Evaluation of low-$Q^2$ fits to $ep$ and $ed$ elastic scattering data
We examined several low-$Q^2$ elastic $ep$ and $ed$ scattering data sets using various models to extract the proton and deuteron rms radii and developed a comprehensive algorithm for estimating the systematic bias of each extracted radius. In each case, we chose the model and upper bound for $Q^2$ that minimized the combination of statistical uncertainty and bias. We attribute the discrepancy between small ($\approx 0.84$~fm) and large ($\approx 0.88$~fm) proton radius extractions to the absence of data that can accurately isolate the linear and quadratic contributions to the form factor at low-$Q^2$. In light of this ambiguity, we estimated the asymmetric model-dependence of each extracted radius by studying the distribution of many possible fits. The resulting radii are 0.842(4)~fm for the proton and 2.092(19)~fm for the deuteron. These values are consistent with those determined from muonic hydrogen measurements within 0.3$σ$ for the proton and 1.8$σ$ for the deuteron.