Long Distance Contribution to $s \to dγ$ and Implications for $Ω^-\to Ξ^-γ, B_s \to B_d^*γ$ and $b \to sγ$

We estimate the long distance (LD) contribution to the magnetic part of the $s \to d\gamma$ transition using the Vector Meson Dominance approximation $(V=\rho,\omega,\psi_i)$. We find that this contribution may be significantly larger than the short distance (SD) contribution to $s \to d\gamma$ and could possibly saturate the present experimental upper bound on the $\Omega^-\to \Xi^-\gamma$ decay rate, $\Gamma^{\rm MAX}_{\Omega^-\to \Xi^-\gamma} \simeq 3.7\times10^{-9}$eV. For the decay $B_s \to B^*_d\gamma$, which is driven by $s \to d\gamma$ as well, we obtain an upper bound on the branching ratio $BR(B_s \to B_d^*\gamma)<3\times10^{-8}$ from $\Gamma^{\rm MAX}_{\Omega^-\to \Xi^-\gamma}$. Barring the possibility that the Quantum Chromodynamics coefficient $a_2(m_s)$ be much smaller than 1, $\Gamma^{\rm MAX}_{\Omega^-\to \Xi^-\gamma}$ also implies the approximate relation $\frac{2}{3} \sum_i \frac{g^2_{\psi_i}(0)}{m^2_{\psi_i}} \simeq \frac{1}{2} \frac{g^2_\rho(0)}{m^2_\rho} + \frac{1}{6}\frac{g^2_\omega(0)}{m^2_\omega}$. This relation agrees quantitatively with a recent independent estimate of the l.h.s. by Deshpande et al., confirming that the LD contributions to $b \to s\gamma$ are small. We find that these amount to an increase of $(4\pm2)\%$ in the magnitude of the $b \to s \gamma$ transition amplitude, relative to the SD contribution alone.