terms of the observed current curvature value, and ΛCDM has no solution beyond appealing to inflation.