A close up of Doppler Shift

Doppler shift is a phenomenon which is commonly observed by the lay person, yet still confuses many amateur satellite operators. This page intends to de-mystify Doppler and presents typical Doppler scenarios for low Earth orbiting satellite operation.

Many amateurs have been confused by the Doppler shift encountered during satellite operation. In essence, Doppler shift is simply the apparent change in frequency that is observed when an object moves towards or away from an observer. It is a property common to wave propagation. Almost everyone has experienced Doppler shift in everyday life. Here's a couple of common examples:

1. You are waiting at a level crossing for a train to pass. The train is an express train, and is moving at high speed. As the train approaches, the driver sounds the horn and keeps it on until the locomotive is well past the level crossing. Sitting in your car, you hear the pitch of the horn go lower as the train passes and moves away. At the same time, a passenger on the train hears the horn. However, the horn's pitch doesn't appear to vary, but the bells of the level crossing appear to become lower in pitch as the passenger's carriage passes the level crossing. Sound familiar?

2. You are walking along a main road when an ambulance passes with its siren on. As the ambulance goes past where you're walking, the pitch of the siren becomes lower.

3. You live near a small airport, where the aircraft are mainly propeller driven. A plane takes off on a course over your roof. As it passes, the note of the engine becomes deeper (like the sound effects in old war movies!).

4. Similarly, the "VRRROOOOOOOOOOOMMMM" sounds that kids make when imitating cars driving past at high speed was learnt from their (unconscious) observations of the effects of Doppler shift on the sound of the engines as cars drive past (anyone listened closely to Formula 1 telecasts on the TV?). :-)

There's just a few everyday examples.

Similarly, radio waves are affected by Doppler shift. However, because the speed of light (and therefore radio waves) is much higher than that of sound, everyday Doppler shifts are quite small, maybe a few Hz for a VHF station mobile in a car. The only terrestrial operators who would normally notice Doppler shift are those who attempt SSB mobile on the microwave bands, and those who work aurora scatter. Doppler shift is proportional to the frequency of operation, so it becomes more significant on the higher bands.

Satellites travel at much higher speeds, typically 27,000 km/h for a low Earth orbiting satellite. At this speed, Doppler shift becomes very significant for SSB operators on all satellite bands (21 MHz and up), and is noticeable to FM operators on 145 MHz and must be compensated for on any higher band. I have mentioned techniques for Doppler compensation in other articles, so instead, I'll present information on typical Doppler effects encountered on LEO amateur satellites.

The following information assumes a satellite in a circular orbit at an altitude of 800 km. This is a common orbit for amateur satellites (in fact, the raw data was obtained from pass predictions of UO-14, which is in such an orbit). The data was collected by using Winorbit to generate tables of Doppler shift against time every 5 seconds for various elevation passes. The raw data was then plotted in Excel to give a graphical presentation of Doppler shift as a pass progresses. Finally, the frequency shifts were scaled from 70cm (UO-14's downlink band) to several amateur bands, to show the effect of carrier frequency on Doppler.

Firstly, the amount of Doppler shift for LEO at 800 km varies within these ranges:

Table 1. Maximum Doppler Shift Vs Frequency for Popular Amateur Bands for an LEO at 800km Altitude.

The table above shows how Doppler shift increases with frequency. For SSB/CW, it should be obvious that Doppler will significantly impact operations on any of the bands given, and must be compensated for. However, for FM, the Doppler shift on 2 metres (3.27 kHz) is still small enough to be workable (with some distortion) on a fixed frequency receiver. On 70cm, even FM receivers much be retuned 3 or 4 times during a typical pass. By the time one gets to 10 GHz, only wideband modes and/or computer controlled stations would be able to cope with the severe Doppler shift one would encounter. However, 10 GHz isn't used on any current LEOs, but will be active on Phase 3D, where the higher orbit and slower satellite motion will mean the Doppler shift will be less in magnitude, and less variable over a given short time period. Just for comparison, here's typical Doppler shifts for a car travelling at 100 km/h. Hardly enough to keep you reaching for the VFO dial, unless you operate on bands over 23cm, but 10 GHz mobile SSB would be interesting indeed! :-)

Table 2. Maximum Doppler Shift Vs Frequency for Popular Amateur Bands for a car travelling at 100 km/h.

The other issue with Doppler shift is how it varies during the satellite pass itself. This is highly dependent on the pass itself. As the graphs show, low elevation passes have a fairly linear variation of Doppler shift, spread out over the pass. On the other hand, a pass directly overhead has most of the Doppler shift variation concentrated around the middle of the pass, which means a relaxed start and finish, but lots of VFO twiddling around mid pass! :) Below are 4 graphs of Doppler shift Vs time for 4 different UO-14 passes (11 degree, 20 deg, 47 deg and overhead). The downlink (on which the Doppler was calculated) is on 435.070 MHz, and the graphs show how far removed the actual frequency received on the ground would be from the nominal carrier frequency. Note the shape of the curve for each pass.

With the above graphs, notice how the Doppler shift starts positive early in the pass (i.e. when the satellite approaches), and ends up at the maximum negative value at the end of the pass. This is just like the everyday examples of Doppler shift with sounds given at the start of this article. On the overhead pass (the bottom graph), the Doppler shift takes over 5 minutes to change by 2 kHz. The next 2 kHz takes just under a minute, and by mid pass, the rate of change is around 5.8 kHz/minute. Compare this behaviour to the 11 degree pass in the top graph, which shows a much flatter curve spread across the whole pass.

Conclusion:

Doppler shift is inherent in any mobile radio communications, but is only significant to amateurs at extremely high frequencies, or for satellite operation. All satellite operators (with the possible exception of SO-35 parrot users) need to take Doppler into account.

Satellite Doppler shift doesn't vary in a simple linear fashion during a pass, but instead has the greatest rate of change at mid pass. Overhead passes show more extreme variation in the rate of change of Doppler shift than low elevation passes. The implication is that Doppler compensation is best done either manually or by computer control, as simple linear frequency changes over time will not track the Doppler well.

Email me