Because of the orbital period, the satellite follows the same track every day and is subject to forces that tend to pull the position of the ground track away from the ideal position. This is due to the Earth not being perfectly spherical and the satellites trying to move to positions of equilibrium in the gravitational field. It is the same effect that causes geosynchronous satellites to move away from their positions and having to be nudged back.

The following is a presentation of results from a concentrated period of tracking the active SDS satellites during 2008.

Seen from the UK, the Doppler curve exhibited by a satellite in a Molniya-type HEO is 'S' shaped with two well-defined points at the bottom and top of the 'S'. They represent repectively the minimum and maximum frequencies of the received Doppler-shifted signal. For an nine week period during 2008 July-August, I measured the times and received frequencies at Doppler Minimum and Doppler Maximum on the European orbital loop. I did this for all all three of the operational SDS satellites and the newly-demoted USA 125. Initially, the aim was to determine whether the timings would yield a value for the orbital period.

Interestingly, although it would prove difficult to measure the orbital period, results reveal that orbital adjustments are easy to detect and that the operational satellites' orbits are adjusted just about simultaneously.

The operational satellites at the time were USA 137, USA 179 and USA 198. Transmissions were also monitored from USA 125, which was replaced by USA 198 in 2008 March. All four satellites were tracked on Channel 9 of the USAF Space-Ground Link System (SGLS-9) at 2242.5 MHz

Timings did not yield highly-precise values for the orbital period. The orbit plane drifts slowly eastward or westward from day to day, depending on the actual value of the orbital period. The result is that, from a given location, Doppler Minimum and Doppler Maximum occur at slightly different points on the orbit on successive days. The time between occurrences does not therefore equate to the orbital period.

On each of the daily European orbital loops of the four satellites, I measured the time and received frequency at both Doppler Minimum and Doppler Maximum - eight observations per day. There were notable variations in both, due in part to the fact that the satellite transmitters drift in frequency and, if the transmission frequency is changing around the time of minimum or maximum Doppler Shift, it can shift the perceived time of the event. While frequency measurements are reasonably precise (to the nearest 50 Hz), timings are +/-10 seconds.

For plotting purposes, the interval between successive days' timings was converted to a pseudo value for the Mean Motion - 'n' in Twoline Orbital Element terms. The chart shows the plot for USA 137's Doppler Minimum timings. It shows four distinct phases with changes around July 20, August 5 and August 20. The change points indicate approximate dates on which orbital adjustments were made. The scatter is due in part to accuracy of measurement and may also indicate that minor 'tweaks' to the orbit are occurring. Similar plots result from observations of the other three satellites.

Because of day to day drift in the ground track, the time between Doppler Minimum and Doppler Maximum changes slowly, but exhibits a sudden change when the orbit is adjusted. This plot shows USA 137 (in blue) and USA 125 (in green). USA 137 shows changes around the same dates as in the Timing plot (above). USA 125 has a change too - around July 25. This indicates that it is under control and that that the adjustment engine was still working. There is an interesting, but unexplained, sinusoidal 'wiggle' in the second part of the USA 125 plot.

This is simply the difference in received frequency between the Doppler Minimum and Doppler Maximum points - values are +/-75 Hz. The plot shows all four satellites. USA 179 is again in blue, and the four phases are distinctly visible as in the plots above. USA 125 is again green, and USA 179 and USA 198 are purple and red respectively. All three of the operating satellites show changes around the same dates.

None of the plot types above shows absolutely when orbit adjustments occurred because of the scatter of points. However, by studying the data from them all for the whole constellation, it is possible to narrow down the date range to within +/-1 day. The following pattern merges: