WELP_EDM: velocity appears significantly different from those nearby.

0616_GPS: velocity appears significantly different from those nearby.

OJAI_GPS: velocity appears significantly different from those nearby.

FXHS_GPS: velocity appears significantly different from those nearby.

CCCO_GPS: velocity appears significantly different from those nearby.

LBCH_GPS: velocity appears significantly different from those nearby.

DYHS_GPS: velocity appears significantly different from those nearby.

BKMS_GPS: velocity appears significantly different from those nearby.

VYAS_GPS: velocity appears significantly different from those nearby.

HOLT_GPS: velocity appears significantly different from those nearby.

CABA_GLA: velocity may be biased by post-Landers transient

SAND_EDM, SAND_VLB, SAND_GPS:

   We have included pre-Landers velocities for this site because it is the
  only location close to that earthquake for which there is much
  pre-earthquake data.  Unfortunately the velocities do not agree well, and
  none are without possible problems.  There is potential for the _EDM
  result to be off if the Homestead Valley earthquake was not properly
  allowed for.  SAND_GPS has a short span for determining velocities.
  SAND_VLB has only 4 epochs:  in 1985 that with a large uncertainty, one
  in 1987.2 and two closely spaced at 1988.0; these do not well fit a
  linear trend.

JPLM_GJP: As has been noted by Heflin et al 1998 (Geophys. Res. Lett.,
  vol 25, pp. 93-96) the rate of this station has changed over time; it
  is not clear that a single velocity after the post-Northridge transient
  is an adequate representation.

    _VLB: There is a possible systematic difference between the VLBI and GPS
  velocities, which is difficult to evaluate because these two measurement
  programs (in Southern California) did not overlap much in time.