Model-independent test of the cosmic distance duality relation

A validation of the cosmic distance duality (CDD) relation, η(z)≡ (1+z)²d_A(z)/d_L(z) = 1, coupling the luminosity (d_L) and angular-diameter (d_A) distances, is crucial because its violation would require exotic new physics. We present a model-independent test of the CDD, based on strong lensing and a reconstruction of the HII galaxy Hubble diagram using Gaussian Processes, to confirm the validity of the CDD at a very high level of confidence. Using parameterizations η(z) = 1 + η₀z and η(z) = 1 + η_1z+ η₂z², our best-fit results are η₀= 0.0147^+0.056_-0.066, and η₁= 0.1091^+0.1680_-0.1568and η₂= -0.0603^+0.0999_-0.0988, respectively. In spite of these strong constraints, however, we also point out that the analysis of strong lensing using a simplified single isothermal sphere (SIS) model for the lens produces some irreducible scatter in the inferred CDD data. The use of an extended SIS approximation, with a power-law density structure, yields very similar results, but does not lessen the scatter due to its larger number of free parameters, which weakens the best-fit constraints. Future work with these strong lenses should therefore be based on more detailed ray-tracing calculations to determine the mass distribution more precisely.