Bound on the variation in the fine structure constant implied by Oklo data

Dynamical models of dark energy can imply that the fine structure constant $\alpha$ varies over cosmological time scales. Data on shifts in resonance energies $E_r$ from the Oklo natural fission reactor have been used to place restrictive bounds on the change in $\alpha$ over the last 1.8 billion years. We review the uncertainties in these analyses, focussing on corrections to the standard estimate of $k_\alpha\!=\!\alpha\,dE_r/d\alpha$ due to Damour and Dyson. Guided, in part, by the best practice for assessing systematic errors in theoretical estimates spelt out by Dobaczewski et al. [in J. Phys. G: Nucl. Part. Phys. 41, 074001 (2014)], we compute these corrections in a variety of models tuned to reproduce existing nuclear data. Although the net correction is uncertain to within a factor of 2 or 3, it constitutes at most no more than 25% of the Damour-Dyson estimate of $k_\alpha$. Making similar allowances for the uncertainties in the modeling of the operation of the Oklo reactors, we conclude that the relative change in $\alpha$ since the Oklo reactors were last active (redshift $z\simeq 0.14$) is less than $\sim 10$ parts per billion. To illustrate the utility of this bound at low-$z$, we consider its implications for the string theory-inspired runaway dilaton model of Damour, Piazza and Veneziano.