DOPPLER
Relative motion moves a carrier in frequency.
For speeds much smaller than the speed of light, the one-way Doppler shift of a carrier can be approximated by:
Δf / f₀ ≈ −vr / c
where f₀ is the emitted carrier frequency, Δf is the received frequency offset, vr is relative radial velocity under the chosen sign convention, and c is the speed of light.
The Deep Space Network uses the same basic physics operationally: relative transmitter-receiver motion shifts carrier frequency, and Doppler observables constrain range rate along the line of sight.
That is the first useful mental model for a drifting narrowband signal. Frequency is not floating randomly through a spectrogram. Relative kinematics can write geometry into the carrier.
DRIFT RATE
Changing Doppler shift points toward changing radial velocity.
If radial velocity changes, received frequency changes with time. To first order:
(1/f) df/dt ≈ −ar/c
where ar is relative radial acceleration.
This normalized quantity is useful because an identical line-of-sight acceleration produces a drift rate proportional to carrier frequency. A 1 Hz/s drift at one observing frequency is not kinematically equivalent to 1 Hz/s at a frequency ten times higher.
Normalizing drift by carrier frequency therefore helps compare candidate events observed in different bands:
normalized drift = (df/dt)/f
It does not by itself identify the transmitter. It expresses one part of the signal in a way tied directly to relative acceleration.
THE OBSERVER MOVES
Earth is not a stationary receiving platform.
A terrestrial telescope participates in Earth's rotation and Earth's orbit around the Sun. The observatory's line-of-sight velocity toward a fixed celestial direction therefore changes over time even if the transmitter itself has no local acceleration relative to an inertial frame.
The exact contribution depends on sky position, observatory location, epoch, and time. A narrowband extraterrestrial transmitter received on Earth can therefore show a predictable Doppler evolution produced partly by the receiver.
Conversely, a signal that appears stationary in a receiver-centered frame may require a transmitter or processing chain that compensates observer motion.
Any serious drift analysis must declare which reference frame the frequency track occupies and whether barycentric, topocentric, spacecraft-relative, or other corrections have been applied.
TRANSMITTER MOTION
The source can contribute its own velocity and acceleration.
A transmitter mounted on a rotating planet, orbiting satellite, spacecraft, binary companion, moon, surface vehicle, or other moving platform can add characteristic frequency structure.
Periodic radial velocity can create periodic Doppler modulation. Orbital motion can create a smoothly varying acceleration that changes sign through the orbit. Rotation can add shorter-period modulation. Maneuvers can create discontinuities. An intentionally stabilized transmitter may actively compensate some or all of these motions.
The same measured drift can therefore be consistent with many physical trajectories unless additional information constrains geometry.
This is why Cyberdelia's long-baseline idea does not treat one drift rate as a fingerprint. It treats drift, drift derivative, time, sky geometry, observing location, recurrence phase, and terrestrial ephemerides as a joint pattern.
SECOND DERIVATIVE
Curvature can carry more information than a straight drift line.
A short observation may make a frequency track look nearly linear. Over longer spans, changing acceleration can produce curvature:
d²f/dt²
That curvature can constrain how radial acceleration itself changes. In an orbital interpretation it may contain information about phase and period, although extracting that information is model-dependent and often degenerate.
For long-baseline recurrence work, the goal is not to fit exotic orbits to every curved trace. It is to ask whether multiple events show kinematic structure consistent with one physically constrained recurrence model better than with terrestrial, instrumental, or random alternatives.
KNOWN SPACECRAFT
Human transmitters are calibration targets, not merely contaminants.
Known satellites and spacecraft provide an excellent ground-truth family for testing Doppler reconstruction.
If historical ephemerides and transmission information are available, a pipeline can predict approximate line-of-sight velocity and compare predicted Doppler evolution with observed signal tracks. That tests coordinate transforms, time handling, observatory location, sign conventions, frequency normalization, and search tolerances using objects whose origin is known.
This is strategically useful for SETI because the terrestrial-rejection system and the hypothetical distant-transmitter system share much of the same kinematic machinery.
A pipeline that cannot reliably recognize known spacecraft geometry has no business making confident claims about unknown transmitters.
TLE LIMITS
Present orbital elements are not a time machine.
For Earth satellites, current two-line element sets can be useful for near-term propagation under their intended model. They are not automatically reliable for reconstructing precise positions years backward across drag, maneuvers, station-keeping, orbit maintenance, fragmentation, or element-update history.
Long-baseline interference reconstruction should prefer time-indexed historical orbital elements, state vectors, maneuver records, or ephemerides appropriate to each observation epoch.
Otherwise the pipeline risks rejecting a real terrestrial source because it propagated the wrong orbit through the past, or falsely associating a source because the model uncertainty was never represented.
CLOCKS
A drifting oscillator can impersonate motion.
Received frequency depends on more than celestial mechanics. Transmitter oscillator stability, receiver local oscillators, reference clocks, synthesizers, thermal behavior, calibration, and processing can all create or modify drift.
The Deep Space Network achieves precision by treating frequency standards and phase measurement as part of the instrument. SETI pipelines deserve the same respect for clocks.
