We used HST to obtain surface brightness fluctuation ( SBF ) observations of four nearby brightest cluster galaxies ( BCG ) to calibrate the BCG Hubble diagram of \markcite lp92 Lauer & Postman ( 1992 ) .
This BCG Hubble diagram contains 114 galaxies covering the full celestial sphere and is volume limited to 15,000 km s ^ { -1 } , providing excellent sampling of the far-field Hubble flow .
The SBF zero point is based on the Cepheid calibration of the ground I _ { KC } method \markcite t97 ( Tonry et al .
1997 ) as extended to the WFPC2 F814W filter by \markcite a97 Ajhar et al .
( 1997 ) .
The BCG globular cluster luminosity functions give distances essentially identical to the SBF results .
Using the velocities and SBF distances of the four BCG alone gives H _ { 0 } = 82 \pm 8 { ~ { } km~ { } s ^ { -1 } ~ { } Mpc ^ { -1 } } in the CMB frame , valid on \sim 4,500 km s ^ { -1 } scales .
Use of BCG as photometric redshift estimators allows the BCG Hubble diagram to be calibrated independently of recession velocities , yielding a far-field H _ { 0 } = 89 \pm 10 { ~ { } km~ { } s ^ { -1 } ~ { } Mpc ^ { -1 } } with an effective depth of \sim 11,000 km s ^ { -1 } .
The error in this case is dominated by the photometric cosmic scatter in using BCG as distance estimators .
The concordance of the present results with other recent H _ { 0 } determinations , and a review of theoretical treatments on perturbations in the near-field Hubble flow , argue that going to the far-field removes an important source of uncertainty , but that there is not a large systematic error to be corrected for to begin with .
Further improvements in H _ { 0 } depend more on understanding nearby calibrators than on improved sampling of the distant flow .