Interspecies clock comparison below $5 \times 10^{-18}$ uncertainty with a transportable clock

We report a measurement of the optical frequency ratio between the $^2\mathrm{S}_{1/2}(F=0)$--${^2\mathrm{F}_{7/2}(F=3)}$ electric-octupole (E3) transition of $^{171}$Yb$^{+}$ and the $^1\mathrm{S}_0$--${^3\mathrm{P}_0}$ transition of $^{87}$Sr, $\nu_{\mathrm{Yb}^{+}}/\nu_\mathrm{Sr} = 1.495\,991\,618\,544\,900\,588\,1(65)$. Reaching a fractional uncertainty of $4.3 \times 10^{-18}$, this result improves upon the previous best by more than a factor of three and is among the few that meet the requirements for interspecies clock comparisons specified by the roadmap towards the redefinition of the SI second. The comparison is between a transportable optical lattice clock and a stationary single-ion clock. It spans a period of nearly two years, during which the transportable clock was intermittently operated off-campus. The ratio was reproducibly measured during four separate campaigns, which are consistent within their statistical uncertainties. The results demonstrate reproducible $10^{-18}$ level operation of the transportable clock and thus validate its application for chronometric geodesy and as a transfer standard for inter-institute clock comparisons, e.g., in the absence of optical fiber links.