An observed linear drift could contain true Doppler, local oscillator drift, transmitter instability, or a combination. A repeatable feature that tracks receiver hardware rather than sky geometry is especially suspicious.
Multi-beam observations, simultaneous reference channels, known-frequency calibrators, redundant instrumentation, and observations from independent sites can help separate sky-linked behavior from receiver-linked behavior.
RFI
Terrestrial interference has geometry too.
Radio-frequency interference is not merely a pile of static lines. Satellites move. aircraft move. ground transmitters appear through sidelobes and reflections. radar sweeps. oscillators drift. digitally generated signals hop or chirp. local electronics cycle with temperature and load.
A candidate filter that asks only whether the signal sits at a protected astronomical frequency is inadequate. A serious system uses time, direction, multiple beams, recurrence, observatory-local behavior, known transmitters, satellite geometry, modulation structure, and instrument context.
Rejected events should not necessarily be discarded. Aggregated RFI detections can become a useful environmental model that improves future rejection and may reveal recurring interferers that were previously treated as unrelated junk.
RECURRENCE
Long baselines turn isolated events into a graph.
Suppose several weak events occur years apart. Each alone is unremarkable. If they share compatible sky geometry, normalized drift, acceleration curvature, observing phase, or predicted recurrence windows, they may deserve joint analysis.
The useful object is not simply a list of detections. It is a recurrence graph in which events are connected by physically testable hypotheses.
For each proposed connection, ask:
Could one known satellite family explain both? Could observatory-local interference explain both? Is the required acceleration physically plausible? Does the model predict additional historical windows? Are compatible raw observations available in those windows? Does the predicted signal appear below the original threshold? Does the hypothesis survive when search criteria are frozen before the backsearch?
Prediction is the discipline that keeps recurrence analysis from becoming pattern worship.
BACKSEARCH
A hypothesis becomes interesting when it tells you where to look before you look.
If a recurrence or kinematic model predicts that a transmitter should cross a particular time-frequency region in an old observation, the archive can be searched specifically around that predicted track.
This can recover signals that were individually too weak to cross the original threshold, provided the search accounts honestly for how the prediction was constructed and how many alternatives were tried.
Backsearch is strongest when parameters are fit on one subset of events and evaluated on held-out archival windows. If every parameter is adjusted after seeing every candidate, the model can become an elaborate description of noise.
The result can still be negative. A predicted signal failing to appear is information. Good research packages preserve those kills rather than quietly moving the hypothesis to a different window.
FREQUENCY ACROSS BANDS
Same dynamics do not require the same carrier frequency.
If multiple transmitters or harmonics belong to the same moving platform, their absolute drift rates scale with carrier frequency while normalized drift may remain related through the common radial acceleration.
This suggests a broader comparison than “did the signal repeat at exactly the same frequency?”
Events at different frequencies can be tested for compatible normalized kinematics. That does not prove common origin. Independent terrestrial sources can share similar motion, and oscillator behavior complicates the picture. But it is a physically meaningful way to search across frequency without demanding identical carriers.
ERROR BUDGET
Drift precision is limited by more than spectral-bin width.
Uncertainty can enter through frequency estimation, time stamps, observatory coordinates, source direction, clock stability, ephemeris error, transmitter-frequency knowledge, acceleration approximation, integration length, signal-to-noise ratio, and spectral leakage.
A claimed kinematic match needs tolerances derived from those uncertainties. Otherwise a broad matching window can connect almost anything, while an unrealistically narrow window can reject the actual source.
The Deep Space Network literature treats Doppler tracking as a measurement with instrumental and external error sources. SETI should do the same rather than treating a spectrogram line as an exact trajectory.
FIELD METHOD
A disciplined drift-analysis pipeline.
1. Preserve the raw observation and instrument metadata. Do not begin with screenshots.
2. Identify frequency and time reference frames. Topocentric, barycentric, spacecraft-relative, corrected or uncorrected.
3. Estimate f, df/dt, and curvature where supported. Preserve uncertainty and estimator settings.
4. Normalize drift by carrier frequency. Compare kinematics across bands when physically meaningful.
5. Model observatory motion. Earth rotation, Earth orbit, location, epoch, and sky direction.
6. Cross-match terrestrial transmitters. Historical satellites/spacecraft, known RFI families, local recurrence, multiple beams, and independent sites.
7. Fit only physically interpretable recurrence hypotheses. Track degrees of freedom and alternative trials.
8. Predict held-out windows. Backsearch old data without retuning the model after every miss.
9. Publish negative results. A killed hypothesis improves the search space.
BOTTOM LINE
A drifting line is a kinematic clue, not a biography.
Before asking who transmitted it, ask what relative motion, clock behavior, and observing geometry are required to make it move through frequency that way.
If terrestrial and instrumental models fail under a reproducible analysis, the residual becomes interesting. The discipline comes from making those models fail first.
SOURCE TRAIL
Doppler tracking and radio measurement references.
JPL Deep Space Network — Doppler Tracking
NASA NTRS / JPL — Radiometric Spacecraft Tracking for Deep Space Navigation
NASA NTRS / JPL — Precision of Spacecraft Doppler Tracking at Low Signal-to-Noise Ratios