Summary
Until recently, the IGS network of global tracking sites was predominently populated y Rogue/Turborogue GPS receivers. Firmware used in these receivers was found to perform poorly during periods of high ionospheric disturbances and, during 1999/2000, a significant number of receivers throughout the world underwent firmware upgrades that improved the performance of the receivers. The effect at some sites was dramatic, causing actual changes in the estimated heights of the sites and introducing discontinuities in the long-term height rate estimates. Here we provide examples of some sites where significant changes have occurred in the coordinate estimates of the sites and show that the time series should no longer be considered as a single time series; rather that the period before and after the firmware upgrade(s) should be considered to be of independent populations.