Speaker
Description
Pulsar timing arrays probe the stochastic gravitational wave background through the angular cross-correlations of pulsar timing residuals. For an isotropic tensor background in general relativity, the expected overlap reduction function is the Hellings-Downs curve. We investigate how ghost-free massive gravity modifies this prediction through a massive dispersion relation and additional vector and scalar polarizations. We correct previous treatments of the vector and scalar modes and show that in the general relativistic limit, their leading contributions scale as $(1 - A^2)^{-1}$ and $(1 - A^2)^{-2}$ respectively. This motivates a phenomenological screening of the non-tensor sectors to ensure that we recover the Hellings-Downs in the massless graviton limit. Using the NANOGrav 15-year dataset, we perform a Bayesian model selection scan over graviton masses $m_g \in [10^{-25},\,8.17\times10^{-24}]\,\mathrm{eV}$ and scalar suppression indices $n_S \in \{2,4,6,8\}$ with vector suppression index $n_V = 1$ fixed. Among the tested parameters, the largest preference over Hellings-Downs occurs for $m_g = 5.82\times 10^{-24}$ eV and $n_S = 2$, with a Bayes factor $\mathcal{B}\mathcal{F}_{10} \simeq 14.7$, indicating strong preference. These results show that PTA angular correlations provide a concrete phenomenological test of massive gravity.